Proceedings of Biological Control of Weeds
Uploaded by
mroslifrimProceedings of Biological Control of Weeds
Uploaded by
mroslifrimProceedings of the
XII International Symposium on
Biological Control of Weeds
La Grande Motte, France, 2227 April 2007
Edited by M.H. Julien, R. Sforza, M.C. Bon, H.C. Evans, P.E. Hatcher, H.L. Hinz and B.G. Rector
CABI is a trading name of CAB International
CABI Head Office
Nosworthy Way
Wallingford
Oxfordshire OX 10 8DE
UK
Tel: +44 (0) 1491 832111
Fax: +44 (0) 1491 833508
Email: [email protected]
Web site: www.cabi.org
CABI North American Office
875 Massachusetts Avenue
7th Floor
Cambridge, MA 02139
USA
Tel: +1 617 395 4056
Fax: +1 617 354 6875
Email: [email protected]
CAB International 2008. All rights reserved. No part of this publication may be reproduced in any form or by
any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission
of the copyright owners.
ISBN-13: 978-1-84593-502-3 (paperback edition)
ISBN-13: 978-1-84593-506-1 (hardback edition)
Typeset by MTC, Manila, Philippines.
Printed and bound in the UK by Cambridge University Press, Cambridge.
How to cite:
Authors (2008) title. In Proceedings of the XII International Symposium on Biological Control
of Weeds (eds. Julien, M.H., Sforza, R., Bon, M.C., Evans, H.C., Hatcher, P.E., Hinz, H.L. and
Rector, B.G.), pp. xxx xxx. CAB International Wallingford, UK.
Contents
Preface
Theme 1: Ecology and Modelling in Biological Control of Weeds
xix
1
Papers
Is modelling population dynamics useful for anything other than keeping a
researcher busy? [Keynote paper]
Y.M. Buckley
Biomass reduction of Euphorbia esula/virgata by insect/bacterial combinations
A.J. Caesar and R.J. Kremer
Rhizosphere bacterial communities associated with insect root herbivory of an invasive plant,
Euphorbia esula/virgata
A.J. Caesar and T. Caesar-Ton That
3
7
13
The endophyte-enemy release hypothesis: implications for classical biological control and plant invasions
H.C. Evans
20
Multiple-species introductions of biological control agents against weeds: look before you leap
F.A.C. Impson, V.C. Moran, C. Kleinjan, J.H. Hoffmann and J.A. Moore
26
Clipping the butterfly bushs wings: defoliation studies to assess the likely impact of a folivorous weevil
D.J. Kriticos, M.S. Watt, D. Whitehead, S.F. Gous, K.J. Potter and B. Richardson
32
Can a pathogen provide insurance against host shifts by a biological control organism?
P.B. McEvoy, E. Karacetin and D.J. Bruck
37
Which haystack? Climate matching to narrow the search for weed biological control agents
M.P. Robertson, C. Zachariades and D.J. Kriticos
43
Nutritional characteristics of Hydrilla verticillata and its effect on two biological control agents
J.F. Shearer, M.J. Grodowitz and J.E. Freedman
44
How sensitive is weed invasion to seed predation?
R.D. van Klinken, R. Colasanti and Y.M. Buckley
52
Abstracts
Altered nutrient cycling as a novel non-target effect of weed biocontrol
I.E. Bassett, J. Beggs and Q. Paynter
56
Interactions of plant quality and predation affect the success of purple loosestrife biocontrol programme
A. Dvalos and B. Blossey
56
An arthropod and a pathogen in combination as biocontrol agents: how do they shape up?
L. Buccellato, E.T.F. Witkowski and M.J. Byrne
57
Impact of invasive exotic knotweeds (Fallopia spp.) on invertebrate communities
E. Gerber, U. Schaffner, C. Krebs, C. Murrell and M. Moretti
57
An experimental test of the importance of climate matching for biological control introductions
F.S. Grevstad, C.E. OCasey, M.L. Katz and K.H. Laukkenen
58
Effect of climate on biological control: a case study with diffuse knapweed in British Columbia, Canada
C.A.R. Jackson, J.H. Myers, S.R. White and A.R.E. Sinclair
58
XII International Symposium on Biological Control of Weeds
The IRA and getting the result you want
M.K. Kay
59
Microclimate effects on biological control: water hyacinth in South Africa
A.M. King, M.P. Hill, M. Robertson and M.J. Byrne
59
Habitat analysis of the rush skeleton weed root moth, Bradyrrhoa gilveolella (Lepidoptera: Pyralidae)
J.L. Littlefield, G.P. Markin, J. Kashefi and H.D. Prody
60
Evaluating the performance of Episimus utilis (Lepidoptera: Tortricidae) on the invasive
Brazilian peppertree in Florida
V. Manrique, J.P. Cuda, W.A. Overholt and D. Williams
60
Successful biological control of diffuse knapweed in British Columbia, Canada
J.H. Myers, H. Quinn, C.A.R. Jackson and S.R. White
61
An integrated approach to invasive plant management: biocontrol and native plant interactions
J.G. Nachtrieb, M.J. Grodowitz, R.M. Smart and C.S. Owens
61
Impact of host-plant water stress on the interaction between Mecinus janthinus and Linaria dalmatica
A.P. Norton
62
Impact of insect herbivory on dispersal in Hydrilla verticillata (L.f.) Royle
C.S. Owens, M.J. Grodowitz and R.M. Smart
62
Dynamics of invasive plant monocultures after the establishment of natural enemies: an example from the
Melaleuca quinquenervia system in Florida
M.B. Rayamajhi, P.D. Pratt, T.K. Van and T.D. Center
63
Modelling of Diorhabda elongata dispersal during the initial stages of establishment for the
control of Tamarix spp.
J. Sanabria, C.J. DeLoach, J.L. Tracy and T.O. Robbins
63
Seed feeders: why do so few work and can we improve our selection decisions?
R.D. van Klinken, R. Colasanti and G. Maywald
64
Theme 2: Benefit/RiskCost Analyses
65
Papers
Return on investment: determining the economic impact of biological control programmes [Keynote paper]
R. McFadyen
67
Post-release non-target monitoring of Mogulones cruciger, a biological control agent
released to control Cynoglossum officinale in Canada
J.E. Andreas, M. Schwarzlnder, H. Ding and S.D. Eigenbrode
75
Assessing indirect impacts of biological control agents on native biodiversity: a communitylevel approach
L.G. Carvalheiro, Y.M. Buckley, R. Ventim and J. Memmott
83
Factors affecting oviposition rate in the weevil Rhinocyllus conicus on non-target
Carduus spp. in New Zealand
R. Groenteman, D. Kelly, S.V. Fowler and G.W. Bourdt
87
Fortieth anniversary review of the CSIRO European Laboratory: does native range research
provide good return on investment?
A.W. Sheppard, D.T. Briese, J.M. Cullen, R.H. Groves, M.H. Julien, W.M. Lonsdale, J.K. Scott
and A.J. Wapshere
91
Abstracts
F1 sterility: a novel approach for risk assessment of biocontrol agents in open-field trials
J.E. Carpenter and C.D. Tate
iv
101
Contents
Impact of biocontrol agents on native biodiversity: the case of Mesoclanis polana
L.G. Carvalheiro, Y.M. Buckley and J. Memmott
A look at host range, host specificity and non-target safety from the perspective of a plant virus
as a weed-biocontrol agent
R. Charudattan, M. Elliott, E. Hiebert and J. Horrell
Novel approaches for risk assessment: feasibility studies on temporary reversible releases
of biocontrol agents
J.P. Cuda, O.E. Moeri, W.A. Overholt, V. Manrique, S. Bloem, J.E. Carpenter, J.C. Medal
and J.H. Pedrosa-Macedo
101
102
102
A wolf in sheeps clothing: potential dangers of using indigenous herbivores as biocontrol agents
J. Ding and B. Blossey
103
Impact of biological control of Salvinia molesta in temperate climates on biodiversity conservation
B.R. Hennecke and K. French
103
Opening Pandoras box? Surveys for attack on non-target plants in New Zealand
Q. Paynter, S.V. Fowler, A.H. Gourlay, M.L. Haines, S.R. Hona, P.G. Peterson, L.A. Smith,
J.R.A. Wilson-Davey, C.J. Winks and T.M. Withers
104
New biological control agents for Cytisus scoparius (Scotch broom) in New Zealand:
dealing with the birds and the bees and predicted non-target attack to a fodder crop
Q. Paynter, A.H. Gourlay, P.G. Peterson, J.R.A. Wilson-Davey, J.V. Myers, S.R. Hona and S.V. Fowler
104
Predicting risk and benefit a priori in weed biological control: a systems modelling approach
S. Raghu, K. Dhileepan and J. Scanlan
105
Comparative risk assessment of Linaria dalmatica and L. vulgaris biological control
S.E. Sing and R.K. Peterson
105
Theme 3: Target and Agent Selection
107
Papers
Latin American weed biological control science at the crossroads [Keynote paper]
R.W. Barreto
Galling guilds associated with Acacia dealbata and factors guiding selection of potential
biological control agents
R.J. Adair
109
122
Biological control of Miconia calvescens with a suite of insect herbivores from Costa Rica and Brazil
F.R. Badenes-Perez, M.A. Alfaro-Alpizar, A. Castillo-Castillo and M.T. Johnson
129
Giving dyers woad the blues: encouraging first results for biological control
G. Cortat, H.L. Hinz, E. Gerber, M. Cristofaro, C. Tronci, B.A. Korotyaev and L. Gltekin
133
Herbivores associated with Arundo donax in California
T.L. Dudley, A.M. Lambert, A. Kirk and Y. Tamagawa
138
Which species of the thistle biocontrol agent Trichosirocalus are present in New Zealand?
R. Groenteman, D. Kelly, S.V. Fowler and G.W. Bourdt
145
Bionomics and seasonal occurrence of Larinus filiformis Petri, 1907 (Coleoptera: Curculionidae)
in eastern Turkey, a potential biological control agent for Centaurea solstitialis L.
L. Gltekin, M. Cristofaro, C. Tronci and L. Smith
All against one: first results of a newly formed foreign exploration consortium for the biological
control of perennial pepperweed
H.L. Hinz, E. Gerber, M. Cristofaro, C. Tronci, M. Seier, B.A. Korotyaev, L. Gltekin, L. Williams
and M. Schwarzlnder
150
154
XII International Symposium on Biological Control of Weeds
Potential biological control agents for fumitory (Fumaria spp.) in Australia
M. Jourdan, J. Vitou, T. Thomann, A. Maxwell and J.K. Scott
160
Expanding classical biological control of weeds with pathogens in India: the way forward
P. Sreerama Kumar, R.J. Rabindra and C.A. Ellison
165
Explorations in Central Asia and Mediterranean basin to select biological control agents for Salsola tragus 173
F. Lecce, A. Paolini, C. Tronci, L. Gltekin, F. Di Cristina, B.A. Korotyaev, E. Colonnelli,
M. Cristofaro and L. Smith
Eriophyoid mites on Centaurea solstitialis in the Mediterranean area
R. Monfreda, E. de Lillo and M. Cristofaro
178
Diclidophlebia smithi (Hemiptera: Psyllidae) a potential biological agent for Miconia calvescens
E.G.F. Morais, M.C. Picano, R.W. Barreto, G.A. Silva, M.R. Campos and R.B. Queiroz
182
A lace bug as biological control agent of yellow starthistle, Centaurea solstitialis L. (Asteraceae):
an unusual choice
A. Paolini, C. Tronci, F. Lecce, R. Hayat, F. Di Cristina, M. Cristofaro and L. Smith
189
Pathogens from Brazil for classical biocontrol of Tradescantia fluminensis
O.L. Pereira, R.W. Barreto and N. Waipara
195
Field and laboratory observations of the life history of the Swiss biotype of Longitarsus
jacobaeae (Coleoptera: Chrysomelidae)
K.P. Puliafico, J.L. Littlefield, G.P. Markin and U. Schaffner
200
Fungal survey for biocontrol agents of Ipomoea carnea from Brazil
D.J. Soares and R.W. Barreto
206
Biological control of lippia (Phyla canescens): surveys for the plant and its natural enemies in Argentina
A.J. Sosa, M.G. Traversa, R. Delhey, M. Kiehr, M.V. Cardo and M.H. Julien
211
Potential biological control agents of field bindweed, common teasel and field dodder from Slovakia
P. Tth, M. Tthova and L. Cag
216
Lewia chlamidosporiformans, a mycoherbicide for control of Euphorbia heterophylla: isolate
selection and mass production
B.S. Vieira, K.L. Nechet and R.W. Barreto
221
Sphenoptera foveola (Buprestidae) as a potential agent for biological control of skeletonweed,
Chondrilla juncea
M.G. Volkovitsh, M.Yu Dolgovskaya, S.Ya Reznik, G.P. Markin, M. Cristofaro and C. Tronci
227
Common buckthorn, Rhamnus cathartica L.: available feeding niches and the importance of
controlling this invasive woody perennial in North America
M.V. Yoder, L.C. Skinner and D.W. Ragsdale
232
Evaluation of Fusarium as potential biological control against Orobanche on Faba bean in Tunisia
M. Zouaoui Boutiti, T. Souissi and M. Kharrat
238
Abstracts
Prospective biological control agents for Nassella neesiana in Australia and New Zealand
F.E. Anderson, J. Barton and D.A. McLaren
245
Biological control of Cirsium arvense by using native insects
G.A. Asadi, R. Ghorbani, M.H. Rashed and H. Sadeghi
245
The degree of polymorphism in Puccinia punctiformis virulence and Cirsium arvense resistance:
implications for biological control
M.G. Cripps, G.R. Edwards, N.W. Waipara, S.V. Fowler and G.W. Bourdt
vi
246
Contents
Field exploration for saltcedar natural enemies in Egypt
M. Cristofaro, F. Di Cristina, E. Colonnelli, A. Zilli and W.M. Amer
The phytophagous insects associated with spotted knapweed (Centaurea maculosa Lam.)
in northeast Romania
A. Diaconu, M. Talmaciu, M. Parepa and V. Cozma
246
247
Parkinsonia dieback: a new association with potential for biological control
N. Diplock, V. Galea, R.D. van Klinken and A. Wearing
247
Ecology, impact and biological control of the weed Tradescantia fluminensis in New Zealand
S.V. Fowler, N.W. Waipara, J.H. Pedrosa-Macedo, R.W. Barreto, H.M. Harman, D. Kelly,
S. Lamoureaux and C.J. Winks
248
Potential for biological control of Rhamnus cathartica and Frangula alnus in North America
A. Gassmann, I. Tosevski and L.C. Skinner
248
Arundo donax (giant reed): an invasive weed of the Rio Grande Basin
J. Goolsby, A. Kirk, W. Jones, J. Everitt, C. Yang, P. Parker, D. Spencer, A. Pepper, J. Manhart,
D. Tarin, G. Moore, D. Watts and F. Nibling
249
Potential agents from Kazakhstan for Russian Olive biocontrol in USA
R.V. Jashenko, I.D. Mityaev and C.J. DeLoach
249
Biology of the Rumex leaf defoliator sawfly Kokujewia ectrapela Konow (Hymenoptera: Argidae)
in Urmia region
Y. Karimpour
What defines a host? Growth ratethe paradox revisited
M.K. Kay
Selection of fungal strains for biological control of important weeds in the Krasnodar
region of Russia
T.M. Kolomiets, E.D. Kovalenko, Zh.. Mukhina, S.N. Lekomtseva, .V. Alexandrova,
O.o. Skatenok, I.Uj. Samokhina, L.F. Pankratova, D.K. Berner and S.A. Volkova
250
250
251
Vegetative expansion and seed output of swallow-worts (Vincetoxicum spp.)
L.R. Milbrath, K.M. Averill and A. DiTommaso
251
A new biological control program for common tansy (Tanacetum vulgare) in Canada and the USA
A.S. McClay, M. Chandler, U. Schaffner, A. Gassmann and G. Grosskopf
252
Surveys in Argentina for the biological control of Brazilian peppertree in the USA
F. McKay, G. Cabrera Walsh, M.I. Oleiro and G.S. Wheeler
252
Natural enemies of balloon vine and pompom weed in Argentina: prospects for biological
control in South Africa
F. McKay, M.I. Oleiro, A. McConnachie and D.O. Simelane
Tamarix biocontrol in US: new biocontrol agents from Kazakhstan
I.D. Mityaev, R.V. Jashenko and C.J. DeLoach
Biological control of aquatic weeds by Plectosporium alismatis, a potential mycoherbicide in
Australian rice crops: comparison of liquid culture media for their ability to produce high yields
of desiccation-tolerant propagules
C. Moulay, S. Cliquet, K. Zeehan, G.J. Ash and E.J. Cother
Herbivorous insects from Brazil for classical biocontrol of Tradescantia fluminensis
J.H. Pedrosa-Macedo, S.V. Fowler, M. Silvrio, K. Doetzer, M. Livramento and L. Suzuki
Nigrospora oryzae, a potential bio-control agent for Giant Parramatta Grass (Sporobolus fertilis)
in Australia
S. Ramasamy, D. Officer, A.C. Lawrie and D.A. McLaren
vii
253
253
254
254
255
XII International Symposium on Biological Control of Weeds
Biological control and ecology of the submerged aquatic weed Cabomba caroliniana
S.S. Schooler, G.C. Walsh and M.H. Julien
255
Hindsight is 20/20: improved biological control of Chromolaena odorata (Asteraceae)
for seasonally dry regions
L.W. Strathie, C. Zachariades, O. Delgado and C. Duckett
256
Surveys for herbivores of Casuarina spp. in Australia for development as biological control
agents in Florida, USA
G.S. Taylor, G.S. Wheeler and M.F. Purcell
256
Differential host preferences of Diorhabda elongata: implications for biological control of Tamarix
H.Q. Thomas
257
Hybridization potential of Saltcedar leaf beetle, Diorhabda elongata, ecotypes
D.C. Thompson, B.A. Petersen, D.W. Bean and J.C. Keller
257
Pathogens as potential classical biological control agents for alligator weed, Alternanthera philoxeroides
M.G. Traversa, M. Kiehr, R. Delhey, A.J. Sosa and M.H. Julien
258
A survey for fungal pathogens with potential for biocontrol of exotic woody Fabaceae in Argentina
M.G. Traversa, M. Kiehr and R. Delhey
258
Applied biocontrol, a landscape comparison of two Dalmatian toadflax agents
S.C. Turner
259
Survey of European natural enemies of Swallow-worts (Vincetoxicum spp.)
A.S. Weed, R. Casagrande and A. Gassmann
259
Climate matching and field ecology of Australian Bluebell Creeper
A.M. Williams, H. Spafford Jacob and E. Bruzzese
260
Theme 4: Pre-release Specificity and Efficacy Testing
261
Papers
The importance of molecular tools in classical biological control of weeds: two case studies
with yellow starthistle candidate biocontrol agents
G. Antonini, P. Audisio, A. De Biase, E. Mancini, B.G. Rector, M. Cristofaro, M. Biondi,
B.A. Korotyaev, M.C. Bon, A. Konstantinov and L. Smith
Fungal pathogens of Schinus terebinthifolius from Brazil as potential classical biological control agents
A.B.V. Faria, R.W. Barreto and J.P. Cuda
Testing the efficacy of specialist herbivores to control Lepidium draba in combination with different
management practices
H.L. Hinz, A. Diaconu, M. Talmaciu, V. Nastasa and M. Grecu
263
270
278
Assessing herbivore impact on a highly plastic annual vine
J.A. Hough-Goldstein
283
The disintegration of the Scrophulariaceae and the biological control of Buddleja davidii
M.K. Kay, B. Gresham, R.L. Hill and X. Zhang
287
Quarantine evaluation of Eucryptorrhynchus brandti (Harold) (Coleoptera: Curculionidae),
a potential biological control agent of tree of heaven, Ailanthus altissima, in Virginia, USA
L.T. Kok, S.M. Salom, S. Yan, N.J. Herrick and T.J. McAvoy
292
The insect fauna of Chondrilla juncea L. (Asteraceae) in Bulgaria and preliminary studies of
Schinia cognata (L.) (Lepidoptera: Noctuidae) as a potential biological control agent
I. Lecheva, A. Karova and G. Markin
301
viii
Contents
Biological control of aquatic weeds by Plectosporium alismatis, a potential mycoherbicide in
Australian rice crops: comparison of liquid culture media for their ability to produce high yields
of desiccation-tolerant propagules
C. Moulay, S. Cliquet, K. Zeeshan, G.J. Ash and E.J. Cother
Feeding and oviposition tests refute hostherbivore relationship between Fragaria spp. and
Abia sericea, a candidate for biological control of Dipsacus spp.
B.G. Rector, V. Harizanova and A. Stoeva
The cereal rust mite, Abacarus hystrix, cannot be used for biological control of quackgrass
A. Skoracka and B.G. Rector
306
311
317
Refining methods to improve pre-release risk assessment of prospective agents: the case of Ceratapion
basicorne
L. Smith, M. Cristofaro, C. Tronci and R. Hayat
321
Host-specificity testing on Leipothrix dipsacivagus (Acari: Eriophyidae), a candidate for
biological control of Dipsacus spp.
A. Stoeva, B.G. Rector and V. Harizanova
328
Impact of larval and adult feeding of Psylliodes chalcomera (Coleoptera: Chrysomelidae)
on Centaurea solstitialis (yellow starthistle)
C. Tronci, A. Paolini, F. Lecce, F. Di Cristina, M. Cristofaro, S.Ya. Reznik and L. Smith
333
Syphraea uberabensis (Coleoptera: Chrysomelidae) potential agent for biological control of
Tibouchina herbacea (Melastomataceae) in the archipelago of Hawaii, USA
C. Wikler and P.G. Souza
340
Host-specificity testing of Prospodium transformans (Uredinales: Uropyxidaceae), a biological
control agent for use against Tecoma stans var. stans (Bignoniaceae)
A.R. Wood
345
Study on the herbicidal activity of vulculic acid from Nimbya alternantherae
M.M. Xiang, L.L. Fan, Y.S. Zeng and Y.P. Zhou
349
Abstracts
Optimization of water activity and placement of Pesta-Pseudomonas fluorescens
BRG100biocontrol of green foxtail
S.M. Boyetchko, R.K. Hynes, K. Sawchyn, D. Hupka and J. Geissler
353
Impact of natural enemies on the potential damage of Hydrellia sp. (Diptera: Ephydridae)
on Egeria densa
G. Cabrera Walsh, F. Mattioli and L.W.J. Anderson
353
Towards to study of the sunflower broomrape fungi disease in Georgia
C. Chkhubianishvili, I. Malania, E. Tabatadze and L. Tsivilashvili
354
Biological control of Imperata cylindrica in West Africa using fungal pathogens
A. Den Breeyen, R. Charudattan, F. Beed, G.E. MacDonald, J.A. Rollins and F. Altpeter
354
Impact of Ischnodemus variegatus (Hemiptera: Blissidae) on the invasive grass
Hymenachne amplexicaulis in Florida
R. Diaz, W.A. Overholt, J.P. Cuda and P.D. Pratt
Ecological basis for biological control of Arundo donax in California
T.L. Dudley, A. Lambert and A. Kirk
Biology and host specificity of Puccinia arechavaletae, a potential agent for the biocontrol
of Cardiospermum grandiflorum
A. Fourie and A.R. Wood
ix
355
355
356
XII International Symposium on Biological Control of Weeds
Potential for host-specific biological control agents at population/subspecies level?
P. Hfliger and B. Blossey
356
Combined effects of herbicides and rust fungi on Rumex obtusifolius
P.E. Hatcher and F.J. Palomares-Rius
357
Host-specificity and potential of Kokujewia ectrapela Konow for the control of Rumex spp.
Y. Karimpour
357
Growth and phenology of three Lythraceae species in relation to feeding by the leaf beetles,
Galerucella spp.
E.J.S. Katovich, R.L. Becker, L.C. Skinner and D.W. Ragsdale
358
Corynespora cassiicola f. sp. benghalensis, a new natural enemy of Commelina benghalensis:
infection parameters
D.C. Lustosa and R.W. Barreto
358
Potential use of Trichilogaster acaciaelongifoliae as a biocontrol agent of Acacia longifolia in Portugal
H. Marchante, H. Freitas and J. Hoffmann
359
Diclidophlebia smithi (Hemiptera, Psylloidea): a potential biocontrol agent for Miconia calvescens
E.G.F. Morais, M.C. Picano, R.W. Barreto, G. Silva, M.R. Campos and R.B. Queiroz
359
Supplementary host-specificity testing of Puccinia melampodii, a biocontrol agent of
Parthenium hysterophorus
K. Ntushelo and A.R. Wood
Is Prosopis meeting its match in Baringo?
W.O. Ogutu, H. Mueller-Schaerer, U. Schaffner, P.J. Edwards and R. Day
A lace bug as biological control agent of yellow starthistle, Centaurea solstitialis L. (Asteraceae):
an unusual choice
A. Paolini, C. Tronci, F. Lecce, R. Hayat, F. Di Cristina, M. Cristofaro and L. Smith
Potential biological control of Lantana camara in the Galapagos using the rust Puccinia lantanae
J.L. Rentera and C. Ellison
360
360
361
361
Biology and host specificity of Puccinia conoclinii for biocontrol of Campuloclinium macrocephalum
in South Africa
E. Retief and A.R. Wood
362
Status of tree of heaven, Ailanthus altissima, in Virginia, USA and quarantine evaluation of
Eucryptorrhynchus brandti (Harold) (Coleoptera: Curculionidae), a potential biological control agent
S.M. Salom, L.T. Kok, S. Yan, N. Herrick and T.J. McAvoy
362
Host use by the biological control agent Longitarsus jacobaeae among closely related plant species?
U. Schaffner, P. Pelser and K. Vrieling
Towards predicting establishment of Longitarsus bethae, root-feeding flea beetle introduced into
South Africa for potential release against Lantana camara
D.O. Simelane
363
363
Host-specificity testing the French broom psyllid Arytinnis hakani (Loginova)
T. Thomann and A.W. Sheppard
364
Prospects for the biocontrol of Banana Passionfruit in New Zealand with a Septoria leaf pathogen
N.W. Waipara, A.H. Gourlay, A.F. Gianotti, J. Barton, L.S. Nagasawa and E.M. Killgore
364
Novel preliminary host-specificity testing of Endophyllum osteospermi (Uredinales)
A.R. Wood
365
Potential of Ustilago sporoboli-indici for biological control of five invasive Sporobolus grasses
in Australia
K.S. Yobo, M.D. Laing, W.A. Palmer and R.G. Shivas
365
Contents
Theme 5: Regulations and Public Awareness
367
Papers
Regulation of biological weed control agents in Europe: results of the EU Policy Support Action
REBECA [Keynote paper]
R.-U. Ehlers
369
Avoiding tears before bedtime: how biological control researchers could undertake better dialogue
with their communities
L.M. Hayes, C. Horn and P.O.B. Lyver
376
Field release of the rust fungus Puccinia spegazzinii to control Mikania micrantha in India:
protocols and raising awareness
K.V. Sankaran, K.C. Puzari, C.A. Ellison, P.S. Kumar and U. Dev
384
What every biocontrol researcher should know about the public
K.D. Warner, J.N. McNeil and C. Getz
390
Abstracts
Is the Code of Best Practices helping to make biological control of weeds less risky?
J. Balciunas and E.M. Coombs
395
The new quarantine facility, St. Paul, MN, USA
R.L. Becker, D.W. Ragsdale, D. Sreenivasam, J. Heil, Z. Wu, M. Hanks, E.J.S. Katovich and L.C. Skinner
395
Biological control of weeds at the USDA-ARS-SABCL in Argentina: history and current program
J.A. Briano
396
A quarter of a century of contributions from the FDWSRU in biological control of weeds
W.L. Bruckart, D.K. Berner and D.G. Luster
396
Protocol for projects on classical biological control of weeds with insects
G. Campobasso and G. Terragitti
397
Weed biological control evaluation process in the United States - past and present
A.F. Cofrancesco, Jr
397
Biocontrol capacity of ARS research group in Central Asia and surrounding areas
R.V. Jashenko and C.J. DeLoach
398
USDA-ARS Australian Biological Control Laboratory
M.F. Purcell, A.D. Wright, J. Makinson, R. Zonneveld, B. Brown, D. Mira and G.W. Fichera
398
Status of biological control in Australia, policy and regulatory influences
J.K. Scott
399
401
Theme 6: Evolutionary Processes
Papers
The primacy of evolution in biological control [Keynote paper]
G.Roderick and M. Navajas
403
Does phylogeny explain the host-choice behaviour of potential biological control agents
for Brassicaceae weeds?
H.L. Hinz, M. Schwarzlnder and J. Gaskin
410
Population structure of an inadvertently introduced biological control agent of toadflaxes: Brachypterolus
pulicarius in North America
R.A. Hufbauer and D.K. MacKinnon
418
xi
XII International Symposium on Biological Control of Weeds
Genetic analysis of native and introduced populations of Taeniatherum caput-medusae
(Poaceae): implications for biological control
S.J. Novak and R. Sforza
422
The use of surrogate herbivores for the pre-release efficacy screening of biological control
agents of Lepidium draba
K.P. Puliafico, M. Schwarzlnder, H.L. Hinz and B.L. Harmon
429
The evolutionary history of an invasive species: alligator weed, Alternanthera philoxeroides
A.J. Sosa, E. Greizerstein, M.V. Cardo, M.C. Telesnicki and M.H. Julien
Landscape genetics and climatic associations of flea beetle lineages and implications for biological
control of tansy ragwort
M. Szcs, C.L. Anderson and M. Schwarzlnder
435
443
Abstracts
Genetic characterization of the whitetop collar gall weevil, Ceutorhynchus assimilis, enhances its
potential as biological control agent
M.C. Bon, B. Fumanal, J.F. Martin and J. Gaskin
Pinpointing the origin of North American invasive Vincetoxicum spp. using phylogeographical markers
M.C. Bon, R. Sforza, W. Jones, C. Hurard, L.R. Milbrath and S. Darbyshire
448
448
Population genetics of invasive North American diffuse and spotted knapweed (Centaurea
diffusa and C. stoebe)
R.A. Hufbauer, R.A. Marrs and R. Sforza
449
Morphological and genetic methods to differentiate and track strains of Phoma clematidina
on Clematis in New Zealand
H.M. Harman, N.W. Waipara, H. Kitchen, R.B. Beever, B. Massey, S. Parkes and P. Wilkie
449
Polyploidy, life cycle, herbivory and invasion success: work on Centaurea maculosa
H. Mller-Schrer, H. Bowman Gillianne, U. Treier, C. Bollig, U. Schaffner and T. Steinger
Use of morphometrics and multivariate analysis for classification of Diorhabda ecotypes from
the old world
J. Sanabria, J.L. Tracy, T.O. Robbins and C.J. DeLoach
450
450
Why are there no species-specific natural enemies for giant hogweed?
M.K. Seier and M.J.W. Cock
451
Specificity and plant host phenology: the case of Gephyraulus raphanistri (Diptera: Cecidomyiidae)
J. Vitou, J.K. Scott and A.W. Sheppard
451
Comparative invasion histories of Australians invading South Africa
J.R.U. Wilson, D.M. Richardson, A.J. Lowe, T.A.J. Hedderson, J.H. Hoffmann, A.W. Sheppard,
A.B.R. Witt and L.C. Foxcroft
452
Theme 7: Opportunities and Constraints for the Biological
Control of Weeds in Europe
453
Papers
Opportunities and constraints for the biological control of weeds in Europe [Keynote paper]
M. Vurro and H.C. Evans
455
Could Fallopia japonica be the first target for classical weed biocontrol in Europe?
D.H. Djeddour, R.H. Shaw, H.C. Evans, R.A. Tanner, D. Kurose, N. Takahashi and M. Seier
463
Biological control of Rumex species in Europe: opportunities and constraints
P.E. Hatcher, L.O. Brandsaeter, G. Davies, A. Lscher, H.L. Hinz, R. Eschen and U. Schaffner
470
xii
Contents
Opportunities for classical biological control of weeds in European overseas territories
T. Le Bourgeois, V. Blanfort, S. Baret, C. Lavergne, Y. Soubeyran and J.Y. Meyer
476
Weed biological control regulation in Europe: boring but important
R.H. Shaw
484
Abstracts
Field evaluation of Fusarium oxysporum as a biocontrol agent for Orobanche ramose
E. Kohlschmid, D. Mller-Stver and J. Sauerborn
489
Potential for biological control of Hydrocotyle ranunculoides in Europe
R. Shaw and J.R. Newman
489
Alien poisonous weeds: a challenge for a biological control of weeds program in Europe
R. Sforza, M. Cristofaro and W. Jones
490
Using augmentative biocontrol against Euphorbia esula: an innovative program in France
R. Sforza, J. Le Maguet, B. Gard and L. Curtet
490
The biological control of Impatiens glandulifera Royle
R.A. Tanner and H.C. Evans
491
Theme 8: Release Activities and Post-release Evaluations
493
Papers
Release strategies in weed biocontrol: how well are we doing and is there room for
improvement? [Keynote paper]
S.V. Fowler, H.M. Harman, J. Memmott, P.G. Peterson and L. Smith
495
Feeding impacts of a leafy spurge biological control agent on a native plant, Euphorbia robusta
J.L. Baker and N.A.P. Webber
503
Variation in the efficacy of a mycoherbicide and two synthetic herbicide alternatives
G.W. Bourdt, G.A. Hurrell and D.J. Saville
507
Ten years after the release of the water hyacinth mirid Eccritotarsus catarinensis in South Africa:
what have we learnt?
J.A. Coetzee, M.P. Hill and M.J. Byrne
512
Release and establishment of the Scotch broom seed beetle, Bruchidius villosus, in Oregon
and Washington, USA
E.M. Coombs, G.P. Markin and J. Andreas
516
Biological control of Mediterranean sage (Salvia aethiopis) in Oregon
E.M. Coombs, J.C. Miller, L.A. Andres and C.E. Turner
521
Preliminary results of a survey on the role of arthropod rearing in classical weed biological control
R. De Clerck-Floate, H.L. Hinz, T. Heard, M. Julien, T. Wardill and C. Cook
528
Beginning success of biological control of saltcedars (Tamarix spp.) in the southwestern USA
C.J. DeLoach, P.J. Moran, A.E. Knutson, D.C. Thompson, R.I. Carruthers, J. Michels, J.C. Herr,
M. Muegge, D. Eberts, C. Randal, J. Everitt, S. OMeara and J. Sanabria
535
Monitoring the rust fungus, Puccinia jaceae var. solstitialis, for biological control of yellow starthistle
(Centaurea solstitialis)
A.J. Fisher, D.M. Woods, L. Smith and W.L. Bruckart
540
Is ragwort flea beetle (Longitarsus jacobeae) performance reduced by high rainfall on the West Coast,
South Island, New Zealand?
A.H. Gourlay, S.V. Fowler and G. Rattray
545
xiii
XII International Symposium on Biological Control of Weeds
Host-range investigations of potential biological control agents of alien invasive
hawkweeds (Hieracium spp.) in the USA and Canada: an overview
G. Grosskopf, L.M. Wilson and J.L. Littlefield
552
Azolla filiculoides Lamarck (Pteridophyta: Azollaceae) control in South Africa: a 10-year review
M.P. Hill, A.J. McConnachie and M.J. Byrne
558
Species pairs for the biological control of weeds: advantageous or unnecessary?
C.A.R. Jackson and J.H. Myers
561
Field studies of the biology of the moth Bradyrrhoa gilveolella (Treitschke) (Lepidoptera: Pyralidae)
as a potential biocontrol agent for Chondrilla juncea
J. Kashefi, G.P. Markin and J.L. Littlefield
568
The release and establishment of the tansy ragwort flea beetle in the northern Rocky Mountains
of Montana
J.L. Littlefield, G.P. Markin, K.P. Puliafico and A.E. deMeij
573
Factors affecting mass production of Duosporium yamadanum in rice grains
D.M. Macedo, R.W. Barreto and A.W.V. Pomella
Biological control of tansy ragwort (Senecio jacobaeae, L.) by the cinnabar moth, Tyria jacobaeae (CL)
(Lepidoptera: Arctiidae), in the northern Rocky Mountains
G.P. Markin and J.L. Littlefield
Establishment, spread and initial impacts of Gratiana boliviana (Chrysomelidae) on
Solanum viarum in Florida
J. Medal, W.A. Overholt, P. Stansly, A. Roda, L. Osborne, K. Hibbard, R. Gaskalla, E. Burns,
J. Chong, B. Sellers, S.D. Hight, J.P. Cuda, M. Vitorino, E. Bredow, J.H. Pedrosa-Macedo and C. Wikler
Dissemination and impacts of the fungal pathogen, Colletotrichum gloeosporioides f. sp. miconiae,
on the invasive alien tree, Miconia calvescens, in Tahiti (South Pacific)
J.-Y. Meyer, R. Taputuarai and E. Killgore
One agent is usually sufficient for successful biological control of weeds
J.H. Myers
Evaluating implementation success for seven seed head insects on Centaurea solstitialis in
California, USA
M.J. Pitcairn, B.Villegas, D.M. Woods, R. Yacoub and D.B. Joley
The ragweed leaf beetle Zygogramma suturalis F. (Coleoptera: Chrysomelidae) in Russia:
current distribution, abundance and implication for biological control of common ragweed,
Ambrosia artemisiifolia L.
S.Ya. Reznik, I.A. Spasskaya, M.Yu. Dolgovskaya, M.G. Volkovitsh and V.F. Zaitzev
Long-term field evaluation of Mecinus janthinus releases against Dalmatian toadflax in Montana (USA)
S.E. Sing, D.K. Weaver, R.M. Nowierski and G.P. Markin
Post-release evaluation of invasive plant biological control agents in BC using IAPP, a novel database
management platform
S.C. Turner
577
583
589
594
601
607
614
620
625
Abstracts
Monitoring of ground cover post release of Aphthona nigriscutis near Lander, Wyoming
J.L. Baker and N.A.P. Webber
Benefits to New Zealands native flora from the successful biological control of mistflower
(Ageratina riparia)
J. Barton and S.V. Fowler
xiv
631
631
Contents
Tracking population outbreaks: impact and quality of Aphthona flea beetles on leafy spurge at
two spatial scales
R.S. Bourchier
632
Are nutrients limiting the successful biological control of water hyacinth, Eichhornia crassipes,
in South Africa?
R. Brudvig, M.P. Hill, M. Robertson and M.J. Byrne
632
Spatial evaluation of weed infestation and bioagent efficacy: an evolution in monitoring technique
V.A. Carney, G.J. Michels Jr and D. Jurovich
633
Influence of release size on the establishment and impact of a biocontrol root weevil
R. De Clerck-Floate
633
Development of Mycoleptodiscus terrestris as a biological control agent of Hydrilla
C.A. Dunlap and M. Jackson
634
Molecular characterization of Striga mycoherbicides Fusarium oxysporum strains:
evidence for a new forma specialis
A. Elzein, M. Thines, F. Brndle, J. Kroschel, G. Cadisch and P. Marley
634
Prioritizing candidate biocontrol agents for garlic mustard based on their potential effect
on weed demography
E. Gerber, H. Hinz, D.A. Landis, A.S. Davis, B. Blossey and V. Nuzzo
635
The accidentally introduced Canada thistle mite Aceria anthocoptes in the western USA:
utilization of native Cirsium thistles?
R.W. Hansen
635
Formulation of Colletotrichum truncatum into complex coacervate biocontrol of scentless
chamomile, Matricaria perforata
R.K. Hynes, P. Chumala, D. Hupka and G. Peng
636
Efficacy of the seed feeding bruchid beetle, Sulcobruchus subsuturalis, in the biological
control of Caesalpinia decapetala in South Africa
F.N. Kalibbala, E.T.F. Witkowski and M.J. Byrne
636
Field studies of the biology of the moth, Bradyrrhoa gilveolla, as a potential biocontrol agent
for Chondrilla juncea
J. Kashefi, G.P. Markin and J.L. Littlefield
637
Release of additional strains of the rust, Phragmidium violaceum, to enhance blackberry
biocontrol in Australia
L. Morin, R. Aveyard, K.L. Batchelor, K.J. Evans, D. Hartley and M. Jourdan
637
Impact of the bridal creeper rust fungus, Puccinia myrsiphylli
L. Morin, A. Reid and A.J. Willis
638
Overview of the biological control of the invasive plant Chromolaena odorata (Asteraceae)
in the Old World
R. Muniappan and G.V.P. Reddy
638
Trichopria columbiana a pupal parasite of the Hydrellia spp. introduced for the
management of hydrilla
J.G. Nachtrieb, M.J. Grodowitz and N. Harms
639
What is responsible for the low establishment of the bridal creeper leaf beetle in Australia?
M. Neave, L. Morin and A. Reid
Introduction, specificity and establishment of Tetranychus lintearius for biological control of
gorse in Chile
H. Norambuena
xv
639
640
XII International Symposium on Biological Control of Weeds
Were ineffective agents selected for the biological control of skeletonweed in North America?
A post-release analysis
L.K. Parsons, L.M. Collison, J.D. Milan, B.L. Harmon, G. Newcombe, J. Gaskin and M. Schwarzlnder
640
Confirming host-specificity predictions for Oxyops vitiosa, a biological control agent of
Melaleuca quinquenervia
P.D. Pratt, M.B. Rayamajhi, T.D. Center and P.W. Tipping
641
Biological control of the ivy gourd, Coccinia grandis (Cucurbitaceae), in the Mariana Islands
G.V.P. Reddy, J. Bamba, T.Z. Cruz and R. Muniappan
Quantifying the impact of biological control: what have we learned from the bridal creeper-rust
fungus system?
A. Reid and L. Morin
641
642
From invasive to fixed-in-place: the transformation of Melaleuca quinquenervia in Florida
P.W. Tipping, P.D. Pratt and T.D. Center
642
Long-term field evaluation of Mecinus janthinus releases against Dalmatian toadflax in Montana (USA)
S.E. Sing, D.K. Weaver, R.M. Nowierski and G.P. Markin
643
Population dynamics and long-term effects of Galerucella spp. on purple loosestrife,
Lythrum salicaria, and non-target native plant communities in Minnesota
L.C. Skinner and D.W. Ragsdale
Midges and wasps gain tarsus hold successful release strategies for two Hieracium biocontrol agents
L.A. Smith, P. Syrett and G. Grosskopf
643
644
Are seedfeeding insects adequately controlling yellow starthistle (Centaurea soltitialis)
in the western USA?
R.L. Winston and M. Schwarzlnder
644
Impact of the rust fungus Uromycladium tepperianum on the invasive tree, Acacia saligna,
in South Africa: 15 years of monitoring
A.R. Wood
645
Success at what price? Establishment, spread and impact of Pareuchaetes insulata on
Chromolaena odorata in South Africa
C. Zachariades, L.W. Strathie, D. Sharp and T. Rambuda
645
Theme 9: Management Specifics, Integration, Restoration
and Implementation
647
Papers
Integration of biological control into weed management strategies [Keynote paper]
J.M. DiTomaso
649
Biological control of Melaleuca quinquenervia: goal-based assessment of success
T.D. Center, P.D. Pratt, P.W. Tipping, M.B. Rayamajhi, S.A. Wineriter and M.F. Purcell
655
Hydrilla verticillata threatens South African waters
J.A. Coetzee and P.T. Madeira
665
Status of the biological control of banana poka, Passiflora mollissima (aka P. tarminiana) in Hawaii
R.D. Friesen, C.E. Causton and G.P. Markin
669
A cooperative research model biological control of Parkinsonia aculeata and Landcare
groups in northern Australia
V.J. Galea
xvi
676
Contents
A global view of the future for biological control of gorse, Ulex europaeus L.
680
R.L. Hill, J. Ireson, A.W. Sheppard, A.H. Gourlay, H. Norambuena, G.P. Markin, R. Kwong and E.M. Coombs
Assigning success in biological weed control: what do we really mean?
J.H. Hoffmann and V.C. Moran
687
Combination of a mycoherbicide with selected chemical herbicides for control of
Euphorbia heterophylla
K.L. Nechet, B.S. Vieira, R.W. Barreto, E.S.G. Mizubuti and A.A. Silva
693
Sustainable management based on biological control and ecological restoration of an alien
invasive weed, Ageratina adenophora (Asteraceae) in China
F. Zhang, W.-X. Liu, F.-H. Wan and C.A. Ellison
699
Abstracts
Trans-Atlantic opportunities for collaboration on classical biological control of weeds with
plant pathogens
D.K. Berner and W.L. Bruckart
704
Factors affecting success and failure of Diorhabda elongata releases for control of Tamarix
spp. in western North America
T.L. Dudley, P. Dalin, D.W. Bean, D.L. Thompson, D. Kazmer, D. Eberts and C.J. DeLoach
704
Advances in Striga mycoherbicide research and development: implications and future
perspective for Africa
A. Elzein, J. Kroschel, P. Marley and G. Cadisch
705
Multispectral satellite remote sensing of water hyacinth at small extents a monitoring tool?
J.T. Fisher, B.F.N. Erasmus and M.J. Byrne
705
Innovative tools for the transfer of invasive plant management technology
M.J. Grodowitz, S.G. Whitaker, J.A. Stokes and L. Jeffers
706
Physiological age-grading techniques to assess reproductive status of insect biocontrol
agents of aquatic plants
M.J. Grodowitz and L. Lenz
706
Use of multi-attribute utility analysis for the identification of aquatic plant restoration sites
M.J. Grodowitz, R.M. Smart, J. Snow, G.O. Dick and J.A. Stokes
707
Induced resistance in plants friend or foe to biological control?
P.E. Hatcher
707
Turning the tide using the sterile insect technique to mitigate an unwanted weed biocontrol agent
S.D. Hight, J.E. Carpenter, S. Bloem and K.A. Bloem
708
Integrated weed control using a retardant dose of glyphosate: a new management tool for water hyacinth
A.M. Jadhav, A. Kirton, M.P. Hill, M. Robertson and M.J. Byrne
708
Avoiding biotic interference with weed biocontrol insects in Hawaii
M.T. Johnson
709
Sustainable management, based on biological control and ecological restoration, of the alien invasive
weed, Ageratina adenophora (Asteraceae), in China
W-X. Liu, F-H. Wan, F. Zhang and C.A. Ellison
Biological control of emerging weeds in South Africa: an effective strategy to halt alien plant
invasions at an early stage
A.J. McConnachie, T. Olckers, A. Fourie, K. Ntushelo, E. Retief, D.O. Simelane, L.W. Strathie,
H. Williams and A.R. Wood
xvii
709
710
XII International Symposium on Biological Control of Weeds
Routine use of molecular tools in Australian weed biological control programmes involving pathogens
L. Morin and D. Hartley
710
An ecological approach to aquatic plant management
R.M. Smart and M.J. Grodowitz
711
A cooperative approach to biological control of Parthenium hysterophorus (Asteraceae) in Africa
L.W. Strathie, A.J. McConnachie and M. Negeri
711
Biological control of Asparagus asparagoides may favour other exotic species
P.J. Turner, H. Spafford Jacob and J.K. Scott
712
The past, present, and future of biologically based weed management on rangeland watersheds
in the western United States
L. Williams, R.I. Carruthers, K.A. Snyder and W.S. Longland
712
An adaptive management model for the biological control of water hyacinth
J.R.U. Wilson, I. Kotz, M.P. Hill, R. Brudvig, A. King and M. Byrne
713
Monitoring garlic mustard populations in anticipation of future biocontrol release
L.C. Van Riper, L.C. Skinner and B. Blossey
713
715
Workshop Reports
Feasibility of biological control of common ragweed (Ambrosia artemisiifolia)
a noxious and highly allergenic weed in Europe
D. Coutinot, U. Starfinger, R. McFadyen, M.G. Volkovitsh, L. Kiss, M. Cristofaro and P. Ehret
Rearing Insects
R. De Clerck-Floate and H.L. Hinz
717
720
Correction to Last Proceedings
721
Author index
723
Keyword Index
729
List of Delegates
733
Symposium Photograph
742
xviii
Preface
Venue and delegates
The XII International Symposium on Biological Control of Weeds was held from 22nd to 27th April 2007 in
Southern France. The venue was the Palais des Congrs at La Grande Motte, on the shores of the Mare nostrum,
the name used by the Romans for the Mediterranean Sea. Two hundred and fifty delegates from 32 countries
attended this 5-day symposium.
Opening ceremony
The symposium was opened on the morning of Monday 27 April 2008, with a welcome to La Grande Motte talk
by the Mayor of La Grande Motte, Mr. Henri Dunoyer. This was followed by an introduction to weed and other
research activities in the region by Prof. Jacques Maillet, SUPAGRO Montpellier. The opening address, on risk
assessment and biological control of weed, was presented by Dr Ernest Delfosse, USDA.
On Sunday evening, before the opening ceremony, a cocktail party was organized for participants and their partners
at the Palais des Congrs.
Sponsors
The organizing committee is very thankful to the sponsors that supported this international event. Their generousity
made the event possible and supported the publication of this Proceedings. They were: CAB International (CABI),
California Department of Food and Agriculture (CDFA), Commonwealth Scientific and Industrial Research
Organisation (CSIRO), Centre de Coopration Internationale en Recherche pour le Dveloppement (CIRAD),
United States Department of Agriculture-Agricultural Research Service (USDA/ARS), The United States Army
Corps of Engineers, and the European Weed Research Society (EWRS).
Symposium programme structure
The scientific program was divided into nine themes with a keynote speaker for all except one theme. There were
68 talks and 180 posters.
Theme chair
Talks and posters
Keynote speakers and titles
Theme: Ecology and modeling in biological control of weeds
Andy Sheppard
9 talks
17 posters
Yvonne Buckley: Is modelling population dynamics useful for anything
other than keeping a researcher busy?
Theme: Benefit-Risk Cost analyses
Ernest (Del) Delfosse
8 talks
6 posters
Rachel McFadyen: Return on investment: determining the economic
impact of biocontrol programs
9 talks
43 posters
Robert W. Barreto: Latin American weed-biocontrol science at the
crossroads
Target and agent selection
Ren Sforza
Pre-release specificity and efficacy testing
Hariet Hinz
7 talks
33 posters
none
xix
XII International Symposium on Biological Control of Weeds
Theme chair
Talks and posters
Keynote speakers and titles
Regulations & Public awareness
Dick Shaw
5 talks
8 posters
Ralf-Udo Ehlers: Regulation of biological weed control agents
results of Policy Support Action REBECA.
7 talks
8 posters
George K. Roderick: Biological control meets evolutionary biology in
the South of France.
Evolutionary processes
Ruth Hufbauer
Opportunities and constraints for biological control of weeds in Europe
Paul Hatcher
Maurizio Vurro & Harry Evans: Opportunities and constraints for
biological control of weeds in Europe.
5 talks
5 posters
Release activities and post-release evaluations
Rosemarie De ClerckFloate
10 talks
40 posters
Simon V. Fowler, et al.: Release strategies in weed biocontrol: how
well are we doing and is there room for improvement?
Management specifics, integration, restoration, implementation
John Hoffmann
9 talks
20 posters
Joe M. DiTomaso: Integration of biological control into weed management strategies.
Six workshops were also organized during the week:
1: Brassicaceae weeds by Hariet Hinz & Mark Schwartzlander.
2: Risk assessmentby Ernest (Del) Delfosse.
3: Aquatic weeds by Michael Grodowitz.
4: Feasibility of biological control of common ragweed (Ambrosia artemisiifolia) in Europe by Dominique
Coutinot , Massimo Cristofaro , Levente Kiss & Pierre Ehret.
5: Rearing insects by Rosemarrie De Clerck-Floate & Hariet Hinz.
6: Swallow worts by Lindsey Milbrath.
Reports on two of these workshops (Biological control of ragweed, and Rearing insects) can be found at the end
of this proceedings.
Mid-symposium tours
Two options were given to delegates: A visit to the Cvennes (foothills of the Massif Central) or to the Camargues
(delta wetlands of the Rhne River). Both tours were held on the sunny day of Wednesday 24 April. The Camargues
tour was organized by Marie-Claude Bon and Brian Rector,and 200 delegates visited this natural reserve and
enjoyed seeing local fauna, such as black bulls, white horses, Grey Heron, greater flamingos, under the guidance
of Nicolas Beck from the Tour du Valat Research Center. Special attention was given to invasive Baccharis sp.,
Pampas grass, Ludwigia spp. The Cvennes tour was organized by Janine Vitou, Mic Julien and Ren Sforza.
One hundred delegates visited a part of the only French national park in the low mountains. This included a short
walk along an ancient Roman road and a scenic picnic. The park guide, Emeric Sulmont, discussed the negative
impacts of the invasives Fallopia japonica and Robinia pseudoacacia and the control methods conducted by local
authourities.
Wine and cheese evening and gala dinner
On the evening of Tuesday 23 April, a wine and cheese party was held. The choice was a selection of succulent
and delightful cheeses of France picked by our specialist, Thierry Thomann. The cheese was accompanied by
red and white wines, and other interesting beverages, from all over our planet, brought by the delegates. It was a
memorable evening with almost no cheese and wine remaining afterwards.
The conference dinner was held on the evening of Thursday 26 April at Le Chteau du Pouget, located at
Vrargues, with historical significance and romantic ambience. After welcome drinks and buffet in the park of the
11th century Chteau a dinner was accompanied by musical entertainment from the band Agate ze bouze. Poster
and oral presentation prizes were awarded during the evening.
xx
Preface
Committees and support
The local organizing committee comprised Janine Vitou, Marie-Claude Bon, Brian Rector, Mic Julien(co-chair),
Ren Sforza (co-chair) and Andy Sheppard.
The scientific committee comprised Mic Julien (convenor), Ren Sforza, Marie-Claude Bon, Brian Rector,
Matthew Cock, Massimo Cristofaro, Paul Hatcher, Hariet Hinz, Walker Jones, Thomas Le Bourgeois, Hlia
Marchante, Heinz Mller-Schrer, Marion Seier, Richard Shaw, Andy Sheppard, and Janine Vitou. Conference
administration was provided by AlphaVisa Congrs. Additional secretariat services was given by Sarah Hague,
and computer logistics was supported by Xavier Chataigner. Lo Ruamps, Benjamin Gard, Christophe Girod and
Steeve Schawann helped with logistics.
The editorial panel for this proceedings comprised Mic Julien, Ren Sforza, Marie-Claude Bon, Harry Evans, Paul
Hatcher, Hariet Hinz and Brian Rector.
Next symposium
The attendees agreed that the next meeting should be held in Hawaii, USA. It will be convened by Dr Tracy
Johnson, USDA Forest Service.
Ren Sforza
USDA-ARS-EBCL
Mic Julien
CSIRO European Laboratory
Local committee (left to right): Brian Rector, Ren Sforza, Mic Julien, Janine Vitou, Marie-Claude Bon.
xxi
This page intentionally left blank
Theme 1:
Ecology and Modelling in Biological
Control of Weeds
Session Chair: Andy Sheppard
This page intentionally left blank
Keynote Presenter
Is modelling population dynamics useful
for anything other than keeping a
researcher busy?
Y.M. Buckley1,2
Summary
Understanding and modelling the population dynamics of weeds and/or biological control agents can
require large investments of time and money; just what are we getting for our modelling efforts? Here
I respond to three persistent critiques of modelling as used in biological control programmes and
present new directions for extending and improving our use of models. Complex models have been
critiqued as resource-intensive, too narrow in scope and difficult to analyse, whereas simple, strategic
models are critiqued as oversimplified and inaccurate in predicting postinvasion population dynamics. I argue that models across this spectrum can be useful and that the dichotomy between simple
and complex models can be broken down. Biological control practitioners often operate in systems
with a high degree of stochasticity and uncertainty; therefore, the incorporation of stochasticity and
uncertainty into population models is essential for the development of robust management strategies.
Close dialogue between managers and modellers is essential for the application of modelling studies
to management. New directions for modelling in biological control include the incorporation of invader impact and complex ecosystem effects such as habitat heterogeneity and disturbance. The right
model used for the right question can bring us insights into the biological control process that would
be difficult or impossible to achieve otherwise.
Keywords: population dynamics, modelling, biological control.
Introduction
biological control programmes has become well-established in the past decade (e.g. Rees and Paynter, 1997;
Shea and Kelly, 1998; Buckley et al., 2005b), but critiques remain on the general use of models, the questions they are brought to answer and the applicability
of their results for management. Here I address three
critiques of population modelling and identify directions where modelling tools are likely to generate useful new insights into the role of biological control in
weed management.
Working out the population dynamics of a species can
keep a large research group going for a long time. This is
generally not possible in a biological control program
(Zalucki and van Klinken, 2006).
Although Zalucki and van Klinken (2006) refer specifically to the use of population modelling for predicting
biological control agent abundance across their potential exotic ranges, I have used this quote to represent a
common critique of modelling projects, which is that
they are time- and data-hungry, too simplistic and contribute little of use to on-ground managers. The use of
different kinds of models to inform and evaluate weed
Critiques of population modelling
Three common critiques of population modelling as a
component of biological control programmes are discussed here:
1. model complexity and simplicity (covering both detail and stochasticity);
2. uncertainty in model structure and parameters; and
. applicability of modelling studies to on-ground
management.
University of Queensland, School of Integrative Biology, St. Lucia,
Brisbane, QLD 4072, Australia <[email protected]>.
2
CSIRO Sustainable Ecosystems, 306 Carmody Road, St. Lucia, QLD
4067, Australia <[email protected]>.
CAB International 2008
XII International Symposium on Biological Control of Weeds
Model complexity and simplicity
cisely, but no amount of measurement will reduce the
yearly fluctuations in seed production. We know that
population dynamics vary in space and time and that
the effect of biological control agents is also likely to
vary; purely deterministic models will therefore fail to
predict the results of the interaction over the range of
conditions likely to be encountered in the field. Does
this mean that deterministic models should be abandoned? I would argue the contrary, as traditional analysis of deterministic models gives an indication of the
likely dynamics under a range or all possible parameter
values.
Buckley et al. (2005b) used a deterministic, coupled, plantherbivore model to explore the qualitative
population dynamics likely to result from the inter
action of the weed Echium plantagineum L. (Boraginaceae) and the weevil Mogulones larvatus Schultze
(Coleoptera: Curculionidae). Ideally, classical biological control would result in a reduced but stable population of the weed supporting a stable population of
herbivores; large population fluctuations of either the
weed or herbivore could lead to extinction of the weevil and subsequent loss of control. Stability boundary
analysis of deterministic models enables identification
of the parameter values that give rise to stable, as opposed to oscillatory, dynamics. These ideal parameter
values can then be compared with estimates from the
field or laboratory. The central critique of studies such
as this one is that factors other than intrinsic population dynamics regulate populations and that stochastic
effects of spatial or temporal variability could dampen
or enhance oscillations resulting from the intrinsic deterministic dynamics alone. This criticism is entirely
valid, but in the EchiumMogulones case, despite the
deterministic origins of the model, it proved possible to
reproduce reasonably well the qualitative and quantitative dynamics in the field observed over seven years
(data not shown), and field densities of plants predicted
by the model before and after introduction of the biological control agent corresponded well with observed
data (Buckley et al., 2005b).
We should expect reasonably tight linkage between
agent and weed dynamics where the biological control
agent has a strong effect on the plant. As the agents
are host-specific, their resource base is greatly simplified, and in the case of M. larvatus, it lives within
stems, with larvae competing strongly with each other
for the plant resource, leading to strong density dependence driving the dynamics. Coupled plantherbivore
models are very rarely explored in a biological control context (Barlow, 1999), so it is currently difficult
to predict what kinds of dynamics are likely to result
from different biological control agent species (e.g.
from various taxonomic groups, feeding guilds). We
do not know in which cases strong intrinsic dynamics are likely to drive the interaction or in which cases
stochastic factors will overwhelm any deterministic
pattern.
One of the primary axes along which different
types of model can be ranged is that which at one of
its extremes has tactical, complex, predictive models
and at the other has strategic, simple, general models
of little predictive power in specific cases. Both extremes have been criticized in relation to their value
in biological control programmes, with tactical models
critiqued as being resource-intensive (Nehrbass and
Winkler, 2007), too narrow in scope and difficult to
analyse (Schreiber and Gutierrez, 1998), whereas strategic models are critiqued as oversimplified and inaccurate in predicting postinvasion population dynamics
(Zalucki and van Klinken, 2006). Models right across
the spectrum have been critiqued as inadequate in contributing to management solutions. It should be noted
that this is a long-standing general debate in applied
ecology and is not confined to the field of biological
control. Models right across this axis of complexity can
be badly and well-applied and the ability of the model
to contribute to understanding and solving the driving
problem should be the criterion used for judging the
success of the modelling approach. In other words, the
type of model to be used depends on the question being asked. The availability of data to validate and test
models is also important, and closer dialogue amongst
modellers, biological control practitioners and empirical biological control researchers will lead to more appropriate modelling approaches and collection of data
necessary for such models. The aim of most modelling
studies is not to reproduce exactly the dynamics seen
in the field but to test hypotheses about how we believe the system to be working. Ability to exactly reproduce field dynamics should not necessarily be the
acid test of the success of a modelling approach. For
example Buckley et al. (2003) constructed a complex
individual-based model of Hypericum perforatum L.
(Clusiaceae) dynamics that incorporated biotic and abiotic drivers of dynamics as well as habitat differences
and characterized the stochasticity in the system at several spatial and temporal scales. However, despite its
realism and ability to accurately represent the structure of field populations, it was not possible to predict
dynamics in the field. The aim of this model was to
produce virtual populations of plants that behaved like
H. perforatum plants on which management strategies
could be tested. The incorporation of stochasticity was
important to determine how robust the management
strategies would be to the variability observed in the
field.
Stochasticity is variability in population model parameters or structure due to underlying processes such
as spatial or temporal variability, e.g. effects of weather
on seed production may give rise to a distribution of
fecundity values through time. Stochasticity cannot be
reduced by applying greater empirical effort, and we
may come to know the distribution of values more pre
Is modelling population dynamics useful for anything other than keeping a researcher busy?
Uncertainty in model structure
and parameters
tion of M. pigra population size over 3 years. The role
of biological control in this IWM strategy was found
to contribute substantially to its success. IWM strategies are relatively complex, and their results may be
unpredictable because of population processes and
interactions between individual control techniques. In
such cases, the use of models is quite germane but still
surprisingly rare.
Buckley et al.s (2005a) study of the population
dynamics of P. nigra was initiated by a managementdriven question about whether the introduction of a
seed-feeding biological control agent would have the
potential to reduce the rate of spread of the invasive
pine. As spread speed was found to be relatively insensitive and inelastic to the fecundity parameters, initial
recommendations were that a seed feeder would not be
highly appropriate. Modelling studies are increasingly
important in the prerelease phase of biological control
programmes where the weed dynamics and vital rates
are examined for potential management targets (Davis
et al., 2006).
Uncertainty differs from stochasticity in that it represents unknown parameter values, distributions or
model structure; it represents the extent of our ignorance of a system. Uncertainty may be reduced through
the collection of more data, but commonly in invasive
plant studies, we cannot afford to invest the time or resources necessary for intensive data collection before
management decisions are made. Even when detailed
data are available, it may still be impossible to determine the correct model to use (e.g. for E. plantagineum,
both scramble and contest competition models fit the
data equally well for M. larvatus density dependence;
Buckley et al., 2005b). Methods for including both parameter and model uncertainty into population models
are therefore highly relevant but relatively underused
in invasive plant management models.
Parameter uncertainty is pervasive and often unacknowledged; only rarely can we determine parameter
estimates with sufficient confidence whilst representing all sources of stochasticity accurately. Buckley et
al. (2005a) provided an example of a population and
spread model of an invasive pine species, Pinus nigra
Arnold, with a high degree of uncertainty in the demographic and dispersal parameters. Traditional matrix
(for population growth) or integro-difference equation (for spread) models are run under one or a few
parameter scenarios. Subsequently calculated sensitivities and elasticities then inform management by highlighting parameters and life history stages to target for
control. However, the particular parameter values used
will change the pattern of sensitivities and elasticities
for population growth rate or spread (Caswell, 2001).
Buckley et al. (2005a) investigated whether, given a
range of possible values, there are consistent patterns
that can be exploited for robust management. Despite
the large range of uncertainty identified in this case,
consistent patterns of sensitivities and elasticities with
non-overlapping confidence intervals did emerge. This
enabled the identification of suitable robust management targets in a number of different habitats. Buckley
et al. (2005a) used a Monte-Carlo sampling approach
to incorporate uncertainty; other suitable methods that
should be explored are information gap theory (BenHaim, 2001) and uncertain number theory (Regan et
al., 2004).
New directions
We can do more to increase the applicability of our models to management. Incorporation of impact and ecosystem effects into population models may have important
implications for biological control programmes.
Impact
Impact is what separates troublesome invaders from
the merely naturalized, and the importance of including
nonlinear, densityimpact relationships in biological
control studies has recently been recognized (Thomas and Reid, 2007). To date, impact has rarely been
broached in management models of invasive plants. It
has implicitly been assumed that a reduction in density will lead to a corresponding reduction in impact.
If however, impact is nonlinearly related to population
density (see Fig. 1 in Thomas and Reid, 2007) and varies amongst weed species, a biological control agent
causing only a small reduction in one weed species
density may be more effective at reducing impact than
another biological control agent having a large effect
on a second weed species density. If we assume a linear weed densityimpact curve that it is in fact nonlinear, we may be incurring large costs, in both lack
of impact and overinvestment in ineffective or wasted
control efforts.
Applicability of modelling studies to
on-ground management
Ecosystem effects
To date, we have had some successes in the use of
models to inform management strategies in the field.
Buckley et al. (2004) used a model of Mimosa pigra L.
population dynamics to make recommendations about
the type of integrated weed management (IWM) strategy that would have the greatest effect on the reduc-
Nonparametric time-series analysis of the dynamics
of the interaction between cinnabar moth, Tyria jacobaeae L. (Lepidoptera: Arctiidae), and its host plant,
ragwort, Senecio jacobaea L. (Asteraceae), revealed
strikingly different dynamics in two different locations
XII International Symposium on Biological Control of Weeds
(Bonsall et al., 2003), demonstrating that environmental context can determine the strength of intrinsic dynamics. Several studies show the habitat specificity
of population dynamics, management actions and/or
biological control agents (Buckley et al., 2003, 2005a;
Shea et al., 2005; Davis et al., 2006), as plant population dynamics differ between locations even within an
invaded range. It is also apparent that plant population
dynamics and hence management will be affected by
disturbance regimes, whether natural or anthropogenic,
including those caused by weed management itself
(Buckley et al., 2004, 2007). The inclusion of broader
ecosystem effects in population models is therefore
highly relevant for management.
Ben-Haim, Y. (2001) Information Gap Decision Theory: Decisions Under Severe Uncertainty. Academic Press, London, UK.
Bonsall, M.B., van der Meijden, E. and Crawley, M.J. (2003)
Contrasting dynamics in the same plantherbivore interaction. Proceedings of the National Academy of Sciences
of the USA 100, 1493214936.
Buckley, Y.M., Briese, D.T. and Rees, M. (2003) Demog
raphy and management of the invasive plant species
Hypericum perforatum. II. Construction and use of an
individual-based model to predict population dynamics
and the effects of management strategies. Journal of Applied Ecology 40, 494507.
Buckley, Y.M., Rees, M., Paynter, Q. and Lonsdale, W.M.
(2004) Modelling integrated weed management of an inva
sive shrub in tropical Australia. Journal of Applied Ecology 41, 547560.
Buckley, Y.M., Brockerhoff, E.G., Langer, E.R., Ledgard, N.,
North, H. and Rees, M. (2005a) Slowing down a pine invasion despite uncertainty in demography and dispersal.
Journal of Applied Ecology 42, 10201030.
Buckley, Y.M., Rees, M., Sheppard, A.W. and Smyth, M.J.
(2005b) Stable coexistence of an invasive plant and biological control agent: a parameterised coupled plant
herbivore model. Journal of Applied Ecology 42, 7079.
Buckley, Y.M., Rees, M. and Bollker, B. (2007) Disturbance,
invasion and reinvasion: managing the weed-shaped hole
in disturbed ecosystems. Ecology Letters 10, 809817.
Caswell, H. (2001) Matrix Population Models: Construction,
Analysis and Interpretation, 2nd edn. Sinauer Associates,
Inc, Sunderland, MA.
Davis, A.S., Landis, D.A., Nuzzo, V., Blossey, B., Gerber, E.
and Hinz, H.L. (2006) Demographic models inform selection of biological control agents for garlic mustard (Alliaria petiolata). Ecological Applications 16, 23992410.
Nehrbass, N. and Winkler, E. (2007) Is the giant hogweed
still a threat? An individual-based modelling approach for
local invasion dynamics of Heracleum mantegazzianum.
Ecological Modelling 201, 377384.
Rees, M. and Paynter, Q. (1997) Biological control of Scotch
broom: modelling the determinants of abundance and the
potential impact of introduced insect herbivores. Journal
of Applied Ecology 34, 12031221.
Regan, H.M., Ferson, S. and Berleant, D. (2004) Equivalence
of methods for uncertainty propagation of real-valued
random variables. International Journal of Approximate
Reasoning 36, 130.
Schreiber, S.J. and Gutierrez, A.P. (1998) A supply/demand per
spective of species invasions and coexistence: applications
to biological control. Ecological modelling 106, 2745.
Shea, K. and Kelly, D. (1998) Estimating biological control
agent impact with matrix models: Carduus nutans in New
Zealand. Ecological Applications 8, 824832.
Shea, K., Kelly, D., Sheppard, A.W. and Woodburn, T.L.
(2005) Context-dependent biological control of an invasive thistle. Ecology 86, 31743181.
Thomas, M.B. and Reid, A.M. (2007) Are exotic natural
enemies an effective way of controlling invasive plants?
Trends in Ecology and Evolution 22, 447453.
Zalucki, M.P. and van Klinken, R.D. (2006) Predicting population dynamics of weed biological control agents: science or gazing into crystal balls? Australian Journal of
Entomology 45, 331344.
Conclusions
Although critiques of the use of population modelling
in biological control programmes remain, I believe that
we have had some success in improving management
strategies before release of agents and in determining
the potential for success in ongoing biological control
programmes. My research group also plans to use models to retrospectively evaluate the effect of biological
control in a historical biological control programme.
Future progress in the use of modelling in biological
control programmes will be in the use of established
techniques earlier in the programme (e.g. prerelease),
the incorporation into population models of measures of
impact of the agents on the weed and of the weed on the
affected ecosystem or industry and the incorporation of
broader ecosystem effects on the population dynamics
of the weed and the biological control agent. Models
from across the spectrum of complexity to simplicity
can be useful at different stages in a biological control
programme. The incorporation of uncertainty directly
into the models will enable us to focus on robust management strategies that are not contingent on a narrow
set of parameters or model structure assumptions.
Acknowledgements
This research is funded by an Australian Research
Council Linkage grant (LP0667489), an Australian
Research Council Discovery grant and Australian Research Fellowship (DP0771387) and the CRC for Australian Weed Management. I thank my research group
for their contributions: Nikki Sims (evaluation of biological control), Hiroyuki Yokomizo (impact), Jennifer
Firn and Alice Yeates (disturbance, community and
ecosystem effects of management).
References
Barlow, N.D. (1999) Models in biological control: a field
guide. In: Hawkins, B.A. and Cornell, H.V. (eds) Theoretical Approaches to Biological Control. Cambridge Uni
versity Press, Cambridge, UK, pp. 4368.
Biomass reduction of
Euphorbia esula/virgata by
insect/bacterial combinations
A.J. Caesar1 and R.J. Kremer2
Summary
Biological control efforts against the perennial invasive Euphorbia esula/virgata in North America
have left 3050% of all treated sites without impact after 1015 years. Those efforts focused almost
exclusively on insect releases. Much evidence is available indicating that soil biotic factors affect
both invasiveness and biocontrol effectiveness. The authors have shown that soilborne bacteria and
fungi are linked to biomass reductions or mortality in conjunction with insect damage. To understand
factors possibly affecting synergistic interaction of the insects with plant pathogens shown to cause
rapid weed mortality, predominant bacteria associated with the flea beetle Aphthona flava Guill. (Coleoptera: Chrysomelidae) released to control E. esula/virgata L. in western North America, were isolated and identified. Two Euphorbia-infested sites with widely differing levels of impact 810 years
after insect release were sampled. From the site that exhibited rapid, sweeping declines in Euphorbia
density, 6 of 12 isolates were Bacillus spp., 4 were coryneform species and 2 were Pseudomonad
aceae. Bacteria isolated from the Cottonwood site included some species often associated with the
biocontrol of soilborne plant pathogens. The results of tests for a range of hydrolytic enzymes showed
that the two groups differed in the frequency of isolates positive for such enzymes as cellulase and
xylanase. Two isolates from each location representative of predominant bacterial species and their
range of traits were selected for testing on E. esula/virgata in combination with Aphthona spp. After
3537 weeks, two isolates positive for cellulase from the Knutson Creek site caused significant (P =
0.05) dry weight reductions of E. esula/virgata plants of 64% and 67%, respectively, in combination
with Aphthona spp. One of the two isolates from the Cottonwood site, also positive for cellulase production, caused a 60% reduction in dry weight compared with the control.
Keywords: trophic interactions, synergism, biological control, bioherbicides, bacteria.
Introduction
ity through accentuating tissue degradation. Previous
studies by the senior author have shown that the effective biological control at the Knutson site was because
of the presence and action of Rhizoctonia solani Kuhn
and Fusarium oxysporum Schlecht. emend. Snyder and
Hansen that were isolated from plants at that site. These
fungal species, obtained from insect-damaged tissue of
E. esula/virgata, were shown to be highly virulent either independently (Caesar, 1994, 1996) or in combination with Aphthona spp. (Caesar, 2003). Hydrolytic
enzymes were chosen as the traits of interest because
of their potential for increasing plant tissue damage as
well as conversely acting against soilborne pathogens
through lysis of fungal hyphae. Bacterial isolates were
tested for hydrolytic enzyme production to determine
whether there were trends in enzyme spectra amongst
isolates from beetles recovered at a successful biocontrol site and isolates from a less successful release site.
The hypothesis addressed in this work is whether the
degree of biological control activity of the flea beetle
Aphthona flava Guill. (Coleoptera: Chrysomelidae) on
the perennial invasive prairie plant, Euphorbia esula/
virgata L. (leafy spurge) is associated with traits within
members of the bacterial community vectored by the
beetle. It is not known whether the microflora associated with the flea beetles contains species that could affect E. esula by either acting as antagonists against the
documented plant pathogens or enhancing pathogenic1
USDAARS, 1500 North Central Avenue, Sidney, MT 59270, USA.
USDAARS, University of Missouri, 269 Engineering Building, Columbia, MO 65211, USA.
Corresponding author: A.J. Caesar <[email protected]>.
CAB International 2008
2
XII International Symposium on Biological Control of Weeds
In vitro tests of bacterial traits
Previous studies by Kremer have documented deleterious rhizobacteria that can damage E. esula (Kremer
and Kennedy, 1996; Kremer et al., 2006).
To investigate the effect of phenotypes that included
a range or varying intensities of hydrolytic enzyme
production might have on the capacity to interact with
insect herbivory, hydrolytic enzyme activities of the
bacterial isolates were tested using published methods.
Filter-sterilized solutions of 0.1 % 4-methylumbelliferyl
N-acetyl -d-glucosamine, 0.1% 4-methylumbelliferyl
N-acetyl -d-glucosaminide (chitin is a homopolymer
of N-acetyl-glucosamine; the latter substrate assays
for -N-acetylhexosaminidase, a chitin oligosaccharidase), 0.25% p-nitrophenyl -d-mannopyranoside
and 0.25% p-nitrophenyl -d-glucopyranoside (Sigma
Chemicals, St Louis, MO) (Fahey and Hayward, 1983)
in pH 7 phosphate buffer in sterile 96-well microtitre
dishes were used to give 150200 l per well. Plates
were inoculated with isolates and incubated at 20C for
1014 days (Santos et al., 1979). Clearing of coloured
substrates on agar media during incubation at 20C for
1014 days was used in tests to indicate xylanase (Biely
et al., 1985) or -1,4-glucanase (Scott and Schekman, 1980) using 0.2% Remazol Brilliant Blue xylan
(4-O-methyl-d-glucurono-d-xylan dyed with Remazol
Brilliant Blue R) (Biely et al., 1985) and 0.2% Ostazin
Brilliant Red hydroxyethylcellulose (hydroxyethylcellulose dyed with Ostazin Brilliant Red H-3B) (both
from Sigma Chemicals), respectively, in 2YT medium
(Sipat et al., 1987) with 1.5% agar. Tests for polygalacturonase (Hankin and Lacy, 1984) and cellulase (Barros and Thomson, 1987) were also performed. Isolates
were also assessed for in vitro antibiosis against two
soilborne fungal pathogens of E. esula: a Pythium spp.
isolate and a R. solani isolate. Bacteria were streaked
near the edge of Petri dishes containing 0.3% TSBA,
and immediately thereafter, agar plugs taken from
colony margins of one of the fungi were placed at the
opposite side of plates. Plates with these bacterial/fungal pairings were incubated at 20C and examined for
zones of inhibition after 36 h. Degree of inhibition was
scored as , +, ++ or +++ based on 0, <1-cm, >1- to
2-cm and >3-cm-wide zones of inhibition, respectively.
Materials and methods
Plant propagation
Plants used in this study were propagated from cuttings of plants obtained from a single E. esula/virgata
infestation in northeast Montana. Plants weighing ca 30 g
or more were selected for the experiment, after being
produced through continuous culture over more than
1 year and were of an overall size nearest to typical
field-size plants as was achievable in the greenhouse
whilst retaining a degree of apparent vigour similar
to that observed in the field. Plants were grown in the
greenhouse at 2028C in a potting medium containing
equal volumes of peat and vermiculite in 15 15 cm
(diameter height) plastic pots.
Source and collection of Aphthona spp.
and associated bacteria
To ascertain whether adults of Aphthona spp. might
vector plant pathogenic bacteria, active adults of the flea
beetles Aphthona nigriscutis Foudras and Aphthona la
certosa (Rosenhauer), were collected using sweep nets
from two sites within the Theodore Roosevelt National
Park, located in western North Dakota. One site, a portion of the flood plain of Knutson Creek, experienced
dramatic reductions in stand density of E. esula/virgata
following establishment of the flea beetle A. lacertosa
and attainment of high populations of the insect. Another site, Cottonwood, contained stands of E. esula/
virgata that had remained apparently unimpacted over
several years following releases of Aphthona spp. despite establishment of the flea beetle. Half of the Aph
thona adults collected from each site were washed by
placing five adult flea beetles per tube in test tubes (five
tubes per lot) containing 9 ml of pH 7 potassium phosphate buffer and vortexing for three 1-minute periods
interspersed with pauses of 30 s. Tenfold serial dilutions were prepared from the insect washes and plated
on triplicate plates of 0.3% tryptic soy agar (TSBA)
and Kings medium B and incubated at 2528C. Five
apparently distinct colonies were selected from dilution plates on which 20200 colonies occurred. To include bacteria that might be internal, the beetles of the
respective companion lots were washed by vortexing in
three changes of a pH 7 phosphate buffer/20% ethanol
solution. After the final wash, beetles in groups of five
were re-suspended in 9 ml of sterile pH 7 phosphate
buffer and ground with a mortar and pestle. Tenfold serial dilutions were plated on media. All cultures were
stored over the short term in pH 7 potassium phosphate
buffer at 4C and in LuriaBertani medium with 15%
w/v glycerol at 80C for long-term storage.
Identification by fatty acid
methyl ester profiles
Bacterial isolates were identified based on wholecell cellular fatty acids, derivatized to methyl esters,
i.e. fatty acid methyl esters. Isolates from frozen cultures were streaked twice successively on 3% TSBA.
After 24 h, cells were harvested and immediately frozen at 20C. Fatty acid methyl esters were obtained by
saponification, methylation and extraction following
the manufacturers procedure. Bacterial isolates were
analysed using the MIDI Microbial Identification Software (Sherlock TSBA40 Library version 4.5; Microbial
ID, Newark, DE). The fatty acid methyl ester profile
of Stenotrophomonas maltophilia (Hugh) Palleroni
Biomass reduction of Euphorbia esula/virgata by insect/bacterial combinations
and Bradbury (ATCC 13637) was used as a reference
for the MIDI determinations. Strains with a similarity
index (SIM) 0.300 are considered a good match and
conclusively identified (Siciliano and Germida, 1999;
Oka et al., 2000).
duce only polygalacturonase amongst nine hydrolytic
enzymes assayed, identified as S. maltophilia, caused a
24% reduction in biomass of E. esula/virgata, although
this was not significant. In vitro antibiosis against R.
solani and Pythium spp. was not a helpful trait in distinguishing the two sets of isolates. The relevance of investigating bacteria associated with adult flea beetles is
based on two premises: (1) that the bacteria carried by
the flea beetles may be active participants in the phyllosphere and/or rhizosphere once they are carried passively to the plant and (2) that the bacteria found on or
in the insects may represent species that predominate in
the host plant/insect system. A further possibility is that
these bacteria are endemic to the insect or to the plant
leaf surface, root zone or perhaps vascular system. Bacteria that have been identified in the few studies done
in these realms include species that were identified in
the present study: Ochrobacterum spp. (Spiteller et al.,
2000), Cellulomonas Bergey et al. 1923, Microbacte
rium Orla-Jensen 1919 (Zinniel et al., 2002), Bacillus
spp. (Cho et al., 2003), P. chlororaphis, S. maltophila,
B. cepacia and Bacillus thuringiensis Berliner (Canganella et al., 1994). The possibility that bacteria affect
herbivory positively or negatively is in need of further
exploration and could lead to some important contributions to a better elucidated understanding of biocontrol
ecology. That the ecology of classical weed biocontrol
is justifiably receiving greater attention seems evident
by many contributions to the proceedings of recent International Weed Biocontrol Symposia (Spencer, 2001;
Cullen et al., 2004).
Although our results show the effects of the bacteria
in reducing biomass of leafy spurge in conjunction with
insect damage, a fuller understanding of the potential
of such bacteria to cause stand reductions in combination with insects would require application of bacteria
in the field following establishment of the flea beetles.
Bacteria with the traits we have described are likely
accessory to the larger, more pronounced effects of
aggressive fungal root and crown pathogens, and they
may provide additive effects. We propose to confirm
this with further studies by distinguishing the comparative effects of fungi and bacteria. Fungi are two and
a half times more likely than insects to be the cause
of mortality when assessed using comparative risk survival analysis (Caesar, 2003). It was beyond the scope
of this study to show a definitive link of hydrolytic enzyme production and growth reduction. This study did
provide indication for simultaneous further screening
of additional candidate isolates, using criteria identified here and the immediate testing in the field of selected bacteria, such as isolates producing cellulase or a
broad spectrum of hydrolytic enzymes in combination
with Aphthona spp., for biological control of E. esula/
virgata. There remain many sites in the field where insects are established without apparent stand reductions
where bacteria can be tested. Further, our work has
shown that bacterial species not previously considered
Tests of insect/microbial interactions on
E. esula/virgata in the greenhouse
Three isolates from each of the two sites were selected based on traits that broadly typified the respective groups in terms of their taxonomic classification
and hydrolytic enzyme spectra. Isolates were grown in
TSBA at 20 to 25C. Plants of appropriate size and mass
were grown as described above. Cages consisting of
nylon netting material (32 mesh or 530 lm mesh openings) supported by an aluminum frame were placed
over all pots and secured with a clamp to prevent escape of flea beetle adults. Suspensions of isolates selected as described above were adjusted to ca 106 cells
per ml and were poured into the potting medium, 200
ml per pot, in which E. esula/virgata was growing.
Within 24 h of addition of bacteria to the pots, adults of
A. flava were released, 15 per cage, into the cages. Ten
caged plants of E. esula/virgata were treated with each
bacterial isolate used, and the experiment was repeated
once. Treated plants were grown in the greenhouse at
2530C for 3537 weeks, dried at 47C for 10 days
upon harvest and weighed. Data were tested to confirm
homogeneity of variances (Bartlett and Kendall, 1946)
before pooling data from both trials for analysis using
Waller and Duncans exact Bayesian k-ratio least significant difference rule (P = 0.05) (Waller and Duncan,
1969).
Results and discussion
Two of the nine assayed of isolates, whether originating from the highly impacted Knutson Creek site or the
static Cottonwood site, had a similar average number
of positive tests of hydrolytic enzymes (Table 1). However, 6 of 12 Knutson Creek isolates were positive for
a suite of three hydrolytic enzymes, -N-acetylhexosaminidase, a chitin oligosaccharidase and two apparently distinct or dissimilar cellulases (all three degrade
-1, 4 sugar residues), whereas only a single isolate
amongst the 12 from adults collected from Cottonwood were positive for these three enzymes. Only the
three isolates with this suite of three enzymes, including two from Knutson Creek amongst the six isolates
tested from the two sites caused significant reductions,
ranging from 61% to 67% (Table 2) in dry weight of
E. esula/virgata in greenhouse tests. The two isolates
tested that had little or no hydrolytic enzyme production (identified as Ochrobacterium anthropii Holmes
et al. and Corynebacterium acquaticum Lehmann and
Neumann) correspondingly failed to reduce biomass of
E. esula/virgata. Interestingly, an isolate shown to pro
XII International Symposium on Biological Control of Weeds
Table 1.
In vitro antibiosis and hydrolytic enzyme production by bacteria associated with the flea beetle Aphthona flava released at two sites, Knudson Creek site and Cottonwood. Tests for enzymes were with chromogenic substrates.
Isolate
Phenotypic traits of isolated bacteriaa
In vitro
antibiosis vs
Pythium spp.
Knudson Creek site
++
Pseudomonas
putida 102
++
Bacillus
cereus 103
B. cereus 104
Arthrobacter
oxydans 113
++
Bacillus
thuringiensis 124
++
B. cereus 129
++
B. cereus 154
++
Burkholderia
cepacia 207
Corynebacterium
acquaticum 207b
Cellumonas tur
bata 213a
+
B. cereus 216
+++
Microbacterium
liquefaciens 223
Cottonwood Creek site
Brevibacterium
iodinium 116
Paenibacillus
glucoanalyticus
117
++
Pseudomonas
chlororaphis 217
Ochrobactrum
anthropi 145
+
Bacillus
thuringiensis
kurstakii 146
++
Bacillus cereus
Pseudomonas
putida 226
++
Pseudomonas
chlororaphis 145
Stenotrophomonas
maltophilia 144
No match
++
No match
No match
In vitro
antibiosis vs
Rhizoctonia
solani
0.25%
p-Nitrophenyl
-d-glucopyranoside
test
0.25%
p-Nitrophenyl -dmannopyranoside
0.1 %
4-Methylumbelliferyl
N-acetyl
-d-glucosamine
+
+
++
+
+
+
++
+
+
+
+
For in vitro antibiosis tests, degree of inhibition was scored as: = no inhibition; + = 1-cm-wide zone of inhibition; ++ = >1- to 2-cm-wide
zone of inhibition; +++ = 3-cm-wide zones of inhibition; NT = not tested. For all other tests: = trait absent; + = trait present.
10
Biomass reduction of Euphorbia esula/virgata by insect/bacterial combinations
0.1%
4-Methyl umbelliferyl
-d-glucoside
0.1%
4-Methylumbelliferyl
N-acetyl
-d-glucosaminide
Ostazin Brilliant Red
hydroxyethylcellulose
Remazol
Brilliant Blue
Xylan
Polygalacturonase
Cellulase
+
+
+
+
+
+
+
+
+
+
+
11
XII International Symposium on Biological Control of Weeds
Table 2.
Effect on Euphorbia esula/virgata of bacteria
with various spectra of hydrolytic enzyme production in vitro in combination with Aphthona
spp. Means with different letters are significantly different (P = 0.05) as determined using
Waller and Duncans (1979) exact Bayesian kratio least significant difference rule.
Origin
Treatment
Knutson Creek
Aphthona + Bacillus
thuringiensis 124
Aphthona +
Microbacterium
liquefaciens 223
Aphthona +
Brevibacterium
iodinum 116
Aphthona +
Stenotrophomonas
maltophilia 144d
Aphthona +
Ochrobacterium
anthropii 145
Aphthona +
Corynebacterium
acquaticum 207b
Control + Aphthona
Knutson Creek
Cottonwood
Cottonwood
Cottonwood
Knutson Creek
Mean dry
weight (g)
16.2 a
17.4 a
19.0 ab
37.1 bc
45.2 c
47.3 c
49.4 c
amongst those that are deleterious to plant growth can,
in combination with insects, cause dramatically negative effects on invasive weed growth compared with
insects alone.
References
Barros, M.E.C. and Thomson, J.A. (1987) Cloning and expression in Escherichia coli of a cellulase gene from Ru
minococcus flavefaciens. Journal of Bacteriology 169,
17601762.
Bartlett, M.S. and Kendall, D.G. (1946) The statistical analysis
of variancesheterogeneity and the logarithmic trans
formation. Journal of the Royal Statistical Society Sup
plement 8, 128138.
Biely, P., Mislovicova, D. and Toman, R. (1985) Soluble
chromogenic substrates for the assay of endo-1, 4-betaxylanases and endo-1, 4-beta-glucanases. Analytical Bio
chemistry 144, 142146.
Caesar, A.J. (1994) Comparative virulence and host range
of strains of Rhizoctonia solani AG-4 from leafy spurge.
Plant Disease 78, 183186.
Caesar, A.J. (1996) Identifcation, pathogenicity and comparative virulence of Fusarium spp. associated with stand declines of leafy spurge (Euphorbia esula) in the Northern
Plains. Plant Disease 80, 13951398.
Caesar, A.J. (2003) Synergistic interaction of soilborne plant
pathogens and root-attacking insects in classical biological control of an exotic rangeland weed. Biological Con
trol 28, 144153.
12
Canganella, F., Paparatti, B. and Natali, V. (1994) Microbial
species isolated from the bark beetle Anisandrus dispar F.
Microbiological Research 149, 123128.
Cho, S.J., Lim, W.J., Hong, S.Y., Park, S.R. and Yun, H.D.
(2003) Endophytic colonization of balloon flower by antifungal strain Bacillus sp. CY22. Bioscience Biotechnol
ogy and Biochemistry 10, 21322138.
Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M.,
Morin, L. and Scott, J.K. (eds) (2004) Proceedings of
the XI International Symposium on Biological Control of
Weeds, CSIRO Entomology, Canberra, Australia.
Fahey, P.C. and Hayward, A.C. (1983) Media and methods
for isolation and diagnostic tests. In: Persley, A.G. and
Fahey, P.C. (eds) Plant Bacterial Diseases: A Diagnostic
Guide. Academic Press, New York, pp. 337378.
Hankin, L. and Lacy, G.H. (1984) Pectinolytic microorganisms. In: Speik, M.L. (ed.) Compendium for the Microbio
logical Examination of Foods. American Public Health
Association, Washington, DC, pp. 176183.
Kremer, R.J. and Kennedy, A.C. (1996) Rhizobacteria as biocontrol agents of weeds. Weed Technology 10, 601609.
Kremer, R.J., Caesar, A.J. and Souissi, T. (2006) Soilborne
microorganisms of Euphorbia are potential biological
control agents of the invasive weed leafy spurge. Applied
Soil Ecology 32, 2737.
Oka, N., Hartel, P.G., Finlay-Moore, O., Gagliardi, J., Zuberer, D.A., Fuhrmann, J.J., Angle, J.S. and Skipper, H.D.
(2000) Misidentification of soil bacteria by fatty acid methyl ester (FAME) and BIOLOG analyses. Biology and
Fertility of Soils 32, 256258.
Santos, T., del Rey, F., Conde, J., Villanueva, J.R. and
Nombela, C. (1979) Saccharomyces cerevisiae mutant
defective in exo-1,3-beta-glucanase production. Journal
of Bacteriology 139, 333338.
Scott, J.H. and Schekman, R. (1980) Lyticase: endoglucanase
and protease activities that act together in yeast cell lysis.
Journal of Bacteriology 142, 414423.
Siciliano, S.D. and Germida, J.J. (1999) Taxonomic diversity
of bacteria associated with the roots of field-grown transgenic Brassica napus cv. Quest, compared to the nontransgenic B. napus cv. Exel and B. rapa cv. Parkland.
FEMS Microbiology Ecology 29, 263272.
Sipat, A., Taylor, K.A., Lo, R.Y., Forsberg, C.W. and Krell, P.J.
(1987) Molecular cloning of a xylanase gene from Bacte
roides succinogenes and its expression in Escherichia coli.
Applied and Environmental Microbiology 53, 477481.
Spencer, N.R. (ed.) (2001) Proceedings of the X International
Symposium on Biological Control of Weeds, July 414,
1999, Montana State University, Bozeman, MT.
Spiteller, D., Dettner K. and Boland W. (2000) Gut bacteria may be involved in interactions between plants, herbivores and their predators: microbial biosynthesis of N
acylglutamine surfactants as elicitors of plant volatiles.
Biological Chemistry 381, 755762.
Waller, R.A. and Duncan, D.B. (1969) A Bayes rule for the
symmetric multiple comparison problem. Journal of the
American Statistical Association 64, 14841499.
Zinniel, D.K., Lambrecht, P., Harris, N.B., Feng Z., Kuczmarski, D., Higley, P., Ishimaru, C.A., Arunakumari, A.,
Barletta, R.G. and Vidaver, A.K. (2002) Isolation and
characterization of endophytic colonizing bacteria from
agronomic crops and prairie plants. Applied and Environ
mental Microbiology 68, 21982208.
Rhizosphere bacterial communities
associated with insect root herbivory of an
invasive plant, Euphorbia esula/virgata
A.J. Caesar1 and T. Caesar-Ton That2
Summary
The invasive perennial plant of Eurasian origin, Euphorbia esula/virgata, has been successfully controlled over large areas in North America with a synergism between larvae of Aphthona spp. and
soilborne plant pathogens. However, a multitude of sites is not yet under control. Studies are needed
on how flea beetle root herbivory may alter the microbial ecology of the rhizosphere of E. esula/virgata and how the resulting rhizosphere community may affect the synergism. Studies were undertaken at Theodore Roosevelt National Park from 2001 to 2003 to identify the predominant culturable
prokaryotic species found in the rhizospheres of E. esula/virgata. The hypothesis was that distinct
rhizosphere communities of E. esula/virgata would be associated with root herbivory by the flea
beetle Aphthona compared with rhizospheres of E. esula/virgata from stands without insect presence.
Stands with and without resident populations of Aphthona spp. were assayed by spiral plating root
washes of E. esula/virgata and selecting colonies from the most dilute portion of the spiral (deemed as
predominant). Gas chromatographic analysis of fatty acid methyl ester was performed on the resulting
pure cultures to identify the isolates and further characterize community structures using principal
component analysis. Pseudomonas syringae van Hall, Pseudomonas cichorii (Swingle) Stapp, Erwinia chrysanthemii Burkholder, all plant pathogens, were associated exclusively with herbivory by
Aphthona flea beetles. Conversely, Variovorax Willems et al. 1991 and Aquaspirillum Hylemon
et al. 1973 spp. were a greater proportion of predominant species from roots without Aphthona present.
There were also differences in the occurrence of the root pathogen antagonistic Pantoea agglomerans
Gavini et al. 1989 and Stenotrophomonas maltophilia (Hugh 1981) Palleroni and Bradbury 1993.
Keywords: synergism, trophic interactions, plant pathogens, soilborne, microbial
ecology.
Introduction
characterized by rapid reductions in stand density, is
caused by insect/plant pathogen synergisms (Caesar,
2003). Given that the mechanisms driving the successful biological control include soilborne microbes such as
plant pathogenic Fusarium spp. Link ex Gray, Rhizoctonia solani Kuhn and other fungi, possible explanations
for the prevalence of unimpacted sites, despite establishment of insect root herbivores, may also be microbial in
nature. This aspect has not been investigated previously.
It has been increasingly accepted within the field of
biological control that microbial interactions are a considerable, significant factor in the biological control of
invasive plants (Bacher et al., 2002; Lym and Carlson,
2002; Sing et al., 2005; Butler et al., 2006). This complements a large body of literature showing that exotic
plant invasion is both affected by and affects the soil microbial ecology (Belnap and Phillips, 2001; Ehrenfeld
et al., 2001; Ehrenfeld, 2003; Kourtev et al., 2002, 2003),
including effects on plant succession (Van der Putten
Biological control of plant species of Eurasian origin
that are invasive in North American has resulted in considerable success in reducing population densities of
several species. One such success concerns the deeprooted perennial Euphorbia esula/virgata, regarded as
a fully achieved case of biological control of an invasive plant species. However, the proportion of impacted
sites amongst all infested locations has remained at ca
33% (Caesar, 2003; Kalischuk et al., 2004; Hodur et
al., 2006). Successful biological control of leafy spurge,
Pest Management Research Unit, USDAARS Northern Plains Agricultural Research Laboratory, Sidney, MT 59270, USA.
2
Agricultural Systems Research Unit, USDAARS Northern Plains Agricultural Research Laboratory, Sidney, MT 59270, USA.
Corresponding author: A.J. Caesar <[email protected]>.
CAB International 2008
13
XII International Symposium on Biological Control of Weeds
et al., 1993; Bever et al., 1997). There are also indications that the process of biological control with root herbivore and soilborne microbes, which can be viewed as
an accelerated form of negative feedback (Caesar, 2005),
may affect patterns of succession following E. esula/virgata (Butler et al., 2006). These documented interactions
between plants and soil microbes and amongst plants,
root herbivores and microbial synergists thus have great
implications for biological control, plant succession and
restoration of native plant communities. Several questions arise from this body of findings. Concerning the effects of biological control, these include the following:
populations without Aphthona flea beetle activity. It
was surmised that identification of prokaryotes present
in the highest numbers would be of critical interest to
elucidate their possible role and mode of action in relation to biological control.
Previous studies by the senior author have shown
the effects of fungi that were found in insect-damaged
root tissue of such invasive species as E. esula/virgata,
Acroptilon repens (L.) DC and Centaurea maculosa
Lam. But few previous studies have attempted to identify prominent or predominant members of the prokaryotic microflora in response to exotic plant invasion
and establishment.
1. Is there any as sociation between insect (Aphthona
spp.) damage to roots of E. esula/virgata and predominant culturable prokaryote species? Is the lack
of stand reduction despite the establishment of insect root herbivores as biological control agents
attributable to microbial factors other than plant
pathogens?
2. What alternatives are available when large numbers
of infestations remain unaffected by the most successful agents and can the percentage of impacted
sites be increased through microbial means?
3. How is microbial negative feedback (the accumulation of deleterious microbes in response to individual plant species), shown in a number of cases
with invasive plants, manifested in the predominant
microbial species that occur in response to insect
damage?
a. How does insect damage to roots of the invasive
perennial E. esula/virgata affect the structure of
prokaryotic microbial communities compared
with roots of plants from populations with little
or no insect presence?
b. Do the predominant or prevailing culturable bacteria and actinomycetes from the rhizospheres of
populations of E. esula/virgata with insect activity act as antagonists to plant pathogens or as
low-level plant pathogens?
Materials and methods
Sites within the Theodore Roosevelt National Park in
North Dakota, United States, with infestations of E.
esula/virgata under observation since 1992 were selected for sampling based on the presence or absence of
adult flea beetles on the stand, the former status being
an indicator of larval attack on the roots earlier in the
season, as confirmed by examining roots in work preliminary and subsequent to the work described herein.
Five plants within each sampled stand, which ranged in
size from, were selected haphazardly for rhizosphere
soil samples but were usually 0.51 m from the edge of
a given stand. Stands ranged in size from 0.2 to 1.2 ha.
Three soil cores of 20 cm in diameter to a depth of ca
15 cm containing roots of leafy spurge were taken from
around each of the plants. In the laboratory, spurge
roots were identified, removed from soil cores and
transferred to plastic bags (90 160 mm; Intersciences
Laboratories, Weymouth, MA) containing 9 ml of pH 7
phosphate buffer and subjected to 1 minute of agitation
with a Stomacher 80 (Seward Medical, London, UK).
Soil suspensions were plated on 0.3% tryptic soy broth
agar (TSBA) medium in triplicate using a spiral plater
(Don Whitley Scientific, West Yorkshire, UK). A spiral
plating method to serially dilute rhizosphere soil was
used to afford a non-random means of selecting colonies from the most dilute portion of the spiral. Plates
were incubated at 2028C for 35 days. Five bacterial
colonies found at the end of each spiral were collected
from each plate and thus represented the predominant
E. esula/virgata rhizosphere bacteria for each sampled
site (Caesar-TonThat et al., 2007). For identification of
isolates, fatty acid methyl ester (FAME) profiles were
obtained. FAME profiles are routinely used to identify
genera, species and strains of bacteria (Cavigelli et al.,
1995; Ibekwe and Kennedy, 1999). In our case, FAMEs
were used both as the basis of identification of isolates
and further afforded the analysis of intraspecific differences amongst isolates or amongst unidentified isolates
with similar taxonomic affinities. FAMEs were obtained
by saponification, methylation and extraction following the MIDI system (Microbial Identification System;
Microbial ID, Newark, NJ). MIDI Microbial Identifi-
No previous study has sought to examine effects of
root herbivory on the prokaryotic rhizosphere community in relation to biological control of the plant host.
Thus, the objectives of this study were to assess communities of culturable prokaryotes associated with rhizospheres of E. esula/virgata at sites with heavy flea
beetle activity and compare with such communities occurring at locations with no detectable insect activity.
We sought to examine and discuss the implications of
any trends that the presence of specific bacteria might
be indicative of. For example, the presence of certain
Erwinia spp. would be indicative of soft rot. We hypothesized that prokaryotic rhizosphere communities
from populations of the invasive plant E. esula/virgata
that were damaged by larvae of the flea beetles Aphthona nigriscutis Foudras and/or A. lacertosa (Rosenhauer) would exhibit considerable distinctions from
rhizosphere microbial communities associated with
14
Rhizosphere bacterial communities associated with insect root herbivory of an invasive plant
cation Software (Sherlock TSBA50 Library; Microbial
ID) was used to identify the isolates. Stenotrophomonas
maltophilia (ATCC 13637) was used as a reference.
Only strains with a similarity index (SIM) of 0.300
were considered a good match (Siciliano and Germida,
1999; Oka et al., 2000). Non-matched isolates were
considered conclusively analysed if the percentage of
their named peaks was >85%, although they were not
assigned identification because of lack of information
in MIDI Aerobic Bacteria Library TSBA50. Therefore,
they were included in all analyses. The FAME structural classes were categorized into saturated straightchain fatty acids, branched straight-chain fatty acids,
monounsaturated fatty acids and hydroxyl fatty acids.
These classes were used as indicators for particular
groups of microorganisms (Zelles et al., 1992; Larkin,
2003). The proportion of fatty acid structural classes
(expressed in percentage of total fatty acids) were combined from bacterial isolates belonging to a same species or to the same genus (in the case of Pseudomonas
spp.), and mean values were compared amongst the
groups of species. Principal component analysis (PCA)
was performed on community FAME data from different treatments (locations with or without insects). The
FAME profiles of bacterial isolates were compared by
PCA using JMP v6 (SAS, Cary, NC)
The objectives of the present study were to elucidate
the bacterial community structure associated with insect herbivory of an exotic, invasive species. Previous
studies have focused on key soilborne fungi that are associated with herbivory that have been attributed with
causing biological control of invasive plants (Caesar,
2003).
Although there have been studies on above-ground
herbivory and soil biodiversity, on the effects of invasive species on soil microbial community structure
and on the effects of above-ground herbivory on plant
invasion (Maron and Vila, 2001), few studies have examined the effects of root herbivory and rhizosphere
microbial community structure. The epicenter of invasiveness may be the rhizosphere interactions amongst
plants, microbes and root herbivores. Understanding
the effects of specific soil biota can be useful to make
predictions about the relative importance of soil organisms in the invasion process, the rate at which stand
reductions occur and the likelihood of successful restoration of native plant communities following successful biological control.
The approach taken here of focusing on culturable
bacteria is justified on several grounds. Many, if not
most of the important parameters that relate to soil and
plant health, such as nitrogen cycling, mineralization
and soil structure (aggregation, porosity for water holding capacity and respiration), can at present be linked
exclusively to culturable soil bacteria. Of the soil
microbes known to contribute to such important soil
processes as the control of plant diseases, insects and
weed pests; beneficial symbiotic associations between
bacteria and plants; the recycling of plant nutrients;
and the maintenance of soil structure, all are culturable microbes (Caesar-TonThat et al., 2007). Although
other species clearly may also play prominent roles in
soil biology, the development of tools for assessing the
phenotypes and thus the functional role(s) of such microbes is still at a nascent stage (Liu et al., 2006). Also,
sheer numbers of organisms are a likely indicator of
key roles they play in any ecological realm, thus isolation and study of the bacterial species present at the
highest population levels (which we have deemed predominant) should be the point of departure in assessing
the significance of rhizosphere community composition in relation to insect herbivory or herbicide application, for example.
Results and discussion
There were large differences in the composition of
the predominant Gram-negative rhizosphere bacteria
based on the presence or absence of the root-attacking
(as larvae) species A. nigriscutis and A. lacertosa, in
each of the 3 years of this study (Figs. 14 and Tables
13). Particularly striking were the percent differences,
based on the presence or absence of the insects, of such
plant pathogenic species as Pseudomonas syringae van
Hall, Pseudomonas cichorii (Swingle) Stapp and Erwinia chrysanthemi, Burkholder, all found either exclusively or with greater frequency in rhizospheres of
E. esula/virgata with Aphthona spp. present.
Bacteria such as Stenotrophomonas spp. Palleroni
and Bradbury 1993, Pseudomonas chlororaphis Guignard and Sauvageau 1894 and Pantoea agglomerans
Gavini et al. 1989, with implications for possibly protecting the plant from the more lethal fungal root infections (which are operative as a key factor for rapid
E. esula/virgata stand mortality), were present in both
Aphthona-populated sites and sites lacking the flea
beetle. However, S. maltophilia (Hugh, 1981) Palleroni and Bradbury 1993 was consistently found as a
predominant species with much greater frequency in
Aphthona-populated sites. These sites remained static
in regard to stand density throughout the interval of
the study (data not shown). Despite apparent insect
damage-based stimulation of overall microbial biomass,
Stenotrophomonas spp., well-known as antagonistic to
plant pathogens, was most favored. There were also
Gram-positive bacteria that have shown a capacity in
combination with insects to reduce biomass of E. esula/
virgata (Caesar and Kremer, 2008). Spurge infestations
with both of these characteristics have persisted well
after other infestations of E. esula/virgata have been
dramatically reduced in density at TRNP. This may indicate that the complexity of the microbial community
may contribute to pre-empting or antagonizing the insect/plant pathogen synergisms that cause more rapid
stand reductions in biological control of this highly
15
XII International Symposium on Biological Control of Weeds
Figure 1.
PCA of the rhizosphere community structure of predominant bacteria in relation
to herbivory by Aphthona spp. Cumulative data from stands of four sites: () two
sites with and () two sites without the insect, 2001.
aggressive, deep-rooted perennial invasive plant. Pertaining to complexity, some studies have indicated that
soil microbial complexity is associated with control of
the soilborne plant pathogen Rhizoctonia solani (Garbeva et al., 2006), and others have found no such association per se (Hiddink et al., 2005). Neither study
Figure 2.
could identify specific components of the respective
communities, so it is difficult to assess the underlying
basis for the differing findings. Thus, our approach,
although not as comprehensive as culture-independent
methods of assessing the entire community nonetheless
permits the identification of culturable species occurring
PCA of the rhizosphere community structure of predominant bacteria in relation
to herbivory by Aphthona spp. Cumulative data from stands of four sites: () two
sites with and () two sites without the insect, 2002.
16
Rhizosphere bacterial communities associated with insect root herbivory of an invasive plant
Figure 3.
PCA of the rhizosphere community structure of predominant bacteria in relation
to herbivory by Aphthona spp. Cumulative data from stands of four sites: () two
sites with and () two sites without the insect, 2003.
Figure 4.
PCA of the rhizosphere community structure of predominant bacteria in relation
to herbivory by Aphthona spp. Cumulative data from stands of four sites: () two
sites with and () two sites without the insect, 2003. The ellipses indicate pseudomonad and enteric groupings, wherein some groupings were nearly exclusive
to rhizospheres of stands without Aphthona (enterics and Pseudomonas chlororaphis), whereas other groupings were mixed but with some evident phenotypic
distinctions and differences in numbers (pseudomonads).
17
XII International Symposium on Biological Control of Weeds
Table 1.
Table 3.
Predominant microbial species from roots of
leafy spurge at various sites with or without
Aphthona, 2001.
Microbial species
Pseudomonas putida
P. syringae
Pseudomonas
fluorescens
P. cichorii
Flavimonas spp.
Stenotrophomonas spp.
Sphingomonas spp.
Rhizobium spp.
Zooglea
Species
Percentage of isolates
(of 81 isolates)
With
Aphthona
0
6.8
20.5
Without
Aphthona
22.2
0
13.3
0
0
38
6.8
9.1
4.6
6.7
8.9
6.6
0
8.9
4.4
Pseudomonas spp.
P. syringae
P. cichorii/
viridiflava
P. fluorescens
P. putida
P. chlororaphis
P. huttiensis
Enterobacteria
Erwinia
chrysanthemi
Pa. agglomerans
Other enterobacteria
Other
Stenotrophomonas
spp.
Lysobacter
enzymogenes
Rhizobium spp.
Variovorax
paradoxus
Vibrio hollisiae
Percentage of isolates
(of 121 isolates)
With
Aphthona
Without
Aphthona
17.3
4.8
0
0
12.5
3.3
1.9
1.9
15.6
5.9
0
6.3
2.9
3.9
0
0
6.3
19.2
6.3
4.8
12.5
0
21.9
15.6
2.9
With
Aphthona
Without
Aphthona
21
13.6
9.8
2.7
34.5
5.5
0
7.3
13.6
9.2
1.8
0
7.6
1.3
5.6
3.3
1.8
1.3
7.3
12.7
logical control-based stand reductions (Larson and
Grace, 2004), indicating that soil microbes have a great
effect on whether biological control ultimately results
in stand reductions or in a static state of target species
density. The alterations in soil microbial community
structure also have strong implications for the possibility of restoration of native plant communities. Analyses
of Gram-positive rhizosphere bacterial communities,
which contain isolates shown to cause 2865% reductions in the biomass of E. esula/virgata (Caesar and
Kremer, 2008) associated with Aphthona spp. herbivory were the subject of a companion study intended to
be published separately.
Predominant species from roots at various sites
with and without Aphthona, 2002.
Species
Percentage of isolates
(of 215 isolates)
Pseudomonas spp.
P. putida
P. chlororaphis
P. agarici
P. vancouverensis
Enterobacteria
Pa. agglomerans
Other enteric species
Other species
Stenotrophomonas spp.
Rhizobium spp.
Zoogloea spp.
Aquaspirillum
autotrophicum
at the highest population levels in the rhizospheres of
E. esula/virgata. We propose that complexity within
such functional groups as those that contain pathogenantagonistic or plant-beneficial strains, rather than overall diversity, may be more pertinent to such analyses.
Previous studies have indicated that cultural methods
track the results obtained through culture-independent
methods (Garbeva et al., 2006).
Work by others suggests that biological control insects themselves may not be the prime factors in bioTable 2.
Predominant species from roots at various sites
with and without Aphthona, 2003.
References
Bacher, S., Friedli, J. and Schr, I. (2002) Developing in
diseased host plants increases survival and fecundity in
a stem-boring weevil. Entomologia Experimentalis et Applicata 103, 191195.
Belnap, J., Phillips, S.L., Sherrod, S.K. and Moldenke, A.
(2005) Soil biota can change after exotic plant invasion:
does this affect ecosystem processes? Ecology 86, 3007
3017.
Bever, J.D., Westover, K.M. and Antonovics, J. (1997) Incorporating the soil community into plant population dynamics: the utility of the feedback approach. Journal of
Ecology 85, 561573.
Butler, J.L., Parker, M.S. and Murphy, J.T. (2006) Efficacy
of flea beetle control of leafy spurge in Montana and
South Dakota. Rangeland Ecology and Management 59,
453461.
Caesar, A.J. (2003) Synergistic interaction of soilborne plant
pathogens and root-attacking insects in classical biological control of an exotic rangeland weed. Biological Control 28, 144153.
Caesar, A.J. (2005) Melding ecology, classical weed biocontrol and plant microbial ecology can inform improved
18
Rhizosphere bacterial communities associated with insect root herbivory of an invasive plant
practices in controlling invasive plant species. Biological
Control 35, 240246.
Caesar, A. and Kremer, R.J. (2008) Biomass reduction of E.
esula/virgata by insect/bacterial combinations. In: Julien,
M.H., Sforza, R., Bon, M.C., Evans, H.C., Hatcher, P.E.,
Hinz, H.L. and Rector, B.G. (eds) Proceedings of the XII
International Symposium on Biological Control of Weeds.
CAB International, Wallingford, UK.
Caesar-TonThat, T.C., Caesar, A.J., Gaskin, J.F., Sainju,
U.M. and Busscher, W.J. (2007) Taxonomic diversity of
predominant culturable bacteria associated with microaggregates from two different agroecosystems and their
ability to aggregate soil in vitro. Applied Soil Ecology 36,
1021.
Cavigelli, M.A., Robertson, G.P. and Klug, M.K. (1995)
Fatty acid methyl ester (FAME) profiles as measures of
soil microbial community structure. Plant and Soil 170,
99113.
Ehrenfeld, J.G. (2003) Effects of exotic plant invasions on
soil nutrient cycling processes. Ecosystems 6, 503523.
Ehrenfeld, J.G., Kourtev, P.S. and Huang, W. (2001) Changes
in soil functions following invasions of exotic understory
plants in deciduous forests. Ecological Applications 11,
12871300.
Garbeva, P., Postma, J.J., van Veen, A. and van Elsas, J.D.
(2006) Effect of above-ground plant species on soil microbial community structure and its impact on suppression
of Rhizoctonia solani AG3. Environmental Microbiology
8, 233246.
Hiddink, G.A., Termorshuizen, A.J., Raaijmakers, J.M. and
van Bruggen, A.H.C. (2005) Effect of mixed and single
crops on disease suppressiveness of soils. Phytopathology
95, 13251332.
Hodur, N.M., Leistritz, F.L. and Bangsund, D.A. (2006)
Biological control of leafy spurge: utilization and implementation. Rangeland Ecology and Management 59,
445452.
Ibekwe, A.M. and Kennedy, A.C. (1999) Fatty acid methyl
ester (FAME) profiles as a tool to investigate community
structure of two agricultural soils. Plant and Soil 206,
151161.
Kalischuk, A.R., Bourchier, R.S. and McClay, A.S. (2004)
Post hoc assessment of an operational biocontrol program:
efficacy of the flea beetle Aphthona lacertosa Rosenhauer
(Chrysomelidae: Coleoptera), an introduced biocontrol
agent for leafy spurge. Biological Control 29, 418426.
Kourtev, P.S., Ehrenfeld, J.G. and Haggblom, M. (2002) Exotic plant species alter the microbial community structure
and function in the soil. Ecology 83, 31523166.
Kourtev, P.S., Ehrenfeld, J.G. and Haggblom, M. (2003)
Experimental analysis of the effect of exotic and native
plant communities on the structure and function of soil
microbial communities. Soil Biology and Biochemistry
35, 895905.
Larkin, R.P. (2003) Characterization of soil microbial communities under different potato cropping systems by
microbial population dynamics, substrate utilization and
fatty acid profiles. Soil Biology and Biochemistry 35,
14511466.
Larson, D.L. and Grace, J.B. (2004) Temporal dynamics of
leafy spurge (Euphorbia esula) and two species of flea
beetles (Aphthona spp.) used as biological control agents.
Biological Control 29, 207214.
Liu, Y., Jianrong, L., Lee, S., Goh, C.S., Gerstein, M.B.
and Lussier,Y. (2006) An integrative genomic approach
to uncover molecular mechanisms of prokaryotic traits.
doi:10.1371/journal.pcbi.0020159.eor.
Lym, R.G. and Carlson, R.B. (2002) Effect of leafy spurge
(Euphorbia esula) genotype on feeding damage and reproduction of Aphthona spp.: implications for biological
weed control. Biological Control 23, 127133.
Lym, R.G. and Nelson, J.A. (2000) Biological control of
leafy Spurge (Euphorbia esula) with Aphthona spp. along
railroad right-of-ways. Weed Technology 14, 642646.
Maron, J.L. and Vila, M. (2001) Do herbivores affect plant
invasion? Evidence for the natural enemies and biotic resistance hypotheses. Oikos 95, 363373.
Oka, N., Hartel, P.G., Finlay-Moore, O., Gagliardi, J., Zuberer, D.A., Fuhrmann, J.J., Angle, J.S. and Skipper, H.D.
(2000) Misidentification of soil bacteria by fatty acid
methyl ester (FAME) and BIOLOG analyses. Biology and
Fertility of Soils 32, 256258.
Siciliano, S.D. and Germida, J.J. (1999) Taxonomic diversity
of bacteria associated with the roots of field-grown transgenic Brassica napus cv. Quest, compared to the nontransgenic B. napus cv. Exel and B. rapa cv. Parkland.
FEMS Microbiology Ecol. 29, 263272.
Sing, S.E., Peterson, R.K.D., Weaver, D.K., Hansen, R.W.,
Markin, G.P. (2005) A retrospective analysis of known
and potential risks associated with exotic toadflax-feeding
insects. Biological Control 35, 276287.
Van der Putten, W., Van Dijk, C., Peters, B. (1993) Plantspecific soilborne diseases contribute to succession in
foredune vegetation. Nature 362, 5356.
Van der Stoel, C.D., Van der Putten, W.H. and Duyts, H.
(2002) Development of a negative plantsoil feedback in
the expansion zone of the clonal grass Ammophila arenaria following root formation and nematode colonization. Journal of Ecology 90, 978988.
Zelles, L., Bai, Q.Y., Beck, T. and Beese, F. (1992) Signature
fatty acids in phospholipids and lipopolysaccharides as
indicators of microbial biomass and community structure
in agricultural soils. Soil Biology and Biochemistry 24,
317323.
19
The endophyte-enemy release hypothesis:
implications for classical biological
control and plant invasions
H.C. Evans
Summary
Fungal endophytes are asymptomless colonizers of higher plants for all, or a part, of their life cycles.
They range from latent pathogens to symbionts. There is increasing evidence that some form mutually
beneficial, highly specialized or co-evolved associations with their hosts and that they provide the plant
with an additional armoury to combat abiotic and biotic stresses, including pests and diseases. Thus,
there may be a trade-off between reduced growth (in the short term), as nutrients are sequestered by
the fungal mutualist, but increased overall long-term fitness as natural-enemy pressure is decreased.
This tripartite balance may be lost when plants arrive in exotic ecosystems with incomplete guilds of
both co-evolved endophytes and natural enemies. The enemy release hypothesis (ERH) explains why
alien plants can become invasive. It is now hypothesized that another, more cryptic but still significant
factor could also be involved: the presence or absence of mutualistic endophytes. Those neophytes arriving without co-evolved natural enemies but with mutualistic co-evolved endophytes would have a
double advantage over local competitors. Such endophytic-enriched, alien-invasive weeds and those
that form mutualistic associations with indigenous endophytes could help to explain the inconsistencies of some classical biological control introductions. Similarly, those alien plants that arrive and
remain endophyte-free and without co-evolved natural enemies would have a distinct competitive
advantage because they would have more resources to allocate to growth and reproduction, given,
of course, that there are no significant pressures from indigenous natural enemies or that sufficient
auto-defences are retained to overcome them. Such endophyte-depauperate alien-invasive weeds,
however, remain highly susceptible to co-evolved natural enemies. This may explain the silver bullet phenomenon, whereby the introduction of a single classical biological control agent can achieve
complete control. This endophyte-enemy release hypothesis (E-ERH) is discussed with examples.
Keywords: coevolution, fungal mutualists, plant fitness.
Introduction
partnerships with fungi, which enabled them to survive
the stresses of life on dry land, where water and nutrients were the main constraints to colonization: Once
an endosymbiotic relationship of a fungus with an alga
was achieved, a blueprint for a terrestrial plant was
drawn (Pirozynski and Malloch, 1975).
Here, the form and function of mutualistic endophytic
fungi is reassessed in the light of recent studies, and
their possible significance in plant ecology is explored,
leading to the hypothesis that their presence or absence
may explain, at least in part, why some alien plants become invasive and why classical biological control can
be so unpredictable as a management strategy.
There is now overwhelming evidence that all plants in
natural ecosystems have developed symbiotic relationships with fungi (Rodriguez et al., 2004) and that the
mutualistic ones, especially those involving vesicular
arbuscular mycorrhizae (VAM), are ancient in origin
(Brundrett, 2002). Indeed, it has been suggested that
such associations were pivotal to the colonization of
land by plants (Simon et al., 1993; Blackwell, 2000;
Brundrett, 2002). However, this mycotrophic theory
had been discussed much earlier by Pirozynski and
Malloch (1975), who hypothesized that the evolution
of plants was made possible only through mutualistic
Definitions and concepts
CABI, E-UK, Bakeham Lane, Egham, Surrey TW20 9TY, UK
<[email protected]>.
CAB International 2008
There has been considerable debate and controversy
as to the correct usage of the term endophyte, origi20
The endophyte-enemy release hypothesis
nally coined by the founder of modern mycology,
Heinrich de Bary in 1866 (Wilson, 1995). The ambiguities and confusion have been such (Wilson, 1993,
1995; Wennstrom, 1994) that it has been recommended
the term should be defined according to context (Kirk
et al., 2001). Here, and specifically in relation to the
proposed hypothesis, the use is restricted to fungi that
invade living plants and colonize them without causing
visible or immediate symptoms. Mycorrhizal fungi are
excluded because, as dual plantsoil inhabitants, they
are restricted to root systems in which there is synchronized plantfungus development with nutrient transfer
at specialized interfaces (Schulz and Boyle, 2005). In
contrast, endophytic fungi lack the means of acquiring nutrients from soil but have evolved mechanisms
that enable them to survive and live asymptomatically,
at least initially, within the roots, stems and leaves of
healthy plants (Brundrett, 2002).
Fungal endophyte associations with their host plants
have been described as a continuum (Saikkonen et al.,
1998; Schulz and Boyle, 2005), ranging from parasitism (amensalism) through commensalism to mutualism (Lewis, 1985). This paper concentrates on the
mutualists, those that form intimate associations with
their plant hosts that are beneficial to both, providing
protection from environmental stresses and microbial
competition, as well as nutrients, for the fungus and
increased resistance or tolerance to both abiotic and biotic stresses for the plant.
Most evidence for this increased fitness comes from
studies of forage and turf grasses, particularly from
plant associations in the subfamily Pooideae, with
balansiaceous fungi belonging to the genus Neotyphodium (Clavicipitaceae: Hypocreales) because of their
ecological and economic importance (Schardl and
Phillips, 1997; Clay and Schardl, 2002; Bouton and
Hopkins, 2003). Neotyphodium is a genus of highly
specialized or obligate endophytic species that live
systemically and intercellularly in all the aerial parts
of their hosts and that are transmitted vertically in the
grass seeds. There is increasing evidence, however,
that similar benefits and increased plant fitness are also
conferred to both non-woody and woody dicot hosts by
horizontally transmitted, facultative endophytic fungi
(Narisawa et al., 2000; Wilson, 2000; Arnold et al.,
2003; Clay, 2004).
Obviously, there is a price to pay by the host plant
for harbouring beneficial endophytes; although, as yet,
quantitative data are lacking. Nevertheless, there is circumstantial evidence from field trials with turf grasses
(Poa spp.) that shows that there is a significant cost
involved because endophyte-free plants were notably
more vigorous early in the season than those inoculated with endophytes. In contrast, later in the season,
those with endophytes had markedly outperformed
the plants lacking endophytes, as pests, diseases and
drought stress took their toll (author, Rutgers University Experimental Station, 2001, personal observation).
Moreover, there is indirect evidence of trade-offs in
mycorrhizal fungi, where the carbon costs to the plant
of supporting these mutualists have been found to be
significant (Douds et al., 1988).
Background to the hypothesis
Mutualistic endophytic fungi offer a variety of potential
benefits to their host plants including growth enhancement, tolerance to abiotic factors (including drought,
heat and heavy metals) and resistance to pests and
diseases (Redman et al., 2001; Rudgers et al., 2004;
Schulz and Boyle, 2005). The mechanisms involved
may be wide-ranging and complex, the result of an ancient association (coevolution). In the case of conferring protection against plant pathogens, for example,
these could range from antagonism, mycoparasitism,
competitive displacement, to induced resistance (Evans
et al., 2003). It has now been established that anti-fungal,
as well as anti-herbivore, secondary metabolites are
produced by many endophytes (Latch, 1993; Christensen, 1996; Clay, 1997; Schardl and Phillips, 1997;
Stone et al., 2000). In addition, some produce novel
growth-enhancement compounds (Varma et al., 1999),
whereas endophyte-free plants have been shown to activate defence mechanisms much more slowly than those
with mutualistic endophytes (Rodriguez et al., 2004),
suggesting that they are also involved in inducing host
resistance to pests and diseases. Finally, more recent
studies have revealed unique trophic interrelationships
between endophytes and their plant hosts, which enhance tolerance to both drought and heat (Rodriguez
and Redman, 2005; Marquez et al., 2007).
The endophyte-enemy release hypothesis
Plants in their centres of origin live in mutualistic
relationships with a guild of specialized or co-evolved
endophytic fungi that increase their tolerance of, or resistance to, both abiotic and biotic pressures, including
co-evolved natural enemies. This protection comes at a
price, with a trade-off in plant resources. Alien plants,
especially dicot hosts, arriving in exotic ecosystems
would have a depauperate endophytic mycobiota, freeing up resources for increased growth and reproduction. This, together with the absence of co-evolved
natural enemies (enemy release hypothesis, ERH; Keane and Crawley, 2002), would enhance significantly
their fitness. Given that these endophyte-free aliens
have sufficient auto-defence mechanisms to overcome
the pressure from indigenous natural enemies, they
then would have increased competitive advantage.
The result would be a dominance of these enhanced or
favoured species that would increase over successive
generations. Thus, neophytes with weedy traits would
tend to become dominant and invasive. However, such
plants would be highly vulnerable to co-evolved natural enemies. This could explain the phenomenon of the
21
XII International Symposium on Biological Control of Weeds
ophialum (Morgan-Jones and Gams) Glenn, Bacon and
Hanlin. Tall fescue is a European species of high agronomic importance in North America despite the presence of the endophyte that produces highly toxic ergot
alkaloids (Cross, 2003). Endophyte-infected plants are
more vigorous, drought-tolerant and resistant to herbivores than endophyte-free ones, and one cultivar in
particular (Kentucky 31), with enhanced endophyte
activity, has now become a major invader of natural
communities where it impacts directly on the native
flora and fauna with long-term effects on successional
dynamics and food webs (Clay and Holah, 1999). It has
been argued that this is evidence of ecosystem vulnerability to human-induced invasion by an inbred, highly
competitive exotic species (Saikkonen, 2000; Saikkonen et al., 2006) rather than a natural model. Whatever
the interpretation, indirectly it lends support to the
E-ERH, demonstrating the ecological importance of coevolved, mutualistic endophytes and the invasive threat
from such associations in the absence of co-evolved
natural enemies. Should classical biological control
ever be considered as a management strategy for this
invasive grass, the result would be an arms race between the endophyte and any introduced (co-evolved)
natural enemies. This example also begs the question:
do similar endophyte associations also occur in the invasive African grasses currently threatening the longterm stability not only of the Amazon region but also of
global weather patterns (Mack et al., 2000)?
silver bullet in classical biological control, whereby
the introduction of a single biological control agent
can successfully and often unexpectedly, bring about
the complete control of a rampant, invasive alien weed.
Other alien plants, especially grasses with vertically
transmitted endophytes, may arrive with their mutualistic endophytes, which, in the absence of co-evolved
natural enemies, would give them a double advantage
over local competitors. Such endophyte-enriched,
alien-invasive weeds and those forming mutualistic associations with indigenous endophytes that afford protection from pests and diseases could help to explain
why some classical biological control introductions fail
to live up to expectations or that have only limited impact on the target weed.
The endophyte-enemy release hypothesis (E-ERH)
could help to resolve the on-going debate on the validity of the ERH (Wolfe, 2002; Mitchell and Power, 2003;
Colautti et al., 2004; Parker et al., 2006), as well as clarify inconsistencies in both the new encounter and the
evolution of increased competitive ability hypotheses
(Hokkanen and Pimentel, 1984; Blossey and Notzold,
1995). It also has resonance with the recently proposed
resource-ERH (Blumenthal, 2006), with a possible parallel situation in animal invasions, if protective endophytes can be compared to or are analogous with animal
immune defence systems (Lee and Klasing, 2004).
Evidence for the hypothesis
Monocot hosts
Dicot hosts
Evidence for specialized or co-evolved mutualistic
associations is unequivocal in the grassNeotyphodium systems (Schardl and Phillips, 1997; Schardl and
Moon, 2003). There is no clearer demonstration of
the ecological and practical importance of mutualistic
endophytes than the Poa annuaNeotyphodium association in northern United States, where endophtyeenriched seed is now routinely supplied to the turf-grass
industry (Bouton and Hopkins, 2003). The serendipitous discovery of the co-evolved endophyte in seed of
P. annua L. imported from northern Europe (the centre
of diversity), as part of a breeding programme, led to
research that demonstrated that the fungus not only afforded protection against generalist herbivores and pathogens but also conferred drought tolerance (J.F. White,
Rutgers University, personal communication, 2002).
Tests showed, however, that this fungus is not infective
to Poa pratensis L. (Kentucky blue grass, actually a European species from the Mediterranean region), which
is in high demand as a turf grass in southern United
States. Ecological and economic logic dictate that surveys in southern Europe would pay dividends.
In another example, which at first sight, may appear
to contradict the E-ERH, involves Lolium (Festuca)
arundinaceum (Schreber) S.B. Darbyshire or tall fescue and its co-evolved endophyte, Neotyphodium coen-
In sharp contrast, it is much less likely that co-evolved
endophytes will be carried to new ecosystems with their
dicot hosts, given that these are horizontally transmitted and that most introductions (accidental or deliberate) are from seed. In effect, it would be a similar
situation to that of co-evolved natural enemies, where
there are few examples of them arriving together with
their weed hosts. Therefore, it would be expected that
most invasive alien dicots lack specialist or co-evolved
endophytes.
The degree of specificity of dicot endophytes, however, is not as clear-cut as for the grass endophytes
discussed earlier, and it is probable that these are facultative rather than obligate in that, unlike Neotyphodium, they can survive saprophytically (Wilson, 2000).
Surveys for endophytes of cocoa (Theobroma cacao
L.) and its relatives in their South American centres of
origin revealed that the stems and pods of healthy wild
trees have a rich and unique endophytic mycobiota
that becomes depauperate in plantation trees in exotic
situations (Evans et al., 2003; Crozier et al., 2006). In
vitro studies to test their biological control potential
further demonstrated that some of these novel endophytes, pertaining to the Clavicipitaceae and Hypocreaceae (Hypocreales), are highly antagonistic to the
co-evolved fungal pathogens of cocoa and, in addition,
22
The endophyte-enemy release hypothesis
produce secondary metabolites known to be involved
in plant defence mechanisms (Holmes et al., 2004;
Samuels et al., 2006).
From this work, there is an indication that specialized, perhaps co-evolved, endophytes dominate in native habitats, but these are replaced by generalists when
the host is moved to exotic ecosystems. Similar results
have been reported for other woody plant hosts (Wilson, 2000), and further support is coming from ongoing
surveys of the endophytes associated with Lantana camara L. in both natural and degraded habitats in Brazil,
where this plant is indigenous, as well as in its exotic
invasive range in Pakistan (author, unpublished results).
The endophytes isolated from L. camara in degraded
sites in Brazil showed similarities with those recorded
from Pakistan in that these belonged predominantly to
a few well-known generalist fungal genera (Glomerella/Colletotrichum, Phomopsis), whereas those from a
forest site population comprised an extremely rich mycobiota with many unusual genera being represented. It
is tempting to suggest that these are part of a specialized endophytic guild of fungi that form mutualistic associations with L. camara and that, like the co-evolved
natural enemies, they have been left behind as the plant
host has been moved around the world.
Preliminary studies on Japanese knotweed, Fallopia
japonica (Houtt.) Ronse Decr., are yielding similar results. This plant, in urban situations in the United Kingdom, is virtually free of endophytes, whereas in climax
habitats in Japan, it has a rich and diverse endophytic
mycobiota (H. Evans, unpublished results). So much so
that contaminating endophytes have hampered the culture and study of the fungal component of the plants
co-evolved natural enemies. One of these, belonging to
a monotypic asexual genus, which, unusually, also produces its sexual stage (a new discomycete genus) in culture, can be reinoculated into and readily reisolated from
healthy knotweed leaves. Whether this species and the
other endophytes from Japan are specific or co-evolved
mutualists remains to be proven. However, it is evident
that sophisticated recognition mechanisms are involved,
enabling the fungus to bypass the plants defences.
Y. Ono (Tomley and Evans, 2004). Unexpectedly, host
mortality has been exceptionally high (75%) because
of a lethal combination of rust- and drought-induced
stress, whereas pod set and seedling recruitment have
been almost nonexistent.
Such dramatic impacts and high mortality are atypical of obligate pathogens especially in natural ecosystems and was never observed in Madagascar, where
the rust constitutes part of a guild of natural enemies
keeping the rubber vine population in check but neither
eliminating flowering and fruiting nor killing seedlings
and mature plants.
Discussion
The E-ERH is just one amongst a plethora of hypotheses put forward to explain invasiveness by alien species, especially by plants. Impressively, Colautti et al.
(2004) list no less than eight nonexclusive theories for
invasion success. Others have been added since (MllerSchrer et al., 2004; Blumenthal, 2006). This has led
to confusion and controversy, not to say a heady mix
of acronyms.
Each hypothesis could in itself explain a part of
invasion ecology, or more likely, each invasive weed
needs to be dealt with on a case-by-case basis. Clearly,
the successful silver bullet classical biological control
projects against invasive weeds must have been driven
by the ERH. In these cases, the E-ERH may further
explain why the release of a single natural enemy can
have such a dramatic and profound impact. Conversely,
the presence of co-evolved mutualists, in the absence
of co-evolved natural enemies, offers an explanation
as to why certain grasses have become major invasive
species with the ability to alter plant communities and
reduce biodiversity (Clay and Holah, 1999; Mack et al.,
2000). The E-ERH has direct pragmatic implications
perhaps more so in plant disease rather than weedinvasion ecology. In fact, the germ of the E-ERH was
sown during a project to evaluate the biological control
potential of co-evolved endophytes in the centres of ori
gin of the two major diseases of T. cacao. The finding
of unique endophytes in wild cocoa with demonstrable
antagonism towards the cocoa pathogens, together
with their absence in cultivated cocoa, points to a role
for mutualistic endophytes in plant protection (Evans
et al., 2003; Holmes et al., 2004; Samuels et al., 2006).
In the future, there could be the intriguing possibility
that crop plants, like turf grasses in the United States,
will be marketed as endophyte-enriched, planting
material being inoculated with co-evolved mutualists
to protect not only against pests and diseases but also
against abiotic stresses, notably drought. Such inoculations with mycorrhizal fungi are now standard practice
in tree nurseries.
Perhaps it is fitting to touch on the subject of mycorrhizae because they may also play a role in invasion biology. Indeed, mutualistic ectomycorrhizal (EM) fungi
Classical biological control
The silver bullet examples in classical biological
control could be explained, at least in part, on the basis of the invasive alien weed having lost its protective
co-evolved endophytes and not acquiring indigenous
generalist mutualists to fill this role. Thus, the fitness
of any introduced co-evolved natural enemy would be
increased accordingly. An analysis of the successful
rubber vine project in Australia offers empirical supporting evidence. This Madagascan asclepiad, Cryptostegia grandiflora (Roxb. ex R. Br.) R. Br., which
covered more than 40,000 km2 of northern Queensland,
has now been stopped in its tracks after the release of
a co-evolved rust, Maravalia cryptostegiae (Cummins)
23
XII International Symposium on Biological Control of Weeds
offer an explanation as to why exotic pine species have
increased fitness and why, in some ecosystems, they
have become highly invasive (Richardson et al., 2000).
In contrast, non-specific VAM fungi (Glomales) occur
in all soils and, seemingly, would readily be acquired
by indigenous and non-indigenous plants alike (Read,
1999). They have even been considered to be arguably the most important group of all living organisms
(Brundrett, 2002). Significantly, however, there are
plant families that are predominantly non-mycorrhizal,
including Amaranthaceae, Brassicaceae, Chenopodiaceae, Commelinaceae, Cyperaceae, Polygonaceae and
Urticaceae. Many of these are pioneer colonizers of
marginal habitats or weedy, opportunistic invaders of
disturbed soil, that have expanded into more marginal
environments since mid-Mesozoic, a trend which appears to be accelerated by mans escalating agricultural
and industrial activities (Pirozynski, 1981). Could it be
that they no longer needed VAM fungi and the tradeoffs this entailed, relying instead on mutualistic endophytes for competitive advantages? Indeed, the main
feature of the roots of these plant families is the capacity to actively exclude VAM fungi through the release
of anti-fungal metabolites (Brundrett, 2002). The exclusion of VAM fungi would conserve energy, increase
fitness and, therefore, should be another factor to be
included in the long list of why some plants become
invasive.
Tantalizingly, Richardson et al. (2000) briefly reflect
on fungal endophytes as possible promoters of plant invasions, concluding that: the specificity and the nature
of such associations are poorly known as is their role in
invasion. Here, it is proposed that their role, in the case
of specialized or co-evolved mutualistic endophytes, is
twofold: their presence increasing plant fitness in the
absence of co-evolved natural enemies, especially in
grass hosts with vertically transmitted endophytes;
their absence coupled with release from co-evolved
natural enemies, contributing to increased plant fitness,
especially in dicot hosts with horizontally transmitted
endophytes, but leaving them highly vulnerable to classical biological control agents.
Blackwell, M. (2000) Terrestrial lifefungal from the start?
Science 289, 18841885.
Blossey, B. and Notzold, R. (1995) Evolution of increased
competitive ability in invasive non-indigenous plants: a
hypothesis. Journal of Ecology 83, 887889.
Blumenthal, D.M. (2006) Interactions between resource
availability and enemy release in plant invasions. Ecology
Letters 9, 887895.
Bouton, J.H. and Hopkins, A.A. (2003) Commercial applications of endophytic fungi. In: White, J.F., Bacon, C.W.,
Hywel-Jones, N.L. and Spatafora, J.W. (eds) Clavicipitalean Fungi. Marcel Dekker, New York, pp. 495516.
Brundrett, M.C. (2002) Coevolution of roots and mycorrhizas
of land plants. New Phytologist 154, 275304.
Christensen, M.J. (1996) Antifungal activity in grasses infected with Acremonium and Epichloe endophytes. Australasian Plant Pathology 25, 186191.
Clay, K. (1997) Fungal endophytes, herbivores and the
structure of grassland communities. In: Gange, A.C. and
Brown, V.K. (eds) Multitrophic Interactions in Terrestrial
Systems. Blackwell, Oxford, UK, pp. 151169.
Clay, K. (2004) Fungi and the food of the gods. Nature 427,
401402.
Clay, K. and Holah, J. (1999) Fungal endophyte symbiosis
and plant diversity in successional fields. Science 285,
17421744.
Clay, K. and Schardl, C. (2002) Evolutionary origins and
ecological consequences of endophytic symbionts with
grasses. American Naturalist 160, S99S127.
Colautti, R.I., Ricciardi, A., Grigorovich, I.A. and MacIsaac,
H.J. (2004) Is invasion success explained by the enemy
release hypothesis? Ecology Letters 7, 721733.
Cross, D.L. (2003) Ergot alkaloid toxicity. In: White, J.F.,
Bacon, C.W., Hywel-Jones, N.L. and Spatafora, J.W.
(eds) Clavicipitalean Fungi. Marcel Dekker, New York,
pp. 475494.
Crozier, J., Thomas, S.E., Aime, M.C., Evans, H.C. and Holmes, K.A. (2006) Molecular characterisation of fungal endophytic morphospecies isolated from stems and pods of
Theobroma cacao. Plant Pathology 55, 783791.
Douds, D.D., Johnson, C.R. and Koch, K.E. (1988) Carbon
cost of the fungal symbiont relative to the net leaf P accumulation in a split-root VA mycorrhizal symbiosis. Plant
Physiology 86, 491496.
Evans, H.C., Holmes, K.A. and Thomas, S.E. (2003) Endophytes and mycoparasites associated with an indigenous
forest tree, Theobroma gileri, in Ecuador and a preliminary assessment of their potential as biocontrol agents of
cocoa disease. Mycological Progress 2, 149160.
Hokkanen, H. and Pimentel, D. (1984) New approach for selecting biological control agents. Canadian Entomologist
166, 11091121.
Holmes, K.A., Schroers, H., Thomas, S.E., Evans, H.C. and
Samuels, G.J. (2004) Taxonomy and biocontrol potential
of a new species of Trichoderma from the Amazon basin
of South America. Mycological Progress 3, 199210.
Keane, R.M. and Crawley, M.J. (2002) Exotic plant invasions
and the enemy release hypothesis. Trends in Ecology and
Evolution 17, 164170.
Kirk, P.M., Cannon, P.F., David, J.C. and Stalpers, J.A.
(2001) Dictionary of the Fungi, 9th edn. CAB International, Wallingford, UK.
Latch, G.C.M. (1993) Physiological interactions of endo-
Acknowledgements
This paper contains data, both published and unpub
lished, from studies funded by the Environment Agency
(United Kingdom), Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (Brazil) and USDA
ARS (Beltsville, MD).
References
Arnold, A.E., Mejia, L.C., Kyllo, D., Rojas, E.I., Maynard,
Z., Robbins, N. and Herre, E.A. (2003) Fungal endophytes limit pathogen damage in a tropical tree. Proceedings of the National Academy of Sciences of the USA 100,
1564915654.
24
The endophyte-enemy release hypothesis
phytic fungi and their hosts. Biotic stress tolerance imparted to grasses by endophytes. Agriculture, Ecosystems
and Environment 44, 143156.
Lee, K.A. and Klasing, K.C. (2004) A role for immunology
in invasion biology. Trends in Ecology and Evolution 19,
523529.
Lewis, D.H. (1985) Symbiosis and mutualism: crisp concepts
and soggy semantics. In Boucher, D.H. (ed.) The Biology
of Mutualism. Croom-Helm, London, UK, pp. 2939.
Mack, R.N., Simberloff, D., Lonsdale, W.M., Evans, H.,
Clout, M. and Bazzaz, F.A. (2000) Biotic invasions:
causes, epidemiology, global consequences and control.
Ecological Applications 10, 689710.
Marquez, L.M., Redman, R.S., Rodriguez, R.J. and Rossinck,
M.J. (2007) A virus in a fungus in a plant: three-way
symbiosis required for thermal tolerance. Science 315,
513515.
Mitchell, C.E. and Power, A.G. (2003) Release of invasive
plants from fungal and viral pathogens. Nature 421, 625
627.
Mller-Schrer, H., Schaffner, U. and Steinger, T. (2004)
Evolution in invasive plants: implications for biological
control. Trends in Ecology and Evolution 19, 417
422.
Narisawa, K., Ohki, K.T. and Hashiba, T. (2000) Suppression
of clubroot and Verticillium yellows in Chinese cabbage
in the field by the root endophytic fungus, Heteroconium
chaetospira. Plant Pathology 49, 141146.
Parker, J.D., Burkepile, D.E. and Hay, M.E. (2006) Opposing
effects of native and exotic herbivores on plant invasions.
Science 311, 14591461.
Pirozynski, K.A. (1981) Interactions between fungi and
plants through the ages. Canadian Journal of Botany 59,
18241827.
Pirozynski, K.A. and Malloch, D.W. (1975) The origin of
land plants: a matter of mycotrophism. Biosystems 6,
153164.
Read, D.J. (1999) Mycorrhizathe state of the art. In: Varma,
A. and Hock, B. (eds) Mycorrhiza. Springer-Verlag, Berlin,
Germany, pp. 334.
Redman, R.S., Dunigan, D.D. and Rodriguez, R.J. (2001)
Fungal symbiosis from mutualism to parasitism: who
controls the outcome, host or invader? New Phytologist
151, 705716.
Richardson, D.M., Alsopp, N., DAntonio, E.M., Mitton, S.J.
and Rejmanek, M. (2000) Plant invasionsthe role of
mutualisms. Biological Reviews 75, 6593.
Rodriguez, R. and Redman, R. (2005) Balancing the generation and elimination of reactive oxygen species. Proceedings of the National Academy of Science of the USA 102,
31753176.
Rodriguez, R.J., Redman, R.S. and Henson, J.M. (2004)
The role of fungal symbioses in the adaptation of plants
to high stress environments. Mitigation and Adaptation
Strategies for Global Change 9, 261272.
Rudgers, J.A., Koskow, J.M. and Clay, K. (2004) Endophytic
fungi alter relationships between diversity and ecosystem
properties. Ecology Letters 7, 4251.
Saikkonen, K. (2000) Kentucky 31, far from home. Science
287, 1887.
Saikkonen, K., Faeth, S.H., Helander, M. and Sullivan, T.J.
(1998) Fungal endophytes: a continuum of interactions
with host plants. Annual Review of Ecology and Systematics 29, 319343.
Saikkonen, K., Lehtonen, P., Helander, M., Koricheva, J. and
Faeth, S.H. (2006) Model systems in ecology: dissecting
the endophytegrass literature. Trends in Plant Science
11, 428433.
Samuels, G.J., Suarez, C., Solis, K., Holmes, K.A., Thomas,
S.E., Ismaiel, A. and Evans, H.C. (2006) Trichoderma
theobromicola and T. paucisporum: two new species isolated from cacao in South America. Mycological Research
110, 381392.
Schardl, C.L. and Moon, C.D. (2003) Processes of species
evolution in Epichloe/Neotyphodium endophytes of
grasses. In: White, J.F., Bacon, C.W., Hywel-Jones, N.L.
and Spatafora, J.W. (eds) Clavicipitalean Fungi. Marcel
Dekker, New York, pp. 273327.
Schardl, C.L. and Phillips, T.D. (1997) Protective grass endophytes. Plant Disease 81, 430438.
Schulz, B. and Boyle, C. (2005) The endophytic continuum.
Mycological Research 109, 661686.
Simon, L., Bousquet, J., Levesque, R.C. and Lalonde, M.
(1993) Origin and diversification of endomycorrhizal
fungi and coincidence with vascular land plants. Nature
363, 6769.
Stone, J.K., Polishook, J.D. and White, J.F. (2000) Endophytic fungi. In Mueller, G.M., Bills, G.F. and Foster,
M.S. (eds) Biodiversity of Fungi. Elsevier, Amsterdam,
The Netherlands, pp. 241270.
Tomley, A.J. and Evans, H.C. (2004) Establishment of and preliminary impact studies on, the rust, Maravalia cryptostegiae, of the invasive alien weed, Cryptostegia grandiflora,
in Queensland, Australia. Plant Pathology 53, 475484.
Varma, A., Verma, S., Sudha, A., Sayah, N., Butehorn, B. and
Franken, P. (1999) Piriformospora indica, a cultivable
plant-growth-promoting root endophyte. Applied and Environmental Microbiology 65, 27412744.
Wennstrom, A. (1994) Endophytesthe misuse of an old
term. Oikos 71, 535536.
Wilson, D. (1993) Fungal endophytes: out of sight but should
not be out of mind. Oikos 68, 379384.
Wilson, D. (1995) Endophytesthe evolution of a term and
clarification of its use and definition. Oikos 73, 274276.
Wilson, D. (2000) Ecology of woody plant endophytes. In
Bacon, C.W. and White, J.F. (eds) Microbial Endophytes.
Marcel Dekker, New York, pp. 389420.
Wolfe, L.M. (2002) Why alien invaders succeed: support for
the escape-from-enemy hypothesis. American Naturalist
160, 705711.
25
Multiple-species introductions of
biological control agents against weeds:
look before you leap
F.A.C. Impson,1,2 V.C. Moran,1 C. Kleinjan,1
J.H. Hoffmann1 and J.A. Moore2
Summary
Biological control practitioners have frequently debated the issues behind single vs multiple species
introductions against target weeds. In the case of weed biological control, conventional wisdom is
that multiple species should be used on the assumption that several species are more likely to have a
greater controlling impact than a single species alone. This debate is rehearsed with reference to the
biological control of four species of Australian acacias in South Africa: long-leaved wattle (Acacia
longifolia (Andr.) Willd.), golden wattle (Acacia pycnantha Benth.), Port Jackson willow (Acacia
saligna (Labill.) H. Wendl.) and rooikrans (Acacia cyclops A. Cunn. ex G. Don), where the impacts
of both gall-forming and seed-reducing agents were intended to be additive and possibly synergistic.
Evaluation and observations of these specific cases show that multiple-species introductions can be
beneficial, but in at least one case (A. cyclops), the wisdom of these releases is questionable and potentially even detrimental. This suggests the need for extreme caution when planning multiple-species
introductions against a target weed species.
Keywords: multiple species, Acacia, biological control.
Introduction
For many years, biological control practitioners have
discussed and debated the merits of releasing multiple as opposed to single species of biological control
agents in weed control programmes. The focus of such
discussion has been multifaceted, either in terms of
the effectiveness of the actual control (Myers, 1985;
Myers et al., 1989; Story et al., 1991; Mller-Schrer
and Schroeder, 1993; Hoffmann and Moran, 1998; Anderson et al., 2000), competitive interactions between
agents (Zwlfer, 1973; Ehler and Hall, 1982; Denno
et al., 1995; Woodburn, 1996; Briese, 1997; McEvoy
and Coombs, 2000), the best timing or sequence in
which to introduce agents (Briese, 1991; Syrett et al.,
1996), or in terms of risk, safety and direct and indirect
1
Department of Zoology, University of Cape Town, Rondebosch 7701,
South Africa.
2
Plant Protection Research Institute, Private Bag X5017, Stellenbosch
7599, South Africa.
Corresponding author: F.A.C. Impson, Plant Protection Research Institute, Private Bag X5017, Stellenbosch 7599, South Africa <impsonf@
arc.agric.za>.
CAB International 2008
26
non-target effects (Myers, 1985; Simberloff and Stiling, 1996; Callaway et al., 1999; Denoth et al., 2002;
Pearson and Callaway, 2005).
For most weed biological control projects, the highest levels of success have been achieved using multiple agents, either because there has been a cumulative
or synergistic effect of all agents working together
(e.g. Hoffmann and Moran, 1998) or because as agent
numbers are increased, there is likely to be a greater
probability that the most suitable species will be released, often with a single agent being responsible for
the success (Myers, 1985). Alternatively, the introduction of more agents may ultimately provide a higher
probability of biological control over wider geographical ranges, due to different agent species performing
better under different conditions (DeBach, 1964; Baars
and Heystek, 2003; Day et al., 2003). In many cases,
new and additional agents are released prematurely, either because existing agents have not been adequately
evaluated or because agents have not been provided an
opportunity to achieve their full potential (McFadyen,
1998; McEvoy and Coombs, 2000). Unfortunately, predicting the effectiveness of, and possible interactions
between, potential biological control agents remains an
Multiple-species introductions of biological control agents against weeds: look before you leap
ongoing and daunting challenge (Cullen, 1995; Zalucki
and van Klinken, 2006).
The biological control programmes against four
invasive Australian Acacia species in South Africa
are discussed with respect to these issues. They demonstrate that although multiple-agent releases are usually beneficial, there are times when this may not be
the case and releases of more than one agent should be
planned with caution.
Biological control of Acacia species in
South Africa
During the last 30 years, biological control has been
implemented against nine of the most invasive Australian Acacia species in South Africa (Dennill et al.,
1999). Collectively, these programmes have largely
been governed by conflicts of interest over desires to
control the plants whilst continuing to exploit them
commercially for production of tannin, for timber
and pulp, for fire wood and for dune binding. Consequently, the choice of biological control agents has
been restricted, for the most part, to agents that limit
the reproductive output of their hosts, thereby reducing
invasiveness but not the useful attributes of the plants.
Four of these acacias have been subject to control by
two agent species released sequentially (Table 1), and
they are the subject of discussion here.
Acacia longifolia (Andr.) Willd. (long-leaved wattle):
The gall-forming wasp, Trichilogaster acaciaelon
gifoliae Froggatt (Hymenoptera: Pteromalidae), was
released on A. longifolia in South Africa during 1982
(Dennill and Donnelly, 1991). The wasps dispersed
readily and reduced seed production on A. longifolia
by more than 95%, even causing some suppression of
vegetative growth of the plants (Dennill, 1988). However, there were two situations where T. acaciaelongi
foliae was not fully effective: (a) in the hot, arid, inland
areas and in the elevated, moist, mist-belt regions of
the country where climatic conditions curb population
expansion of the wasps (Dennill and Gordon, 1990)
and (b) A. longifolia plants growing close to rivers do
not suffer water stress and still produce substantial seed
loads despite high levels of galling by the wasp (Dennill et al., 1999).
Although the impact of T. acaciaelongifoliae was
being studied, a second agent, a seed-feeding weevil,
Melanterius ventralis Lea (Coleoptera: Curculionidae),
had been proposed for control of A. longifolia and was
being tested in quarantine. By 1985, the need for an
additional agent was deemed to be necessary, and the first
releases of M. ventralis were made. The seed-feeding
weevils established readily at release sites and have
subsequently played an important supplementary role
in the suppression of seed production by A. longifolia
(Dennill et al., 1999; Donnelly and Hoffmann, 2004).
Acacia pycnantha Benth. (golden wattle): Following
the success of T. acaciaelongifoliae on A. longifolia,
27
a related species of gall-forming wasp, Trichilogaster
signiventris (Girault) (Hymenoptera: Pteromalidae),
was released against A. pycnantha during 1987. After
a slow start, when it was believed that the wrong strain
of T. signiventris may have been imported (Dennill and
Gordon, 1991), and additional releases in 1992, levels
of galling increased dramatically, and the insects became abundant throughout the range of A. pycnantha
by 1998. Besides substantial reductions in seed production due to the wasps, in some cases, extensive galling
caused collapse of branches and toppling of whole trees
(Dennill et al., 1999; Hoffmann et al., 2002). Although
initial indications were that no additional agents would
be required to further reduce seed production, monitoring of pod and gall loads (in 2004 and 2005) demonstrated that many seed pods were still being produced
despite the damage caused by T. signiventris.
The successful combination of the gall former and
a seed feeder in the A. longifolia programme paved the
way for a similar approach with A. pycnantha, and in
2005, the seed-feeding weevil, Melanterius maculatus
Lea (Coleoptera: Curculionidae), was released. Although it is still too early to draw conclusions regarding the combined impact of the two agents, indications
are that both agents will complement each other in reducing seed loads of A. pycnantha plants as is the case
on A. longifolia.
Acacia saligna (Labill.) H. Wendl. (Port Jackson willow): Biological control of A. saligna had been recommended as a priority from the outset of the programme
against the Australian acacias (Neser and Annecke,
1973). The gall-forming rust fungus, Uromycladium
tepperianum (Sacc.) McAlp. (Urediniales: Raveneliaceae), was selected as being a suitably damaging
agent in that it could reduce reproductive output and
also weaken the plants and ultimately cause their death
(van den Berg, 1977). After its release in 1987, U. tep
perianum rapidly dispersed throughout the range of A.
saligna. Long-term evaluation studies demonstrated
that the rust was an extremely effective agent, reducing population densities of adult trees by up to 85%
(Wood and Morris, 2007). However, as in the case of
A. longifolia and A. pycnantha, A. saligna was still able
to produce large seed loads before succumbing to the
effects of high levels of galling.
Again, the need was recognized for a second agent
to target the remaining seeds, and another seed-feeding
weevil, Melanterius compactus Lea (Coleoptera: Curculionidae), was released against A. saligna in 2001.
Although the introduction of M. compactus is relatively
recent, preliminary monitoring indicates that, like its
counterpart on A. longifolia, the weevils are playing an
important supplementary role in curbing the production of viable seeds on A. saligna.
Acacia cyclops A. Cunn. ex G. Don (rooikrans): A.
cyclops was the last of the four species under discussion to be subjected to biological control. In the early
1990s, there was a strong focus on the Melanterius
XII International Symposium on Biological Control of Weeds
group of weevils, which were readily available and
easy to collect and had been shown to be sufficiently
host-specific and damaging to warrant consideration
(Impson and Moran, 2004). In 1991, the first release
of Melanterius servulus (Pascoe) (Coleoptera: Curculionidae) was carried out, followed in 1993 by more
widespread releases. Although the weevils established
successfully, they were relatively slow to build up their
populations, and dispersal was also limited (Impson et
al., 2004; Impson, 2005). Despite this, levels of seed
damage increased with time at many of the release sites,
with up to 95% seed damage being recorded within 5
years of release at some of the sites. Manual redistribution has been used to compensate for slow rates of
natural dispersal.
In 2001, a proposal was made that a second agent,
a flower-galling midge, Dasineura dielsi Rbsaamen
(Diptera: Cecidomyiidae), should be released to supplement the activities of M. servulus. It was anticipated that
the midge would fulfill a complementary role and have
good dispersal abilities, which would thus compensate
for the problem of slow dispersal rates of the weevil.
At the time, some concerns were expressed regarding
possible competitive interactions between the two control agents (i.e. by galling the flowers, the midge would
indirectly remove the food source of the weevils), but
the matter of containing large invasions of A. cyclops
was considered a priority and additional restrictive
measures against this plant were strongly supported.
Following the establishment of D. dielsi, the midge
dispersed extremely rapidly (hundreds of kilometers
per year) throughout the range of A. cyclops (J. Moore,
personal communication, 2003), and with its multivoltine life cycle, populations of the midge exploded. It
initially appeared that the proverbial silver bullet had
been released, and A. cyclops trees had been all but
sterilized by the extremely high levels of galling. However, this situation did not persist, and midge populations have become less stable, resulting in considerable
variation in the amount of pod set between sites and
between years (F. Impson, C. Kleinjan and J. Moore,
unpublished results). This has obvious implications for
M. servulus because the weevils may no longer be able
to sustain their populations when faced with an unpredictable food source, and ultimately, the success of the
biological control programme against A. cyclops may
be compromised.
Discussion
In these four cases of biological control against imported
Australian acacias, there was a clear rationale, based
on available knowledge, which governed the pattern
and sequence of the releases of agents (Table 1), and in
each case, the release of two agents has been justified.
For each of A. longifolia, A. pycnantha and A.
saligna, a gall-forming agent was released before being
followed up by a seed-destroying weevil (Table 1). In
28
all of these programmes, the sequence of releases (i.e. a
gall former preceding a seed feeder) was largely determined by opportunistic and pragmatic considerations.
Agents that were readily available, obviously damaging to the host plant, abundant and easy to collect
and amenable to specificity testing enjoyed priority. In
the case of A. longifolia, the release of two species of
agents occurred within 3 years of each other, and it is
possible that if practical circumstances had been different the order of release could have been reversed.
The cases of A. pycnantha and A. saligna, respectively,
are different in the sense that considerable time elapsed
between the releases of the first and second agents. The
reason for this was a conscious decision to evaluate the
impact of the gall formers acting on their own, before
taking the decision to release a supplementary agent. In
both cases, events were to prove that although the gall
formers were highly effective, there were more than
sufficient seeds left in the system to maintain populations of the host plants at problematic levels. There was
a clear need for the seed-feeding weevils to reduce the
numbers of viable seeds.
The pattern for A. cyclops, however, is different in
that a seed-feeding weevil species was released first,
followed several years later by the release of a gall
midge. Again the sequence of release was determined
by pragmatic and opportunistic circumstances and was
influenced by strong demands for additional control
measures against A. cyclops, particularly in view of the
slow dispersal rates of M. servulus. The, gall midge,
D. dielsi, was not an obvious choice of agent, primarily because of doubts about the effectiveness of gall
midges as biological control agents (Goeden and Louda,
1976; McFadyen, 1985; Wehling and Piper, 1988; Carlson
and Mundal, 1990; Harris and Shorthouse, 1996) and
because from the outset there were some concerns
over a potential conflict with M. servulus. Eight years
elapsed before it was decided that a supplementary
agent was needed.
The A. cyclops programme differs from the others
in one other important respect. In the case of A. longi
folia, A. pycnantha and A. saligna, the gall-forming
agents are essentially univoltine, exert pressure on the
plants and substantially reduce seed production, but in
most circumstances, there are sufficient seeds remaining locally or in a wider area to sustain populations of
the seed-feeding weevils. In other words, the evidence
suggests that the effects of the agents are complementary. The gall-forming cecidomyiid, D. dielsi, on A.
cyclops, by contrast, goes through several generations
a year, most of which coincide with the peak flowering period of the plant (the females lay their eggs in
the flowers), which initially led to enormous gall loads
and the virtual or complete elimination of pods at sites.
Subsequently, levels of pod production have been extremely variable. Of concern is the possibility that the
fluctuations in pod set will destabilize populations of
M. servulus and render the beetles unable to exploit and
Multiple-species introductions of biological control agents against weeds: look before you leap
Table 1.
The four species of Australian acacias targeted for biological control in South Africa, using in each case sequential releases of two agent species, all imported from Australia. In certain cases (marked by asterisk), there were
previous releases, but they were unsuccessful.
Acacia species
(Mimosaceae)
Agent released
1. Trichilogaster
A. A. longifolia
(long-leaved wattle) acaciaelongifoliae
B. A. pycnantha
(golden wattle)
C. A. saligna (Port
Jackson willow)
D. A. cyclops
(rooikrans)
Date of Release interval
Mode of action
between agents
first
release
(years)
1982
3
Induces extensive
gall formation
2. Melanterius ventralis
1985
1. T. signiventris
1992*
2. M. maculatus
2005
1. Uromycladium
tepperianum
1987
2. M. compactus
2001
1. M. servulus
1993*
2. Dasineura dielsi
2001
13
14
destroy the surfeit of seeds that develop when D. dielsi
is less effective. At times when seeds are scarce, the
situation is further exacerbated by rodent and bird predation of the seeds. The adult weevils also feed widely
on the ripening seeds, leaving virtually no seeds that
are in a suitable condition for oviposition. Under these
conditions of extreme seed scarcity, the weevil populations are in danger of becoming extinct locally or even
over wide areas.
It is still too early to predict the long-term outcome
of this programme. Preliminary studies indicate that
the unstable conditions over the last few years have impacted on M. servulus populations at some monitoring
sites, and if this situation persists, it may prove to be
inimical to the biological control programme against A.
cyclops in the long term. Alternatively, midge populations may ultimately stabilize at levels where sufficient
pods are consistently available at sites. Under such
conditions, it is anticipated that M. servulus populations would build up again and that the actions of D.
dielsi and M. servulus could be additive, as it is for the
other Acacia species with two agents. In addition, the
objective of harnessing the high dispersal abilities of
D. dielsi would also have been realized.
Apart from the fact that the introduction of organisms contains inherent risk, the broader ecological consequences of introductions have received little attention
and remain poorly understood. Weed biological control
is only contemplated in situations where mechanical
and/or chemical control of invasive plants is impracti-
References
Dennill, 1988;
Dennill and
Donnelly, 1991
Destroys seed
Dennill and
Donnelly, 1991
Induces extensive
Dennill and Gordon,
gall formation
1991; *Dennill
et al., 1999
Destroys seed
F. Impson,
unpublished results
Induces fungal galls Morris, 1991
on reproductive and
vegetative tissue
Destroys seed
F. Impson,
unpublished results
Destroys seed
*Dennill et al., 1999;
Impson, 2005
Induces galling
Adair, 2004
of flowers
cal or prohibitively expensive. Predicting the outcome
of introductions remains problematic because, frequently, the interacting attributes of the agent, the target weed and the environment are extremely complex.
Furthermore, the introduction of each additional agent
introduces another tier of complexity, complicating the
ability to correctly predict outcomes.
In the case of the releases of a second agent onto A.
pycnantha and A. saligna, sufficient time had elapsed
between the introductions of the gall formers and the
subsequent decision to release seed feeders; the impacts of the gall formers were well understood, and
a clear need for an additional agent that would target
residual seed production was identified. In addition,
extensive knowledge of the attributes of Melanterius
spp. and their potential as biological control agents in
South Africa was available. With A. cyclops, sufficient
time had elapsed after the introduction of M. servulus
for adequate evaluation of its performance and the
recognition of its limitations. However, the ability to
predict the outcome for the midge and its possible interactions with M. servulus was limited. The situation
with A. cyclops in South Africa highlights the need for
extreme caution when contemplating multiple species
introductions and adds credence to the rule that biological control agents in any situation should only be
introduced where circumstances demand and where the
best predictions, as a result of experience, intuition or
modeling, suggest that these multiple species introductions will not worsen the situation.
29
XII International Symposium on Biological Control of Weeds
References
Anderson, G.L., Delfosse, E.S., Spencer, N.R., Prosser, C.W.
and Richard, R.D. (2000) Biological control of leafy
spurge: an emerging success story. In: Spencer, N.R.
(ed.) Proceedings of the X International Symposium on
Biological Control of Weeds. Montana State University,
Bozeman, MT, pp. 1525.
Baars, J.R. and Heystek, F. (2003) Geographical range and
impact of five agents established on Lantana camara in
South Africa. BioControl 48, 743759.
Briese, D.T. (1991) Current status of Agrilus hyperici (Coleoptera: Buprestidae) released in Australia in 1940 for the
control of St Johns wort: lessons for insect introductions.
Biocontrol Science and Technology 1, 207215.
Briese, D.T. (1997) Biological control of St. Johns wort: past,
present and future. Plant Protection Quarterly 12, 7380.
Callaway, R.M., DeLuca, T.H. and Belliveau, W.M. (1999)
Biological control herbivores may increase competitive
ability of the noxious weed Centaurea maculosa. Ecology
80, 11961201.
Carlson, R.B. and Mundal, D. (1990) Introduction of insects
for the biological control of leafy spurge in North Dakota.
North Dakota Farm Research 47, 78.
Cullen, J.M. (1995) Predicting effectiveness: fact and fantasy. In: Delfosse, E.S. and Scott, R.R. (eds) Proceedings
of the VIII International Symposium on Biological Con
trol of Weeds. DSIR/CSIRO, Melbourne, Australia, pp.
103109.
Day, M.D., Wiley C.J., Playford, J. and Zaluki, M.P. (2003)
Lantana: current management status and future prospects.
ACIAR Monograph Series 102, Canberra, Austalia.
DeBach, P. (1964) Biological control of insect pests and
weeds. Chapman & Hall, London, UK.
Dennill, G.B. (1988) Why a gall former can be a good
biocontrol agentthe gall wasp Trichilogaster acaciae
longifoliae and the weed Acacia longifolia. Ecological
Entomology 13, 19.
Dennill, G.B. and Donnelly, D. (1991) Biological control of
Acacia longifolia and related weed species (Fabaceae) in
South Africa. Agriculture, Ecosystems and Environment
37, 115135.
Dennill, G.B. and Gordon, A.J. (1990) Climate-related differences in the efficacy of the Australian gall wasp
(Hymenoptera: Pteromalidae) released for the control of
Acacia longifolia in South Africa. Environmental Ento
mology 19, 130136.
Dennill, G.B. and Gordon, A.J. (1991) Trichilogaster sp. (Hymenoptera: Pteromalidae), a potential biocontrol agent for
the weed Acacia pycnantha (Fabaceae). Entomophaga 36,
295301.
Dennill, G.B., Donnelly, D., Stewart, K. and Impson, F.A.C.
(1999) Insect agents used for the biological control of
Australian Acacia species and Paraserianthes lophantha
(Willd.) Nielsen (Fabaceae) in South Africa. African En
tomology Memoir 1, 4554.
Denno, R.F., McClure, M.S. and Ott, J.R. (1995) Interspecific interactions in phytophagous insects: competition reexamined and resurrected. Annual Review of Entomology
40, 297331.
Denoth, M., Frid, L. and Myers, J.H. (2002) Multiple agents
in biological control: improving the odds? Biological
Control 24, 2030.
30
Donnelly, D. and Hoffmann, J.H. (2004) Utilization of an unpredictable food source by Melanterius ventralis, a seedfeeding biological control agent of Acacia longifolia in
South Africa. BioControl 49, 225235.
Ehler, L.E. and Hall, R.W. (1982) Evidence for competitive
exclusion of introduced natural enemies in biological control. Environmental Entomology 11, 14.
Goeden, R.D. and Louda, S.M. (1976) Biotic interference
with insects imported for weed control. Annual Review of
Entomology 21, 325342.
Harris, P. and Shorthouse, J.D. (1996) Effectiveness of gall
inducers in weed biological control. Canadian Entomolo
gist 128, 10211055.
Hoffmann, J.H. and Moran, V.C. (1998) The population dynamics of an introduced tree, Sesbania punicea, in South
Africa, in response to long-term damage caused by different combinations of three species of biological control
agents. Oecologia 114, 343348.
Hoffmann, J.H., Impson, F.A.C., Moran, V.C. and Donnelly,
D. (2002) Trichilogaster gall wasps (Pteromalidae) and
biological control of invasive golden wattle trees (Aca
cia pycnantha) in South Africa. Biological Control 25,
6473.
Impson, F. (2005) Biological control of Acacia cyclops in
South Africa: the role of an introduced seed-feeding weevil, Melanterius servulus (Coleoptera: Curculionidae) together with indigenous seed-sucking bugs and birds. MSc
thesis. University of Cape Town, South Africa.
Impson, F.A.C. and Moran, V.C. (2004) Thirty years of exploration for and selection of a succession of Melanterius
weevil species for biological control of invasive Australian acacias in South Africa: should we have done anything differently? In: Cullen, J.M., Briese, D.T., Kriticos,
D.J., Lonsdale, W.M., Morin, L. and Scott, J.K. (eds) Pro
ceedings of the XI International Symposium on Biological
Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 127134.
Impson, F.A.C., Moran, V.C. and Hoffmann, J.H. (2004) Biological control of an alien tree, Acacia cyclops, in South
Africa: impact and dispersal of a seed-feeding weevil,
Melanterius servulus. Biological Control 29, 375381.
McEvoy, P.B. and Coombs, E.M. (2000) Why things bite
back: unintended consequences of biological weed control. In: Follett, P.A. and Duan, J.J. (eds) Nontarget effects
of biological control. Kluwer Academic Publishers, Boston, MA, pp. 167194.
McFadyen, P.J. (1985) Introduction of the gall fly, Rhopalo
myia californica from the USA into Australia for the control of the weed Baccharis halimifolia. In: Delfosse, E.S.
(ed.) Proceedings of the VI International Symposium on
Biological Control of Weeds. Agriculture Canada, Vancouver, Canada, pp. 779796.
McFadyen, R.E.C. (1998) Biological control of weeds. An
nual Review of Entomology 43, 369393.
Morris, M.J. (1991) The use of plant pathogens for biological weed control in South Africa. Agriculture, Ecosystems
and Environment 37, 239255.
Mller-Schrer, H. and Schroeder, D. (1993) The biological
control of Centaurea spp. in North America: do insects
solve the problem? Pesticide Science 37, 343353.
Myers, J.H. (1985) How many insect species are necessary
for successful biocontrol of weeds? In: Delfosse, E.S.
(ed.) Proceedings of the VI International Symposium on
Multiple-species introductions of biological control agents against weeds: look before you leap
Biological Control of Weeds. Agriculture Canada, Vancouver, Canada, pp. 7782.
Myers, J.H., Higgins, C. and Kovacs, E. (1989) How many
insect species are necessary for the biological control of
weeds? Environmental Entomology 18, 541547.
Neser, S. and Annecke, D.P. (1973) Biological control of weeds
in South Africa. African Entomology Memoir 28, 27.
Pearson, D.E. and Callaway, R.M. (2005) Indirect nontarget effects of host-specific biological control agents: implications for biological control. Biological Control 35,
288298.
Simberloff, D. and Stiling, P. (1996) How risky is biological
control? Ecology 77, 19651974.
Story, J.M.K., Boggs, K.W., Good, W.R., Harris, P. and
Nowierski, R.M. (1991) Metzneria paucipunctella Zeller (Lepidoptera: Gelechiidae), a moth introduced against
spotted knapweed: its feeding strategy and impact on two
introduced Urophora spp. (Diptera: Tephritidae). Cana
dian Entomologist 123, 10011007.
Syrett, P., Fowler, S.V. and Emberson, R.M. (1996) Are chrys
omelid beetles effective agents for biological control of
weeds? In: Moran, V.C. and Hoffmann, J.H. (eds) Pro
ceedings of the IX International Symposium on Biological
Control of Weeds. University of Cape Town, Stellenbosch,
South Africa, pp. 399407.
Van den Berg, M.A. (1977) Natural enemies of certain acacias in Australia. In: Proceedings of the Second National
Weeds Conference of South Africa, Stellenbosch, South
Africa. A.A. Balkema, Cape Town, South Africa, pp.
7582.
Wehling, W.F. and Piper, G.L. (1988) Efficacy diminution of
the rush skeletonweed gall midge, Cystiphora schmidti
(Diptera: Cecidomyiidae), by an indigenous parasitoid.
Pan-Pacific Entomologist 64, 8385.
Wood, A. and Morris, M.J. (2007) Impact of the gall-forming
rust Uromycladium tepperianum on the invasive tree
Acacia saligna in South Africa: 15 years of monitoring.
Biological Control 41, 6877.
Woodburn, T.L. (1996) Interspecific competition between
Rhinocyllus conicus and Urophora solstitialis, two biocontrol agents released in Australia against Carduus nutans.
In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings
of the IX International Symposium on Biological Control
of Weeds. University of Cape Town, Stellenbosch, South
Africa, pp. 409415.
Zalucki, M.P. and van Klinken, R.D. (2006) Predicting population dynamics of weed biological control agents: science or gazing into crystal balls? Australian Journal of
Entomology 45, 331344.
Zwlfer, H. (1973) Competition and coexistence in phytophagous insects attacking the heads of Carduus nutans L. In:
Dunn, P.H. (ed.) Proceedings of the II International Sym
posium on the Biological Control of Weeds. Miscellaneous
Publication 6. Commonwealth Institute of Biological
Control, Commonwealth Agricultural Bureaux, Farnham
Royal, Slough, UK, pp. 7477.
31
Clipping the butterfly bushs wings:
defoliation studies to assess the likely
impact of a folivorous weevil
D.J. Kriticos,1 M.S. Watt,2 D. Whitehead,3
S.F. Gous,4 K.J. Potter5 and B. Richardson4
Summary
Predicting agent success is a topic that has attracted much attention from the biological control community. Although the likely success of agents establishing in a new environment remains elusively
unpredictable, we can often gain an impression of the likely nature of the agents impact in different
environments should it establish in reasonable numbers. The butterfly bush, or buddleia (Buddleja
davidii Franch.), is a major weed problem in many regions with temperate or Mediterranean climates
and has been identified as the highest priority for biological control in Europe. In New Zealand, it
has invaded disturbed sites such as plantation forest coups, roadsides, earth slips and gravel river
beds. To combat buddleia in New Zealand, a biological control programme was commenced around
1990. Whilst host-specificity testing was being completed on Cleopus japonicus Wingelmller, a
leaf-feeding weevil, defoliation experiments were undertaken to assess its likely impact on the growth
and survival of its prime host, buddleia. Seasonal defoliation studies revealed that in the absence of
plant competition, buddleia was quite resilient and able to recover rapidly from severe defoliation.
Experiments with plant competition, leaf consumption rates and insect developments rates were used
to develop a model to explore the likely impact of C. japonicus.
Keywords: Buddleja davidii, Cleopus japonicus, compensatory growth, growth
modelling, simulated herbivory.
Introduction
species can be predicted in the country of release (McFadyen, 1998).
Although numerous examples of complete or partial
control of weed species by biological control agents
have been reported, there are also many instances
where control of the target weed has been negligible
(McEvoy et al., 1991; Ooi, 1992; Hoffmann, 1995;
McFadyen, 1998; Julien and Griffiths, 1999). Whilst
predicting the success of individual agents in establishing in a new environment remains elusive, we may be
able to at least gain an impression of the likely nature
of the agents impact in different environments should
it establish in reasonable density. For agents that defoliate plants, it may be appropriate to undertake studies
to gauge the impact of different defoliation regimes on
various aspects of the plants natural history.
A broad understanding of how attack by a biological control agent influences a weeds growth and life
history traits is helpful for prioritizing guilds of insects
or pathogens for inclusion in biological control programmes and quantifying the level of control that can
be expected from individual agents (Kriticos, 2003;
Predicting the likely success of a biological control
agent is a topic that has attracted much attention from
the biological control community. The prime challenge
for biological control practitioners after ensuring agent
safety is to select agents that have a high probability of
establishing and, if established, will have a significant
negative impact on the target weed. The success of this
endeavour depends partly on how well the effects of
the agent on the growth and survival of the target weed
Ensis Forest Biosecurity and Protection, PO Box E4008, Kingston,
ACT 2604, Australia.
2
Scion Forest Biosecurity and Protection, PO 29237, Christchurch, New
Zealand.
3
Landcare Research, PO Box 40, Lincoln 7640, New Zealand.
4
Scion Forest Biosecurity and Protection, Private Bag 3020, Rotorua,
New Zealand.
5
CSIRO Forest Biosecurity and Protection, Private Bag 12, Hobart, TAS
7001, Australia.
Corresponding author: D.J. Kriticos <[email protected]>.
CAB International 2008
32
Clipping the butterfly bushs wings: defoliation studies to assess the likely impact of a folivorous weevil
merer and Farquhar, 1984; Trumble et al., 1993). Leaf
tissue removal has also been shown to either increase
(Mabry and Wayne, 1997) or reduce (Dirzo, 1984; Mabry and Wayne, 1997) longevity of remaining leaves.
The butterfly bush, or buddleia (Buddleja davidii
Franch., Buddlejaceae), is a major weed problem in
many regions with temperate or Mediterranean climates
(Fig. 1a, b), and it has been identified as the number
one priority for biological control in Europe (Sheppard
et al., 2006). Cleopus japonicus Wingelmller (Coleoptera: Curculionidae) is a leaf-feeding weevil that has
been identified as a biological control agent for buddleia. After extensive host-specificity testing, C. japonicus was released in New Zealand in late 2006. Initial
results indicate that it appears to be establishing well
in the field, although the field populations are yet to
experience a winter in New Zealand. Before releasing
this agent, we undertook a study to assess the potential
impact of defoliation and improve biological control
practice. The method outlined in this paper provides
Kriticos et al., 2003). Knowledge of the per capita
impacts of putative agents and relative ranges of their
natural rate of increase can provide practitioners with
an indication of the likely relative impacts that agents
with different modes of attack might have on the target
plant (Raghu and Dhileepan, 2005).
For folivorous biological control agents, accurate
determination of their influence on plant growth, and
how these interactions change across environmental
gradients, requires an understanding of the mechanisms by which leaf area reductions influence growth
processes. In many species, reductions in biomass
are proportionately lower than reductions in leaf area
(Langstrom and Hellqvist, 1991; Lavigne et al., 2001),
as plants can respond to defoliation through compensatory growth (McNaughton, 1983; Strauss and Agrawal,
1999). Compensatory responses that have been observed include increased biomass allocation to leaves
(Pinkard and Beadle 1998) and increases in photosynthetic activity (Heichel and Turner, 1983; von Caem-
Figure 1.
The global distribution of Buddleja davidii. (a) The known distribution and (b) the climatic suitability (potential distribution) modelled using CLIMEX (D.J. Kriticos, K.J. Potter and N. Alexander, 2005, unpublished
internal report 37986, Ensis, Rotorua, New Zealand).
33
XII International Symposium on Biological Control of Weeds
a framework for quantifying the net growth impact
of feeding by folivorous biological control agents on
weeds. This method also provides a means of understanding critical levels of defoliation needed to achieve
target levels of weed suppression.
ten blocks and a two-row perimeter buffer. This spacing ensured that plants were not subject to competition
from adjacent plants for light, water or other resources.
The 40 plants within the experiment were randomly
allocated to ten blocks, which included the following
four treatments: (1) undefoliated control, (2) removal
of 33% leaf area, (3) removal of 66% leaf area and
(4) removal of 100% leaf area. For the defoliation treatments, entire leaves were removed on a monthly basis
manually, to simulate the effect of insect defoliation,
from late spring to late summer, initially (November)
on all leaves present, and thereafter (December to February) on newly emerged leaves following the previous
defoliation.
A simple process-based growth model was fitted to
measurements to identify compensatory mechanisms
induced by defoliation and quantify their influence on
above-ground plant biomass (Wp) and the ratio of leaf
to total biomass (Wl/Wp).
Above-ground biomass growth was modelled using
the light use efficiency model. This model determined
on a daily basis the sum of utilizable intercepted radiation from canopy characteristics (leaf area index,
crown diameter), radiation and temperature. Aboveground biomass was then determined as the product
Materials and methods
The experimental site was located adjacent to the Ensis nursery at Rotorua, New Zealand (lat. 38.2S, long.
176.3E). In midwinter of 2004, small B. davidii seedlings were transplanted into single row plots (3 3 m)
laid out in a randomized complete block design, with
A B C D
2.5
Plant height (m)
2.0
1.5
1.0
0.5
0.0
A
1200
Aboveground biomass (g)
Basal diameter (mm)
75
60
45
30
15
400
200
Leaf area (m2)
Crown diameter (m)
1.6
0.4
0.0
May
Figure 2.
Jul
Sep
Nov Jan
Month
Mar
May
600
0
4
0.8
3
2
1
0
Nov
Jul
Figure 3.
Seasonal changes in Buddleia davidii: (a)
height, (b) basal diameter and (c) crown diameter for plants in treatments D0 (thick solid line),
D33 (dotted line), D66 (dashed line) and D100 (thin
solid line). Each point shown is the mean
standard error of ten sample plots. The arrows
A to D indicate the times of defoliations.
34
800
1.2
1000
0
2.0
Jan
Mar
May
Month
Jul
Modelled (a) above-ground biomass and (b) leaf
area for D0 (thick solid line), D33 (dotted line), D66
(dashed line) and D100 (thin solid line). For both
graphs, measured values are shown for D0 (open
triangles), D33 (closed triangles), D66 (closed
diamonds) and D100 (open diamonds). The arrows A to D indicate the times of defoliations.
Clipping the butterfly bushs wings: defoliation studies to assess the likely impact of a folivorous weevil
of utilizable radiation and light use efficiency, and a
fraction was allocated to the leaves. Both estimated
leaf and biomass growth were then added to the value
for the previous day to obtain cumulative total values.
Estimates of plant leaf area were then determined as
the product of specific leaf area and cumulative leaf
mass, from which estimates of radiation interceptance
and biomass growth were then made over the next time
step. Full details of the derivation of the model were
given by Watt et al. (2007).
uncertainties around the population dynamics of exotic
agents before their release and establishment in a new
range, it is unlikely that a precise prediction of an individual agents success could be made using this model.
However, this type of model could at least help assess
the likely effects of folivores compared with agents
from other guilds.
Although mechanical defoliation experiments may
not accurately reflect the full range of effects of herbi
vores (Lehtil and Boalt, 2004; Schooler et al., 2006),
they have been found to be useful for accurately assessing plant responses to various levels of defoliation
(Strauss, 1988; Inouye and Tiffin, 2003; Hjltn, 2004;
Raghu and Dhileepan, 2005; Wirf, 2006; Raghu et al.,
2006; Schooler et al., 2006). Artificial and real herbivory have their respective strengths and weaknesses.
Artificial herbivory can be precisely applied and does
not involve any biosecurity considerations, although
it may not accurately reflect the process of interest. It
can also be applied in situations where the agent cannot
be applied because of, say, biosecurity considerations.
Conversely, real herbivory may be a more direct application of the treatment effect, but it may be difficult
to achieve or measure treatment levels or covariates.
Ideally, both artificial and real herbivory effects should
be measured to draw on the strengths of each approach
(Lehtil and Boalt, 2004; Wirf, 2006).
Results
Values of Wp for treatments D33, D66 and D100 were 61%,
44% and 8%, respectively, compared with the undefoliated control (D0). The defoliation treatments also resulted in significant reductions in plant height, basal
diameter and crown diameter (Fig. 2). The model fitted
data well (Fig. 3) and indicated that increased defoliation was also positively related to light use efficiency,
daily allocation of biomass to leaves and the specific
leaf area and negatively related to rates of natural leaf
loss (M. Watt, unpublished data). Although the plants
were able to change growth characteristics, they were
unable to catch up to the control plants in the course of
a single growing season.
Discussion
Buddleja davidii has a strong tolerance for leaf loss,
including the ability to recover from complete defoliation to a balanced allometric state in a relatively short
period. This would allow it to commence growing rapidly if environmental conditions were favourable and
if the cause of defoliation was removed after the initial
defoliation episode. Nonetheless, there are several fac
tors that give cause for optimism for the chances of
C. japonicus controlling B. davidii under field conditions.
Despite the obvious resiliency, there was a substantial
reduction in plant size at the end of the experiment. If
defoliation by a folivore can reduce the vigour of B.
davidii sufficiently, then desirable vegetation may gain
a competitive advantage over the weed. It is also likely
that repeated defoliation over successive growth seasons would cause further depletion of energy and nutrient reserves. A separate study is examining the effect
over multiple seasons.
Selection of biological control agents is very timeconsuming and costly (McFadyen, 1998). The modelbased approach outlined in this paper could provide
a rapid cost-effective solution for assessing the likely
impacts of candidate biological control agents. Once
parameterized for a particular weed species from field
measurements, the model could be used to examine how
a large number of potential biological control agents,
with a wide range of per capita defoliating intensities,
influence growth of the target species. Given the sensitivity of net defoliation rates to agent abundance and
35
Acknowledgements
Thanks to Samantha Alcaraz for cartography and to
Lindsay Bulman, Mick Crawley, Susan Ebeling and
Nod Kay for providing distribution data for the map
in Fig. 1a. We are also very grateful for the assistance
of Natalie Watkins for measurements undertaken in
the field. This project was funded by the New Zealand
Foundation for Research Science and Technology.
References
Dirzo, R. (1984) Herbivory: a phytocentric overview. In:
Dirzo, R. and Sarukhan, J. (eds) Perspectives on Plant
Population Ecology. Sinauer Associates, Inc, Sunderland,
UK, pp. 141165.
Heichel, G.H. and Turner, N.C. (1983) CO2 assimilation of
primary and regrowth foliage of red maple (Acer rubrum
L.) and red oak (Quercus rubra L.): responses to defoliation. Oikos 57, 1419.
Hjltn, J. (2004) Simulating herbivory: problems and possibilities. In: Weisser, W.W. and Siemann, E. (eds) Insects
and Ecosystem Function. Springer, Heidelberg, Germany,
pp. 244255.
Hoffmann, J.H. (1995) Biological control of weeds: the way
forward, a South African perspective. In: McKinley, R.G.
and Atkinson, D. (eds) Proceedings of the British Crop
Protection Council Symposium. BCPC, Farnham, UK, pp.
7789.
XII International Symposium on Biological Control of Weeds
Inouye, B.D. and Tiffin, P. (2003) Measuring tolerance to herbivory with natural or imposed damage: a reply to Lehtila.
Evolution 57, 681682.
Julien, M.H. and Griffiths, M.W. (1999) Biological control
of weeds. A world catalogue of Agents and Their Target
Weeds, 4th edn. CABI, Wallingford, UK.
Kriticos, D.J. (2003) The roles of ecological models in evaluating weed biological control agents and projects. In:
Spafford-Jacob, H.S. and Briese, D.T. (eds) Improving
the Selection, Testing and Evaluation of Weed Biological
Control Agents. Proceedings of the CRC for Australian
Weed Management Biological Control of Weeds Symposium and Workshop. CRC for Australian Weed Management, Adelaide, Australia, pp. 6974.
Kriticos, D.J., Brown, J.R., Maywald, G.F., Radford, I.D.,
Nicholas, D.M., Sutherst, R.W. and Adkins, S.A. (2003)
SPAnDX: a process-based population dynamics model to
explore management and climate change impacts on an
invasive alien plant, Acacia nilotica. Ecological Modelling 163, 187208.
Langstrom, B. and Hellqvist, C. (1991) Effects of different
pruning regimes on growth and sapwood area of Scots
pine. Forest Ecology and Management 44, 239254.
Lavigne, M.B., Little, C.H.A. and Major, J.E. (2001) Increasing the sink: sources balances enhances photosynthetic
rate of 1-year-old balsam fir foliage by increasing allocation of mineral nutrients. Tree Physiology 21, 417
426.
Lehtil, K. and Boalt, E. (2004) The use and usefulness of
artificial herbivory in plantherbivore studies. In: Weisser,
W.W. and Siemann, E. (eds) Insects and Ecosystem Function. Springer, Heidelberg, Germany, pp. 258275.
Mabry, C.M. and Wayne, P.W. (1997) Defoliation of the annual herb Abutilon theophrasti: mechanisms underlying
reproductive compensation. Oecologia 111, 225232.
McEvoy, P.B., Cox, C.S. and Coombs, E.M. (1991) Successful biological control of ragwort. Ecological Applications
1, 430432.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
McNaughton, S.J. (1983) Compensatory plant growth as a
response to herbivory. Oikos 40, 329336.
Ooi, P.A.C. (1992) Biological control of weeds in Malaysian
plantations. In: Combellack, J.H., Levick, K.J., Parsons, J.
36
and Richardson, R.G. (eds) Proceedings of the 1st International Weed Control Congress, 1721 February 1992,
Melbourne, Australia. Weed Science Society of Victoria,
Melbourne, Australia, pp. 248255.
Pinkard, E.A. and Beadle, C.L. (1998) Above ground biomass partitioning and crown architecture of Eucalyptus
nitens following green pruning. Canadian Journal of Forest Research 28, 14191428.
Raghu, S. and Dhileepan, K. (2005) The value of simulating
herbivory in selecting effective weed biological control
agents. Biological Control 34, 265273.
Raghu, S., Dhileepan, K. and Trevio, M. (2006) Response of
an invasive liana to simulated herbivory: implications for
its biological control. ACTA Oecologia 29, 335345.
Schooler, S., Baron, Z. and Julien, M. (2006) Effect of simulated and actual herbivory on alligator weed, Alternanthera philoxeroides, growth and reproduction. Biological
Control 36, 7479.
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006) Top 20
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 93117.
Strauss, S.Y. (1988) Determining the effects of herbivory using naturally damaged plants. Ecology 69, 16281630.
Strauss, S.Y. and Agrawal, A.A. (1999) The ecology and evolution of plant tolerance to herbivory. Trends in Ecology
and Evolution 14, 179185.
Trumble, J.T., Kolodny-Hirsch, D.M. and Ting, I.P. (1993)
Plant compensation for arthropod herbivory. Annual Review of Entomology 38, 93119.
Von Caemmerer, S. and Farquhar, G.D. (1984) Effects of
partial defoliation, changes of irradiance during growth,
short-term water stress and growth at enhanced p(CO2)
on the photosynthetic capacity of leaves of Phaseolus vulgaris L. Planta 160, 320329.
Watt, M.S., Whitehead, D., Kriticos, D.J., Gous, S.G. and
Richardson, B. (2007) Using a process-based model to
analyse compensatory growth in response to defoliation:
simulating herbivory by a biological control agent. Biological Control 43, 119129.
Wirf, L.A. (2006) The effect of manual defoliation and Macaria pallidata (Geometridae) herbivory on Mimosa pigra:
implications for biological control. Biological Control 37,
346353.
Can a pathogen provide insurance against
host shifts by a biological control organism?
P.B. McEvoy,1 E. Karacetin1,2 and D.J. Bruck3
Summary
The cinnabar moth, Tyria jacobaeae (L.) (Lepidoptera: Arctiidae), is an icon in population ecology
and biological control that has recently lost its shine based on evidence that (a) it is less effective than
alternatives (such as the ragwort flea beetle Longitarsus jacobaeae (Waterhouse) Coleoptera: Chrysomelidae) for controlling ragwort, Senecio jacobaea L. (Asteraceae), (b) it eats (harms) non-target
plant species (including arrowleaf ragwort, Senecio triangularis Hook. (Asteraceae), a native North
American wildflower, and potentially harms the animals that depend on these native plant species
and (3) it carries a disease (caused by a host-specific microsporidian Nosema tyriae). We used a life
table response experiment (LTRE) combining a factorial experiment and a matrix model to estimate
the independent and interacting effects of Old World and New World host plant species (first trophic
level) and the entomopathogen (third trophic level) on the life cycle and population growth of the
cinnabar moth (second trophic level). Host shifts are expected if herbivore fitness is higher on novel
compared with conventional host plants, perhaps because the advantage of reduced effectiveness
of herbivore natural enemies outweighs the disadvantage of herbivore malnutrition associated with
novel host plants. Contrary to this hypothesis, we found the population growth rate of the cinnabar
moth is sharply reduced on novel compared with conventional host plants by interacting effects of
disease and malnutrition. Paradoxically, a pathogen of the cinnabar moth may enhance weed biological control by providing insurance against host shifts.
Keywords: modelling tritrophic interactions, Tyria jacobaeae, pathogenhost
interaction, host specificity, microspora.
Introduction
A persistent concern hangs over the practice of classical
biological control: If some biological control organisms adopt new hosts, what more can be done to contain them? A growing body of evidence suggests that
phytophagous insects commonly adopt new hosts if
given sufficient ecological opportunity, genetic variation in traits related to host use and fitness advantage
to insects adopting new host plant species (Thompson,
2005). The cinnabar moth, Tyria jacobaeae (L.) (Lepidoptera: Arctiidae), introduced to control ragwort, Senecio jacobaea L. (Asteraceae), matches at least two
of three of these requirements: ecological opportunity
1
Department of Botany and Plant Pathology, Oregon State University,
2082 Cordley Hall, Corvallis, OR 97333, USA.
2
Erciyes University, Kayseri, Turkey.
3
USDAARS, Horticultural Crops Research Laboratory, 3420 Northwest Orchard Avenue, Corvallis, OR 97330, USA.
Corresponding author: P.B. McEvoy <[email protected]
.edu>.
CAB International 2008
37
and genetic variation. The cinnabar moth was introduced to control ragwort on farms in lowlands of the
Pacific Northwest in the United States; the unintended
consequence was that it ended up feeding on native
wildflowers in the mountains. The current distribution
of this insect overlaps with potential non-target plant
species (ecological opportunity) (Diehl and McEvoy,
1990), populations of the cinnabar moth vary in heritable traits affecting plant use (genetic variation) (Richards and Myers, 1980) and performance of cinnabar
moths on one non-target species closely matches that
on the target (fitness) (Diehl and McEvoy, 1990). Here
we combine observational, experimental and modelling approaches to investigate how an entomopathogen
might be used to contain an errant control organism.
We use laboratory and modelling studies to show how
an entomopathogen might be operating in this system;
we use field observations on prevalence of pathogen
infection in the wild to document how tritrophic interactions involving an entomopathogen species, an insect species and two plant species are operating in the
field. We outline plans for future research emphasizing
XII International Symposium on Biological Control of Weeds
details of transmission. We conclude with implications
that this research holds for the science, technology and
policy of biological control.
superior colonizer). Third, natural enemies of the cinnabar moth abound. Predators (Myers and Campbell,
1976), parasitoids (Cornell and Hawkins, 1993) and
pathogens (Hawkes, 1973) have been reported to attack
cinnabar moth in North America. One natural enemy,
the pathogen Nosema tyriae, stands out as more prevalent than the rest, with a median prevalence of 70%
measured across 15 populations in the states of California, Oregon and Washington in the United States
(Hawkes, 1973). Diet breadth might be the cinnabar
moths ace in the hole. The fundamental host range
(physiological host range) measured in the laboratory
includes 132 North American plant species and infraspecific taxa, including 20 species in Oregon (Chambers
and Sundberg, 2001). Its realized host range (ecological host range) expressed in the field appears to be
much narrower. One candidate to become a new host
plant, arrowleaf ragwort S. triangularis Hook., stands
out above the rest as accessible, acceptable, suitable
and vulnerable.
If the quality of life for the cinnabar moth has
sharply declined on the Old World host plant species
in North America, then would life be better there on a
New World host plant species (taking all abiotic and
biotic factors into account)?
A model system
Biological control of ragwort has been an economic
and ecological success along the west coast of North
America from British Columbia to Washington, Oregon and northern California (Coombs et al., 1991;
McEvoy et al., 1991). Ragwort has declined to 13%
of its former abundance in that region after introduction of three insect species during a 10-year period:
Tyria jacobaeae (L.) (Lepidoptera: Arctiidae) (cinnabar moth) starting in 1959, Botanophila seneciella
(Meade) (Diptera: Anthomyiidae) (ragwort seed fly,
formerly Hylemia seneciella) starting in 1966 and Longitarsus jacobaeae (Waterhouse) (Coleoptera: Chrysomelidae) (ragwort flea beetle) starting in 1969. There is
a potential downside as well as an upside to biological
control because biological control organisms share attributes of some our worst invaderscapacity to harm,
multiply, spread and evolve. The cinnabar moth is not
a particularly promising biological control organism.
It is less effective than alternatives (such as the ragwort flea beetle L. jacobaeae) for controlling ragwort
(McEvoy et al., 1993; McEvoy and Coombs, 1999). It
eats (harms) non-target plant species (including S. triangularis, a native North American wildflower) (Diehl
and McEvoy, 1990) and potentially harms the animals
that depend on these native plant species. It carries a
disease (caused by a host-specific microsporidian, Nosema tyriae) (Bucher and Harris, 1961; Hawkes, 1973;
Canning et al., 1999). We ask: can we make lemonade
out of this lemon?
Tritrophic interactions
Circumstances favoring host changes by
the cinnabar moth
Quality of life for the cinnabar moth in the New
World has declined on its Old World host plant (ragwort). First, the plant resource has collapsed. Under
pressure from the ragwort flea beetle, ragwort has declined to 13% of its former abundance, leaving little
resource for the cinnabar moth. Second, on the plant resource that remains, competitors of the cinnabar moth
are overpowering it. The cinnabar moth is an inferior
competitor relative to the ragwort flea beetle (McEvoy
et al., 1993; McEvoy and Coombs, 1999), but a superior competitor relative to the ragwort seed head fly
(Crawley and Pattrasudhi, 1988). Markreleaserecapture studies show that the cinnabar moth is inferior as
a colonizer on ragwort relative to both the ragwort flea
beetle and ragwort seed head fly (Harrison and Thomas,
1991; Harrison et al., 1995). Thus, there appears to be
no possibility of coexistence of cinnabar moth with its
competitors on ragwort explained by a colonization/
competition trade-off (when an inferior competitor is a
38
We studied interspecific interactions within a tritrophic system consisting of a host-specific pathogen,
the microsporidian, N. tyriae; the cinnabar moth, T. jacobaeae; and two host plants species, the Old World
host S. jacobaea and the New World host S. triangularis.
Microspora is a phylum of protozoa found as highly
specialized, obligatory, intracellular parasites in nearly
all major animal groups, being especially common in
insects. They are diverse, with approximately 150 genera containing 1200 species. The disease they cause
is called microsporidiosis. They possess unicellular
spores, containing a uninucleate or binucleate sporoplasm and an extrusion apparatus always with a polar
filament and polar cap. Transmission from one host insect to another occurs both horizontally (oral ingestion;
within the same generation) and vertically (mother to
progeny; between generations).
Materials and methods
Life table response experiment
We designed and carried out an LTRE (Caswell,
2001) to estimate the independent and interacting effects of two diets (foliage from Old World and New
World hosts) and five pathogen levels (doses of 0,
101, 102, 103 and 104 spores per individual) on cinnabar moths life cycle and population growth rate. An
Can a pathogen provide insurance against host shifts by a biological control organism?
LTRE combines a factorial experiment and population
model as a way of linking environmental conditions,
vital rates (rates of growth, development, survival, reproduction and movement) and population dynamics.
An LTRE is a powerful way of translating data from
individuals to implications for populations, linking a
populations structure with its dynamics and analysing
the demographic and population-dynamic consequences
of environmental factors.
Methods for a doseresponse experiment
We collected cinnabar moth larvae for these experiments from three field sites in western Oregon,
USA: Santiam Pass (442408N 1215101W) in
the Cascade Mountains, Basket Slough (445708N
1231609W) in the Willamette Valley and Neskowin
(45623N, 1235846W) on the Pacific Coast, anticipating that there might be genetic variation in cinnabar
moths from different geographic locations that could
affect insectplant interactions. We collected infected
larvae from a single population (Neskowin, OR). Infected and uninfected larvae were reared together to
facilitate horizontal transmission of the pathogen. Microsporidium spores were isolated from infected larvae
and suspended in distilled water at different concentrations (0, 101, 102, 103 and 104 spores/l). Spore suspensions were stored at 5 2C for at most 2 months. We
used the same mixtures for every test unit (individual
larva) regardless of the diet. Nosema tyriae was introduced along with the cinnabar moth, and only a single
Nosema sp. (with unusually small spores) is known to
occur in this insect. The microsporidium infecting the
cinnabar moth collected from Neskowin matches the
species description for N. tyriae (Canning et al., 1999).
We did not observe any insects infected with Nosema
sp. We reared insects under optimal conditions (long
day, 16:8 h L/D; temperature, C, 25:15 L/D; humidity, 90%), reared individually (1 oz cup) and fed them
ad lib. There were two diets (foliage of Old World and
New World hosts) five pathogen doses per individual
(spore concentrations, 0, 101, 102, 103 and 104 spores)
= 10 treatment combinations. We collected New World
host plant (S. triangularis) leaves from Marys Peak
(443016N, 1233300W). We grew the Old World
host plant (S. jacobaea) in our greenhouse in individual potsnatural day lengths, temperature (C, 25:15
L/D), humidity (90%). Leaves from both plants were
fresh. We reared uninfected larvae individually through
the first and second instars on both New and Old World
host plants and then fed newly molted third instars 2mm2 leaf disks topically treated with 1 l of each spore
dose, corresponding to a pulse of horizontal transmission. We followed insect development daily for nearly
two generations, allowing for vertical transmission. We
measured vital rates of growth, development, survival
and reproduction in response to diet and pathogen
treatments.
Construction and analysis of a matrix
population model
We constructed and analysed a linear deterministic
matrix model N(t + 1) = A N(t), where N(t) and N(t + 1)
represent vectors of the abundances in each stage from
one time step (t) to the next (t + 1) and A the projection matrix. The life cycle graph (Figure 1) illustrates
the eight life cycle stages representing egg, five larval
stages, pupa and adult. The life cycle graph also illustrates the 16 life cycle transitions in the model, with
seven representing growth g, eight representing stasis
s and one representing fertility f. The time step in the
model is 1 day. The life cycle graph can be represented
as an 8 8 matrix A, which, in turn, can be used to
project the dynamics.
A=
0 0
0 0
s1,1 0 0
f
0 0
0 0
g2,1 s2,2 0
0
0 0
0 g3,2 s3,3 0 0
0
0 0
0 0 g4,3 s4,4 0
0
0 0 0 g5,4 s5,5 0 0
0
0 g6,5 s6,6 0
0 0 0
0
0 0 g7,6 s7,7 0
0 0 0
0 0
0 g8,7 s8,8
0 0 0
The factorial experiment yielded parameter estimates for 20 matrices, one matrix for each of ten
f18
g21
s11
Figure 1.
g32
s22
g43
s33
g54
s44
g65
s55
g76
s66
g87
s77
s88
Life-cycle graph showing the eight stages and 16 transitions in the matrix model used to project cinnabar
moth population growth. The eight life-cycle stages are egg (E), five larval stages (L1, L2, L3, L4 and
L5), pupa (P) and adult (A). The 16 life-cycle transitions in the model include seven representing growth
g, eight representing stasis s and one representing fertility f. The time step in the model is 1 day.
39
XII International Symposium on Biological Control of Weeds
Figure 2.
The relationship between population growth (finite rate of increase ) of the
cinnabar moth population and the treatment factors diet (foliage of New and
Old Host plant species) and pathogen infection (spore dose) for the case of
horizontal transmission only.
Results
treatment combinations (two diets five pathogen doses) two transmission assumptions (case 1, horizontal
transmission only; case 2, horizontal and vertical transmission combined). The finite rate of increase , the
dominant eigenvalue associated with each matrix, was
used as the response variable (population growth rate)
in our experiment.
Figure 3.
Case 1: Horizontal transmission only
Population growth rates of the cinnabar moth declined with increasing Nosema spore dose; the negative
slope of this relationship indicates that the pathogen
has adverse effects (Figure 2). The New World host (S.
The relationship between population growth (finite rate of increase ) of the
cinnabar moth population and the treatment factors diet (foliage of New and
Old Host plant species) and pathogen infection (spore dose) for the case combining horizontal and vertical transmission.
40
Can a pathogen provide insurance against host shifts by a biological control organism?
Figure 4.
The relationship between prevalence of the pathogen Nosema tyriae and elevation in meters
for cinnabar moth populations on Old and New Host plant species. Prevalence is measured
as the percentage of host individuals infected by the pathogen within each host population.
triangularis) was inferior to the Old World host (S. jacobaea) as food; the lower intercept indicates that population growth was lower on New World as compared
with Old World host species. The lines for each host are
parallel, suggesting that diet and pathogen do not interact in their effects. However, qualitative description of
the relationship among population growth, spore dose
and host plant species changes when we increase realism by adding vertical transmission.
prevalence of disease in insects on the Old World host
plant species compared with the New World host plant
species.
Discussion
The strength of the pathogeninsect interaction depends on the plant speciesit is weaker on the Old
World host (S. jacobaea) than on the New World host
(S. triangularis) for mild infections in the laboratory
environment. In other words, mild infections are relatively benign in cinnabar moth populations on Old
World hosts while comparatively virulent in cinnabar
moth populations on New World hosts, under identical
optimal laboratory conditions. This asymmetry tilts the
odds against the non-target host being more acceptable
or more suitable than the target, especially if cinnabar
moth is given a choice between Old World and New
World host plant species.
A remaining challenge is to reconcile our laboratory and field results. If a pathogen is relatively influential in the laboratory and relatively rare in the field
on New World compared with Old World host plants,
it would be wrong to conclude that the pathogen is not
influential in insects on novel host plants in the field.
Pathogens tend to die out as their hosts become rare:
but are cinnabar moths rare because of past epizootics, cool temperatures, unsuitable hosts or some other
causal factor(s)? Mathematical theory of pathogen
host interactions (Anderson and May, 1981) suggests
that (1) there is a minimum, threshold host population
size needed for persistence of a pathogen and (2) intermediate levels of virulence are optimal for increase of
pathogen prevalence. It follows that higher extinction
rates of the pathogen might be expected if, consistent
with our observations, the pathogen is more virulent
Case 2: Horizontal and vertical
transmission combined
When we combined horizontal and vertical transmission, diet and pathogen interacted in their effects
(Figure 3). At low spore doses (left side of the graph),
there was no detectable effect of pathogen infection
in caterpillars on the Old World species and devastating effect of pathogen infection on the New World
host species. The host plant species effect was nil in
uninfected insects and huge in infected insects. At
high spore doses (right side of the graph), the effect of
pathogen infection was so overpowering that no effect
of diet (host plant species) was expressed.
To summarize the results thus far, mild pathogen
infections were devastating on New World host plants
and inconsequential on Old World host plants. By contrast, severe pathogen infections were devastating on
both New and Old World host plant species.
Field Observations
Field prevalence of the pathogen varied with elevation and host plant species (Figure 4). Prevalence
declined with increasing elevation (associated with decreasing temperature) over a range in elevation from
0 to 1645 m. At similar elevations, there was a higher
41
XII International Symposium on Biological Control of Weeds
and cinnabar moths is rarer (due to some combination
of disease and malnutrition) on New World compared
with Old World host plants at a given elevation (and
corresponding ambient temperature).
Cause and effect cannot be established by passive
observation. To investigate a feedback relationship,
we need to interrupt the feedback. It would be useful
to create an outbreak of cinnabar moths at high elevations and see if microsporidian epizootics develop. It
would be useful to know why cinnabar moth populations are smaller at high elevations (>800 m), whether
due to past epizootics, cool temperatures or unsuitable hosts. The ability of pathogens to kill ectothermic
hosts has been shown to depend on host body temperature, which fluctuates with environmental conditions
(Thomas and Blanford, 2003). The thermal sensitivities of plant, insect and pathogen vital rates must all
be taken into account when weighing the outcome of
tritrophic interactions. But for the moment at least, it
seems that entomopathogens can help prevent non-
target effects in the event that an insect biological control agent strays from its target host.
Finding ways to rein in errant classical biological
control organisms is likely to be difficult and costly.
It is better to predict and prevent adverse effects than
to try to mitigate them after the fact. Some scientists
worry that new organisms released into the environment are a potent form of pollution: not only with
the power to have adverse effects on the environment
(like chemicals), but with powers of evolution, replication and autonomous dispersal (unlike chemicals) that
make adverse effects harder to predict and manage.
The same scientists worry that the epidemic of plant
and pest invasions is still not under control. Biological
control should help in the war on weeds. Classical biological control has had the advantage over other control
methods: it is a technology that operates on a scale that
matches the scale of the problem. The obvious bears repeating: do not make things worse by moving the cinnabar moth and other risky control organisms to new
geographic areas containing potential non-target species; that would be counterproductive.
Acknowledgements
We are grateful to Eric Coombs for assistance in all
phases of our work, to other members of our laboratory
for critiquing this work and to Jason Fuller for inspiring this line of research.
References
Anderson, R.M. and May, R.M. (1981) The population dynamics of microparasites and their invertebrate hosts.
Philosophical Transactions of the Royal Society of London B Biological Sciences 291, 451524.
42
Bucher, G.E. and Harris, P. (1961) Foodplant spectrum and
elimination of disease of cinnabar moth larvae, Hypocrita
jacobaeae (L.) (Lepidoptera: Arctiidae). Canadian Entomologist 93, 931936.
Canning, E.U., Curry, A., Cheney, S.A., Lafranchi-Tristem,
N.J., Kawakami, Y., Hatakeyama, Y., Iwano, H. and Ishihara, R. (1999) Nosema tyriae n.sp. and Nosema sp., microsporidian parasites of cinnabar moth Tyria jacobaeae.
Journal of Invertebrate Pathology 74, 2938.
Caswell, H. (2001) Matrix Population Models: Construction,
Analysis and Interpretation. Sinauer, Sunderland, MA.
Chambers, K.L. and Sundberg, S. (2001) Oregon Vascular
Plant Checklist: Asteraceae. Oregon Flora Project, Oregon State University, Corvallis, OR.
Cornell, H.V. and Hawkins, B.A. (1993) Accumulation of
native parasitoid species on introduced herbivores: a comparison of hosts as natives and hosts as invaders. American Naturalist 141, 847865.
Crawley, M.J. and Pattrasudhi, R. (1988) Interspecific competition between insect herbivores: asymmetric competition
between cinnabar moth and the ragwort seed-head fly.
Ecological Entomology 13, 243249.
Diehl, J. and McEvoy, P.B. (1990) Impact of the cinnabar
moth (Tyria jacobaeae) on Senecio triangularis, a nontarget native plant in Oregon. In: Delfosse, E.S. (ed.) Proceedings of the VII International Symposium on Biological Control of Weeds. Ministero dellAgricoltura e delle
Foreste, Rome, Italy/CSIRO, Melbourne, Australia, pp.
119126.
Harrison, S. and Thomas, C.D. (1991) Patchiness and spatial
pattern in the insect community on ragwort Senecio jacobaea. Oikos 62, 512.
Harrison, S., Thomas, C.D. and Lewinsohn, T.M. (1995)
Testing a metapopulation model of coexistence in the insect community on ragwort (Senecio jacobaea). American
Naturalist 145, 546562.
Hawkes, R.B. (1973) Natural mortality of cinnabar moth in
California. Annals of the Entomological Society of America 66, 137146.
McEvoy, P.B. and Coombs, E.M. (1999) Biological control
of plant invaders: Regional patterns, field experiments
and structured population models. Ecological Applications 9, 387401.
McEvoy, P.B., Cox, C. and Coombs, E. (1991) Successful
biological control of ragwort, Senecio jacobaea, by introduced insects in Oregon. Ecological Applications 1,
430442.
McEvoy, P.B., Rudd, N.T., Cox, C.S. and Huso, M. (1993)
Disturbance, competition and herbivory effects on ragwort Senecio jacobaea populations. Ecological Monographs 63, 5575.
Myers, J.H. and Campbell, B.J. (1976) Predation by carpenter
ants: a deterrent to the spread of cinnabar moth. Journal of
the Entomological Society of British Columbia 73, 79.
Richards, L.J. and Myers, J.H. (1980) Maternal influences on
size and emergence time of the cinnabar moth. Canadian
Journal of Zoology 58, 14521457.
Thomas, M. and Blanford, S. (2003) Thermal biology in
insectparasite interactions. Trends in Ecology and Evolution 18, 344350.
Thompson, J.N. (2005) The Geographic Mosaic of Coevolution. University of Chicago Press, Chicago, IL.
Which haystack? Climate matching
to narrow the search for weed
biological control agents
M.P. Robertson,1 C. Zachariades2 and D.J. Kriticos3
Summary
The shrub Chromolaena odorata (L.) King and Robinson (Asteraceae) is highly invasive in southeastern Africa and is the subject of a South African biological control programme. The biotype of C.
odorata growing in South Africa differs in several respects from the more common type noted to be
invasive elsewhere, including its apparent better adaptation to a cool climate. One challenge facing
the biological control programme is the identification of agents that are both suited to develop on this
host biotype and persist in the relatively cool conditions found in South Africa. C. odorata is native
to the Americas, where it has a very extensive distribution spanning a wide range of climates. Two
climate matching computer programmes (CLIMEX and FloraMap) were used to focus the agent
search effort by identifying areas in the Americas that are climatically similar to the invaded region in
southern Africa (SA). Several higher-latitude and higher-altitude areas in South and Central America
were identified by both CLIMEX and FloraMap as being similar to the region invaded by C. odorata
in South Africa. In many areas, the two models agreed, but in others, there were discrepancies, which
are discussed. There was little overlap between the region from which the SA biotype is thought to
have originated and climatically suitable/similar areas in the Americas indicated by either model.
Keywords: agent selection, Chromolaena odorata, CLIMEX, FloraMap.
This article has been published in full as Robertson, M.P., Kriticos, D.J. and Zachariades, C. (2008) Climate
matching techniques to narrow the search for biological control agents. Biological Control, doi: 10.1016/
j.biocontrol.2008.04.002.
Department of Zoology and Entomology, University of Pretoria, Pretoria 0002, South Africa.
2
Plant Protection Research Institute, Agricultural Research Council, Private Bag x6006, Hilton 3245, South Africa.
3
Forest Biosecurity and Protection Unit, Ensis, PB 3020, Rotorua 3201,
New Zealand. Presently at the Forest Biosecurity and Protection Unit,
Ensis, PO Box E4008, Kingston, ACT 2614, Australia.
43
Nutritional characteristics of Hydrilla
verticillata and its effect on two
biological control agents
J.F. Shearer, M.J. Grodowitz and J.E. Freedman
Summary
A complex of abiotic and biotic factors is known to impact the establishment and success of biological
control agents. Experiments using the ephydrid fly Hydrellia pakistanae Deonier have demonstrated
that hydrilla, Hydrilla verticillata (L.f.) Royle, containing low protein content appears to impact
larval development time and the number of eggs oviposited per female. Eggs per female were over
twofold higher for larvae reared on hydrilla containing 2.4-fold more protein. Mean adult female fly
weight peaked when emergence is low (i.e. low crowding) and leaf protein content is high. The hydrilla biological control pathogen Mycoleptodiscus terrestris (Gerd.) Ostazeski also responds to plant
nutritional condition. The nutritional status of hydrilla shoots affects M. terrestris vegetative growth,
disease development and conidia and microsclerotia production. High protein content in shoot tissues
was associated with a more than threefold increase in conidia production and maximum disease severity. In contrast, low protein content in shoot tissues stimulated a 3.7-fold increase in melanized microsclerotia, reproductive structures that are more persistent in the environment than conidia. These
studies suggest that the nutritional condition of target plants cannot be excluded as an important factor
in efficacy of biological control agents. Both agents responded to favorable conditions by reproducing
prolifically, which ultimately resulted in increased host damage.
Keywords: Hydrellia pakistanae, Mycoleptodiscus terrestris, evaluation.
Introduction
stanae individuals have been released with established
populations occurring in Florida, Arkansas, Alabama,
Georgia and Texas (Center et al., 1997; Julien and Griffiths, 1998; Grodowitz et al., 1999). Field establishment has generally been excellent with close to 90%
establishment observed (Center et al., 1997; T. Center,
unpublished data). Populations are now found far removed from their original release sites, indicating the
fly is spreading naturally throughout the southeastern
United States. Significant Hydrellia spp. impact has
been observed at sites in Texas, Florida and Georgia
(Grodowitz et al., 2003a,b) but significant increases
in fly populations and subsequent impact have not occurred at many sites. Reasons are not completely understood.
Mycoleptodiscus terrestris reproduces asexually by
thin-walled conidia and by melanized survival structures called microsclerotia. To date, sexual reproduction
of the fungus has not been observed, therefore sexual
spores were not an issue in this study. Conidia develop
from spore-producing structures called sporodochia
following ingress by the pathogen. The sporodochia
form on tissue surfaces within 5 to 7 days postinoculation
In aquatic systems, there is scant information on the
impact of biological control agents relative to the
physical and nutritional characteristics of submersed
aquatic macrophytes. Two agents of hydrilla, Hydrilla
verticillata (L.f.) Royle, the Asian leaf-mining fly
Hydrellia pakistanae Deonier (Diptera: Ephydridae)
and the pathogenic fungus Mycoleptodiscus terrestris
(Gerd.) Ostazeski (Ascomycota: Magnaporthaceae),
perform extremely well under laboratory, greenhouse
and experimental conditions (Doyle et al., 2002;
Shearer, 2002; Shearer and Nelson, 2002; Grodowitz
et al., 2003a,b; Owens et al., 2006; Shearer and Jackson, 2006) but at times are inconsistent in their ability
to successfully reduce hydrilla populations under field
conditions. Since 1987, more than 20 million H. paki
U.S. Army Engineer Research and Development Center, 3909 Halls
Ferry Road, Vicksburg, MS 39180, USA.
Corresponding author: J.F. Shearer <[email protected].
mil>.
CAB International 2008
44
Nutritional characteristics of Hydrilla verticillata and its effect on two biological control agents
followed within a day by commencement of spore production. It has been documented that spore production
may vary in relation to the substrata available and to
environmental variables such as stress or disturbance
(Dix and Webster, 1995). Under optimum conditions in
greenhouse studies, the hydrilla pathogen M. terrestris
is consistently pathogenic to hydrilla and can reduce
shoot biomass by 97% to 99% (Shearer, 2002). How-
Figure 1.
ever, subjecting field populations of hydrilla to similar
rates of M. terrestris inoculum has often produced inconsistent results.
Potential factors that might limit agent performance
on hydrilla include parasites, predators, temperature,
water flow, turbidity, plant density, age and plant nutritional status. To better understand the importance of
plant nutrition on agent performance, hydrilla plants of
Proximate analysis of hydrilla shoot tissues for (a) insect and (b) pathogen study. Sediments were nutrient-
deficient (Used) or nutrient-enriched (Fert = fertilized). Plants received ambient air (Air) or air enriched with
carbon dioxide (CO2). Growth periods were 4 weeks (Short) or 10 weeks (Long).
45
XII International Symposium on Biological Control of Weeds
Figure 2.
Correlation between crude protein and days to first emergence of Hydrellia pakistanae.
varying nutritional status were challenged with H. pakistanae and M. terrestris.
Materials and methods
Plant growth
Hydrilla plants of known nutritional composition
were produced by growing them in used or fertilized
sediments under different aeration conditions (high or
low CO2) using procedures described by Grodowitz
and McFarland (2002) and Shearer et al. (2007). The
used sediment was rendered nitrogen-poor because of
previous growth of submersed macrophytes. Fertilized
sediments were amended with 0.7 g NH4Cl per liter of
wet sediment. Additionally, for the insect experiment,
period of growth was varied (long vs. short) to produce plants having varying degrees of leaf hardness as
measured by a penetrometer. Nutritional parameters,
including percent ash, crude protein, ether-extractable
compounds, crude fiber and nitrogen-free extract, were
determined using a standard feed analysis known as a
proximate analysis described in detail by Grodowitz
and McFarland (2002). Phosphorous concentration was
determined using atomic absorption techniques.
Insect biological control agent
Insects were reared in a greenhouse, beginning with
50 eggs per container, on hydrilla plants of varying
46
nutritional composition in 3.5-l containers in a water
bath maintained at 2225C (Freedman et al., 2001).
Emerged adults were removed from the containers daily
and released into oviposition chambers (30.5 30.5
30.5 cm). Percent emergence was calculated. Each
treatment was replicated five times.
Within the oviposition chambers, females were allowed to oviposit freely onto five to seven hydrilla apical shoots held within an open 100 15-mm (d h)
Petri dish containing deionized water. After the adults
died, the sex ratio was recorded and dessicated females
were weighed. Hydrilla shoots were removed from
the oviposition chambers every 3 to 5 days, eggs were
identified and counted and number of eggs per female
was calculated.
Pathogen biological control agent
Hydrilla apical shoots (5 cm) of variable nutritional
compositions were placed in 250-ml Erlenmeyer flasks
containing 150 ml sterile water and 20 l wet inoculum. Inoculum was prepared as described by Shearer
et al. (2007). Control flasks received an additional 20
l of sterile water. Each treatment was replicated five
times. The flasks were randomly arranged on a rotary
shaker (Innova 2300, New Brunswick Scientific, Edison, NJ) set at 50 rpm and incubated at room temperature for 2 weeks.
At 7 and 14 days postinoculation, the hydrilla shoots
were visually assessed for disease development based
Nutritional characteristics of Hydrilla verticillata and its effect on two biological control agents
Figure 3.
Figure 4.
Correlation between hydrilla phosphorous content and eggs per female for Hydrellia pakistanae.
3-D surface plot with data points marked for percent emergence of Hydrellia pakistanae vs weight per
female and leaf protein content.
47
XII International Symposium on Biological Control of Weeds
Figure 5.
Effects of hydrilla fertilization levels on Mycoleptodiscus terrestris
(a) disease development, (b) asexual spore production in the form of
conidia and (c) production of survival structures or microsclerotia.
Statistical analysis
on a disease rating scale from 0 to 4, where 0 = green
and healthy, 1 = slight chlorosis, 2 = general chlorosis, 3 = tissues flaccid and disarticulating and 4 = complete tissue collapse. At 14 days postinoculation, the
flasks were gently shaken to dislodge any spores that
had developed on infected tissue surfaces. The number
of spores released into the water was then determined
using a hemacytometer. Three leaves were randomly
retrieved from each flask to count microsclerotia that
had developed within leaf tissues.
Statistical analyses were performed using Statistica
version 7.1 (Statsoft, 2005) and included ANOVA,
correlation analysis and a distance-weighted least
square means graphing technique to visualize threedimensional trends with corresponding measures of
the amount of variance explained (i.e. R). Statistical
significance was assumed at or below P = 0.05, unless
otherwise noted.
48
Nutritional characteristics of Hydrilla verticillata and its effect on two biological control agents
Results and discussion
tween percent crude protein and days to first emergence
was observed where higher crude protein values were
associated with fewer days to first emergence (Fig. 2).
This was not surprising, as similar results were noted in
experiments conducted by Wheeler and Center (1996),
where larvae reared on harder hydrilla leaves (and lower protein) resulted in longer developmental times.
Plants grown in fertilized sediments gave rise to female flies that laid more than twice as many eggs (df
= 1, 32, P < 0.00017), as female flies that were reared
on plants in used sediments. Mean number of eggs per
female was 7.8 for used sediments compared with 17.2
for fertilized sediments. Although there was a strong
linear relationship between phosphorous and protein
content in plant tissues, egg production appeared to be
more strongly correlated with phosphorous than protein. The r values were higher when egg numbers were
correlated with phosphorous (Fig. 3, r = 0.87) than with
crude protein (r = 0.65).
There is an interesting relationship among weight
per female (an indication of fecundity), crude protein
and percent emergence as an indicator of crowding
(Fig. 4). Female fly weight peaks when emergence is
low (i.e. low crowding) and protein is high. However,
as percent emergence increases, leading to increased
crowding and competition amongst larvae, the emerging female weight remains low even at high protein
levels. Hence, crowding strongly influences female
weight and most likely fecundity.
Plant nutritional status
By manipulating growing conditions, hydrilla plants
were produced with significant differences in nutritional composition for percent nitrogen-free extract
(soluble sugar, starch and some hemicelluloses), crude
fiber (cellulose and some lignin), ether-extractable
compounds (lipids and fats), crude protein (total nitrogen) and ash (mineral content) (Fig. 1). Of particular
note was that crude protein, as a measure of total nitrogen, was approximately twofold higher in plants grown
in fertilized sediments compared with plants cultured
in used or nutrient-depleted sediments. Protein levels
were similar for corresponding treatments for plants
used for both the insect and pathogen experiments.
Insect response to plant nutrition
Significant difference in days to first emergence (an
indication of development time) was noted for both the
fertilized (df = 1, 32, P = 0.0009) and growth period
(df = 1, 32, P = 0.001) main effects only. Time to first
emergence was 2 days shorter in fertilized sediments as
compared with used sediments and 2 days longer for
plants grown for longer periods under cooler temperatures compared with shorter growth periods at higher
temperatures. As expected, a significant correlation be-
Figure 6.
Relationship between percent crude protein in hydrilla leaf tissues and production of Mycoleptodiscus
terrestris microsclerotia.
49
XII International Symposium on Biological Control of Weeds
Pathogen response to plant nutrition
Fourteen days postinoculation with M. terrestris,
disease ratings between plants grown in fertilized and
used sediments were significantly higher (df = 1, 16, P
= 0.0001; Fig. 5a) than for plants grown in used sediments. Although the leaves of plants grown in low-
fertility sediment were chlorotic and becoming flaccid,
the stems remained intact. In field situations, such plants
would probably recover and regrow from undamaged
root crowns (Netherland and Shearer, 1996). The highest disease severity rating (Fig. 5a) was consistently
found on inoculated hydrilla that had high leaf-protein
content. These plants collapsed to the bottom of the
flasks and, lacking cell integrity, would have had no
possibility of recovery. Other studies have documented
that high leaf-protein content is often associated with
increases in disease severity (Ghorbani et al., 2002;
Latty et al., 2002).
Plant nutritional status also affected the pathogens
reproductive ability. Mycoleptodiscus terrestris conidial production appeared to be influenced by the substrate. This is indicated by significantly higher numbers
of spores produced in flasks containing hydrilla plants
grown in high-fertility sediments (df = 1, 16, P = 0.0021)
(Fig. 5b). In contrast to conidia, significantly higher
numbers of vegetative reproductive structures, microsclerotia, were present in leaves of hydrilla plants grown
in low-fertility or used sediment at 14 days postinocu-
Figure 7.
lation (df = 1, 16, P = 0.0028) (Fig. 5c). Lacking nutrients for continued mycelial growth, M. terrestris, in all
likelihood, used the available nutrients in plant tissues
and mycelium for production of survival structures.
The highest number of microsclerotia developed in
leaves from plants that had the lowest available nitrogen. The response was strongly curvilinear, suggesting that microsclerotia production may be triggered by
some threshold level of leaf protein, perhaps <9% (Fig.
6). Limited nitrogen availability apparently induced
changes in the pathogen that altered growth from active
proliferation, i.e. conidial production, to preparation for
a period of dormancy or lack of resources, i.e. microsclerotia production. A similar curvilinear response for
microsclerotia numbers and spore production was observed, indicating that such a shift had occurred (Fig. 7).
Implications for biological control
Hydrilla plant nutrition affected the two very different biological control agents in similar ways. High
hydrilla leaf-protein content stimulated agent reproduction as indicated by increased H. pakistanae female
weight and fecundity and increased conidial production
and disease severity in the case of M. terrestris. Low
leaf-protein content in hydrilla negatively affected reproduction of both biological control agents, resulting
in increased developmental times and reduced fecundity
for the insect and a shift by the pathogen to a higher
Relationship between Mycoleptodiscus terrestris microsclerotia and spore production.
50
Nutritional characteristics of Hydrilla verticillata and its effect on two biological control agents
production of microsclerotia. Based on the case studies, it appears that plant nutritional quality could have
been a factor in inconsistent field results in the past and
that higher quality plants should lead to increased agent
establishment and impact. Support for this conclusion
can be found in a study using the salvinia weevil, Cyrtobagous salviniae Calder and Sands, where fertilization
of giant salvinia plants, Salvinia molesta D.S. Mitchell, in the field substantially aided the establishment of
the agent and ultimately lead to a successful biological
control effort (Room and Thomas, 1985).
Acknowledgements
This research was conducted under the U.S. Army
Corps of Engineers Aquatic Plant Control Research
Program, Environmental Laboratory, U.S. Army Engineer Research and Developmental Center, Waterways
Experiment Station, Vicksburg, MS. Permission was
granted by the Chief of Engineers to publish this information. We thank Jenny Goss and Harvey Jones for
technical assistance. We would also like to thank inhouse reviewers Linda Nelson and Chetta Owens and
the symposium editors whose comments significantly
improved the manuscript.
References
Center, T.D., Grodowitz, M.J., Cofrancesco, A.F., Jubinsky,
G., Snoddy, E. and Freedman, J.E. (1997) Establishment
of Hydrellia pakistanae (Diptera: Ephydridae) for the
biological control of the submersed aquatic plant Hydrilla
verticillata (Hydrocharitaceae) in the Southeastern United
States. Biological Control 8, 6573.
Dix, N.J. and Webster, J. (1995) Fungal Ecology. Chapman
& Hall, London.
Doyle, R.D., Grodowitz, M.J., Smart, R.M. and Owens, C.S.
(2002) Impact of herbivory Hydrellia pakistanae (Diptera:
Ephydriadae) on growth and photosynthetic potential of
Hydrilla verticillata. Biological Control 24, 221229.
Freedman, J.E., Grodowitz, M.J, Cofrancesco, A.F. and Bare,
R. (2001) Mass-rearing Hydrellia pakistanae Deonier,
a biological control agent of Hydrilla verticillata (L.f.)
Royle, for release and establishment. ERDC/EL TR01-24,
U.S. Army Engineer Research and Development Center,
Vicksburg, MS.
Ghorbani, R., Scheepens, P.C., Zweerde, W.V.D., Leifert, C.,
McDonald, A.J.S. and Seel, W. (2002) Effects of nitrogen availability and spore concentration on the biocontrol
activity of Ascochyta caulina in common lambsquarters
(Chenopodium album). Weed Science 50, 628633.
Grodowitz, M.J., Freedman, J.E., Cofrancesco, A.F. and
Center, T.D. (1999) Status of Hydrellia spp. (Diptera:
Ephydridae) release sites in Texas as of December 1998.
Miscellaneous Paper A-99-1, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
51
Grodowitz, M.J. and McFarland, D.G. (2002) Developing
methodologies to assess the influence of nutritional and
physical characteristics of Hydrilla verticillata on its
biological control agents. ERDC/EL TN-APCRP-BC-05,
U.S. Army Engineer Research and Development Center,
Vicksburg, MS.
Grodowitz, M.J., Smart, R.M., Doyle, R.D., Owens, C.S.,
Bare, R., Snell, C., Freedman, J. and Jones, H. (2003a)
Hydrellia pakistanae and H. balciunasi insect biological
agents of hydrilla: boon or bust? In: Cullen, J.M., Briese,
D.T., Kriticos, D.J., Lonsdale, W.M., Morin L. and Scott,
J.K. (eds) Proceedings of the XI International Symposium
on Biological Control of Weeds. CSIRO Entomology,
Canberra, Australia, pp. 529538.
Grodowitz, M.J., Cofrancesco, A.F., Stewart, R.M., Madsen,
J. and Morgan, D. (2003b) Possible impact of Lake Seminole Hydrilla by the introduced leaf-mining fly Hydrellia
pakistanae. ERDC/EL TR-03-18, U.S. Army Engineer
Research and Development Center, Vicksburg, MS.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A World Catalogue of Agents and Their Target
Weeds, 4th edn. CABI Publishing, Wallingford, UK.
Latty, E.F., Canham, C.D. and Marks, P.L. (2002) Beech bark
disease in northern hardwood forests: the importance of
nitrogen dynamics and forest history for disease severity.
Canadian Journal Forest Research 33, 257268.
Netherland, M.D. and Shearer, J.F. (1996) Integrated use of
fluridone and a fungal pathogen for control of Hydrilla.
Journal Aquatic Plant Management 34, 48.
Owens, C.S., Grodowitz, M.J., Smart, R.M., Harms, N.E.
and Nachtrieb, J.M. (2006) Viability of hydrilla fragments
exposed to different levels of insect herbivory. Journal
Aquatic Plant Management 44, 145147.
Room, P.M. and Thomas, P.A. (1985) Nitrogen and establishment of a beetle for biological control of the floating
weed Salvinia in Papus New Guinea. Journal of Applied
Ecology 22, 139156.
Shearer, J.F. (2002) Effect of a new growth medium on Mycoleptodiscus terrestris (Gerd.) Ostazeski. Techical Note
TN-APCRP-BC-04, U.S. Army Engineer Research and
Development Center, Vicksburg, MS.
Shearer, J.F., Grodowitz, M.J. and McFarland, D.G. (2007)
Nutritional quality of Hydrilla verticillata (L.F.) Royle
and its effects on a fungal pathogen Mycoleptodiscus
terrestris (Gerd.) Ostazeski. Biological Control 41, 175
183.
Shearer, J.F. and Jackson, M.A. (2006) Liquid culturing of
microsclerotia of Mycoleptodiscus terrestris, a potential
biological control agent for the management of hydrilla.
Biological Control 38, 298306.
Shearer, J.F. and Nelson, L.S. (2002) Integrated use of endothall and a fungal pathogen for management of the
submersed macrophyte Hydrilla verticillata (L.f.) Royle.
Weed Technology 16, 224230.
StatSoft. (2005) Statistica Version 7.1. StatSoft, Tulsa, OK.
Wheeler, G.S. and Center, T.D. (1996) The influence of hydrilla leaf quality on larval growth and development of the
biological control agent Hydrellia pakistanae (Diptera:
Ephydridae). Biological Control 7, 19.
How sensitive is weed invasion to
seed predation?
R.D. van Klinken,1 R. Colasanti1 and Y.M. Buckley2,3
Summary
Seed predators are typically easy to test, rear, release and establish and can be particularly useful
when targeting invaders that also provide benefits to some parts of the community. However, seed
predation rates are generally considered too low to cause significant population regulation of invasive
plants. They may, however, impact on invasion rates (here defined as the combined effects of spread
and infill), although this has proved difficult to demonstrate. In this paper, we use a cellular automaton
model to test whether the effect of seed predation on the regulation of existing populations is influenced by the seed dispersal mechanism and how the addition of a seed predator to an existing population affects invasion rates. We found that population regulation occurs at significantly lower seed
predation rates for poor dispersers as compared with good dispersers (existing models commonly
assume random dispersal) and that seed predation impacts on invasion rate at predation rates that are
commonly observed in the field. Further analysis is required to test how robust these conclusions are
for different plant parameters, for a range of dispersal mechanisms (including long-distance dispersal)
and in heterogeneous landscapes.
Keywords: agent selection, population regulation, seed dispersal mechanisms, spread
rates.
Introduction
Seed predators have a mixed record in weed biological
control. They are typically easy to test, rear, release and
establish and can be particularly useful when targeting
beneficial plants. However, population modelling and
field observations suggest that seed predation rates need
to be very high to regulate plant populations, typically in
the order of 9099% seed mortality (Myers and Risling,
2000; Sheppard et al., 2002; Buckley et al., 2005), as invasive plants are normally microsite-limited (Crawley,
1992). Some biological control agents do cause sufficient seed mortality to regulate populations (Louda and
Potvin, 1995; Hoffmann and Moran, 1998), but most bi
ological control agents probably do not (Crawley, 1992;
R.D. van Klinken, unpublished data).
CSIRO Entomology and CRC for Australian Weed Management, 120
Meiers Road, Indooroopilly, Brisbane, QLD 4068, Australia.
2
University of Queensland, School of Integrative Biology, St. Lucia,
Brisbane, QLD 4072, Australia.
3
CSIRO Sustainable Ecosystems, 306 Carmody Road, St. Lucia, QLD
4067, Australia.
Corresponding author: R.D. van Klinken <[email protected].
CAB International 2008
52
Seed predators may, however, impact on invasive
plant populations in ways other than population regulation per se (van Klinken et al., 2004). Of those, reduced
invasion rate (defined here as the combined effects of
spread and infill) is likely to offer the most substantial
benefits for management. Invading populations may be
seed-limited rather than microsite-limited along their
invading front and may therefore be most sensitive to
reductions in viable seed through seed predation or
other mechanisms. If biological control agents can or
do reduce invasion rates, then it is important to demonstrate such benefits. However, it is difficult to do so
empirically, and to our knowledge, this has not been
done. However, modelling provides opportunities for
testing the likely sensitivity of invasion rates to seed
predation under a range of circumstances.
A range of mathematical modelling approaches
have been used to test the relationship between fecundity and spread speed (Buckley et al., 2005; Hastings
et al., 2005). However, modelling approaches used so
far are not well-suited to exploring the spatial effects of
demographic change and landscape heterogeneity on
plant invasion (With, 2002). We therefore developed
a cellular automaton model to explore the interactions
amongst dispersal mechanism, seed predation and invasion. We present early results of this work, which tests
How sensitive is weed invasion to seed predation?
(1) whether seed dispersal mechanisms influence the
way populations are regulated by seed predators and
(2) what effect adding a seed predator to an existing
population has on invasion rates.
We restrict our discussion to seed predators. However, results and conclusions are equally applicable to
factors that might limit seed production, including a
wide range of biological control agents that attack immature reproductive organs or stress plants.
automaton models are spatial models that treat time
and space in discrete elements. In our model, an iteration, which represents a single growing year, comprises
plant establishment and survival, seed production and
seed dispersal. Each cell of the cellular automaton can
contain a single plant, which has an individual age and
can produce a number of individual seeds. At maturity, a
plant can produce seed but is subject to a death rate that
is proportional to its age, with a probability of death of
95% after 20 years. Plant parameters were set to model
a simplified perennial woody shrub (Table 1).
Seed dispersal between grid cells was modelled us
ing a random-walk algorithm, which results in a normally distributed dispersal kernel. A proportion of seeds
within each cell fall to the ground, and the remainder
are randomly distributed between the cell and its imme
diate neighbours. This algorithm is applied to each cell
Methods
Model design
The model is based on a cellular automaton model
of plant populations (Colasanti et al., 2007). Cellular
8000
a)
.Total population
Population in original
40x40 grid cells
6000
4000
Average number of plants
2000
0
8000
b)
6000
4000
2000
0
0%
Figure 1.
20%
40%
60%
Seed predation
80%
100%
A comparison of the effect of adding seed predators to
an existing 40 40-grid cell population (maximum popu
lation size = 1600 adults), assuming (a) good dispersal
(after 25 years) and (b) poor dispersal (after 50 years) (average SE). Data shows the number of adults in the source
area and the total adult population after 25 and 50 years of
seed predation, respectively.
53
XII International Symposium on Biological Control of Weeds
Table 1.
Variables used in the model and parameter
values.
Variables
Time step
Seed production
Seed bank
Time to maturity
Adult lifespan
Probability to maturity
(from seed)
Seed predation
repeated 20 times, and the average values and standard
deviations were calculated.
Parameters
Results and discussion
Annual
1024/plant/year
None
5 years
20 years (95% probability)
0.01
Effect of seed predation on population
regulation
099%
and repeated until there are no more seeds to distribute.
The fall rate was set to 10% to model a good disperser
and 90% for a poor disperser.
The effect of seed predation was modelled for an
established population that occupied a 40 40-cell grid
block in the centre of a 160 160-cell matrix. The num
ber of plants that occurred within the established block
was counted separately from those plants that were
outside this block. This was done at the end of a fixed
number of iterations (years), 25 for the good disperser
and 50 for the poor disperser; the former had a much
greater rate of spread and was thus stopped sooner.
Predispersal seed predation was simulated by reducing the number of seeds produced by the plant from
a maximum of 1024 to a minimum of 2 seeds using
an exponential series (1024, 512, 2562). The experiments for each level of seed predation (099%) were
When seed dispersal approached random, seed predation needed to be at least 90% before it began having
a significant impact on plant density within the estab
lished 40 40-grid cell population (Fig. 1a). This thresh
old is sensitive to model parameters, including seed
production per plant per year (here set at 1024 seeds).
However, a threshold of 90% or greater is a common
conclusion for population modelling.
Our model showed that seed dispersal mechanisms
can affect the way in which even relatively small populations are regulated. Where dispersal was poor, less
seed predation is required to have an impact on plant
population density (Fig. 1b). Poor seed dispersal results in seeds being aggregated under and adjacent to
parent trees, thereby producing a more heterogeneous
response to seed predation across the site.
Effect of seed predation on invasion rate
Invasion rate in our model was restricted by the
surface area of the invasion front, as there were no
long-distance dispersal events. It was also delayed because plants took 5 years to reach maturity. As would
be expected, invasion rate was much faster for a good
100%
Proportion of trees if no seed predation
Good disperser
Poor disperser
80%
60%
40%
20%
0%
0%
20%
40%
60%
80%
100%
Seed predation
Figure 2.
The effect of seed predation on invasion rate (as a proportion of the total number of trees present
outside the source population in the absence of seed predation) assuming good and poor dispersal.
54
How sensitive is weed invasion to seed predation?
Acknowledgements
disperser (Fig. 1a), with the adult population increasing
fivefold after only 25 years compared with only a 1.8fold increase after 50 years for a poor disperser (Fig.
1b), in the absence of seed predation.
Seed predation had an immediate effect on invasion
rate (as a proportion of invading trees in the absence
of seed predation), with substantial reductions occurring at seed predation rates well below that required to
affect population regulation (Fig. 1). The effect was,
however, proportionally greatest when dispersal was
poor (Fig. 2).
We thank the Australian Weed Management Co-
Research Centre and the Department of Agriculture,
Fisheries and Forestry for the funding support that
made this research possible.
References
Other factors that might be important
Our model is currently very simplistic and begs the
question of how conclusions will differ with different
plant parameters, dispersal mechanisms and landscape
structure. Other models have already partially explored
the effect of demographic parameters on invasion rates
and found the system to be relative insensitive to rea
listic variation in some plant parameters, including adult
longevity and time to reproduction (Buckley et al.,
2005). Our results show that dispersal mechanisms can
clearly interact with the way seed predation affects both
population regulation and invasion rates. The next step
will be to test the effect of rare to frequent long-distance
dispersal events. Landscape heterogeneity is also likely
to be important, through its effect on recruitment prob
abilities (Buckley et al., 2005), seed production (Payn
ter et al., 1996) and dispersal probabilities. Other likely
moderating factors include Allee effects (which result
in seed production increasing with adult density), density dependence of seed predators and seed predators
attacking seeds before and/or after seed dispersal.
Implications for biological control
Our results suggest that the seed predation rates required to regulate plant populations may not need to be
as high as previously thought but that it will vary with
seed dispersal mechanism. However, the required seed
predation rates may still be too high for most species of
seed predators to realistically achieve (R.D. van Klinken,
unpublished data).
In contrast, invasion rates are much more sensitive
to seed reductions. Sensitivities will depend on dispersal mechanisms and landscape heterogeneity. However,
our results suggest that even commonly observed seed
predation rates may be sufficient to result in significantly reduced invasion rates.
The likelihood that even relatively low seed predation rates can reduce invasion rates supports the suggested practice of releasing seed predators as early as
possible in the invasion process, including at the time
of introduction of new agroforestry species (Zimmermann and Neser, 1999). However, further work is still
required to simulate the effect of seed predation for
specific species and landscapes.
55
Buckley, Y.M., Brockerhoff, E., Langer, L., Ledgard, N.,
North, H. and Rees, M. (2005) Slowing down a pine invasion despite uncertainty in demography and dispersal.
Journal of Applied Ecology 42, 10201030.
Colasanti, R.D., Hunt, R. and Watrud, L. (2007) A simple cellular automaton model for high-level vegetation dynamics
Ecological Modelling 203, 363374.
Crawley, M.J. (1992) Seed predators and plant population
dynamics. In: Fenner, M. (ed.) The Ecology of Regeneration in Plant Communities. CAB International, Wallingford, UK, pp. 157192.
Hastings, A., Cuddington, K., Davies, K.F., Dugaw, C.J., El
mendorf, S., Freestone, A., Harrison, S., Holland, M., Lambrinos, J., Malvadkar, U., Melbourne, B.A., Moore, K.,
Talyor, C. and Thomson, D. (2005) The spatial spread
of invasions: new developments in theory and evidence.
Ecology Letters 8, 91101.
Hoffmann, J.H. and Moran, V.C. (1998) The population dynamics of an introduced tree, Sesbania punicea, in South
Africa, in response to long-term damage caused by different combinations of three species of biological control
agents. Oecologia 114, 343348.
Louda, S.M. and Potvin, M.A. (1995) Effect of inflorescencefeeding insects on the demography and lifetime fitness of
a native plant. Ecology 76, 229245.
Myers, J.H. and Risling, C. (2000) Why reduced seed production is not necessarily translated into successful biological
weed control. In: Spencer, N. (ed.) Proceedings of the X
International Symposium on Biological Control of Weeds.
Montana State University, Bozeman, MT, pp. 569581.
Paynter, Q., Fowler, S.V., Hinz, H.L., Memmott, J., Shaw,
R., Sheppard, A.W. and Syrett, P. (1996) Are seed-feeding
insects of use for the biological control of broom? In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of the IX
International Symposium on Biological Control of Weeds.
University of Cape Town, Cape Town, South Africa, pp.
495501.
Sheppard, A.W., Hodge, P. Paynter, Q. and Rees, J. (2002)
Factors affecting invasion and persistence of broom Cytisus scoparius in Australia. Journal of Applied Ecology
39, 721734.
van Klinken, R.D., Kriticos, D., Wilson, J. and Hoffmann, J.
(2004) Agents that reduce seed productionessential ingredients or fools folly? In: Cullen, J.M., Briese, D.T.,
Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott, J.K.
(eds) Proceedings of the XI International Symposium on
Biological Control of Weeds. CSIRO Entomology, Canberra, ACT, Australia, pp. 621622.
With, K.A. (2002) The landscape ecology of invasive spread.
Conservation Biology 16, 11921203.
Zimmermann, H.G. and Neser, S. (1999) Trends and prospects for biological control of weeds in South Africa. African Entomology Memoir 1, 165182.
XII International Symposium on Biological Control of Weeds
Altered nutrient cycling as a novel non-target
effect of weed biocontrol
I.E. Bassett,1,2 J. Beggs1 and Q. Paynter2,3
University of Auckland, Private Bag 92019, Auckland Mail Centre, Auckland 1142, New Zealand
2
Cooperative Research Centre for Australian Weed Management, Waite Campus, PMB 1,
Glen Osmond, SA 5064, Australia
3
Landcare Research, Private Bag 92170, Auckland Mail Centre, Auckland 1142, New Zealand
Invasive weeds have been shown to alter decomposition rates and nutrient cycling in invaded systems.
However, little attention has been paid to the role of introduced biological control agents in influencing
nutrient cycling. Alligator weed (Alternanthera philoxeroides) invades waterways in northern New
Zealand. It is partially controlled in these environments by the alligator weed flea beetle, Agasicles
hygrophila, which was introduced as a biological control agent in the early 1980s and has since become widespread. Annual biomass dynamics and litter decomposition rates of alligator weed were
compared with those of two native sedges, Schoenoplectus tabernaemontani and Isolepis prolifer, in
a New Zealand lake. Herbivory by the alligator weed flea beetle caused a 75% decrease in alligator
weed above-ground biomass over a 3-month period in late spring and early summer. This contrasted in
both timing and magnitude of litter input by either native plant species. In addition, alligator weed litter decomposed significantly faster than the litter of either native species. This combination is likely to
alter the annual availability of nutrients within this ecosystem, with potential flow on effects including
facilitation of further weed invasion.
Interactions of plant quality and predation affect the
success of purple loosestrife biocontrol programme
A. Dvalos and B. Blossey
Department of Natural Resources, Cornell University, Fernow Hall, Ithaca, NY 14853, USA
Biocontrol of purple loosestrife (Lythrum salicaria) in North America is generally considered highly
successful. Yet, at certain sites, biocontrol agents fail to control their host plant. Field observations
indicate that leaf-feeding biocontrol agents Galerucella pusilla and Galerucella calmariensis are more
abundant in flooded than in dry sites. We tested two hypotheses: (1) leaf beetles suffer increased predation in dry areas and (2) superior plant quality in flooded areas results in improved leaf beetle performance. To test Hypothesis 1, we conducted predator exclusion experiments at multiple sites. We found
marginal effects of exposure to predation on leaf beetle survival, but survival was always higher under
flooded conditions. To test Hypothesis 2, we conducted a common garden experiment where we grew
L. salicaria plants under three water levels. We measured several plant traits that are potentially related
to plant quality and leaf beetle performance. Plants grown in flooded treatments had higher water
content and lower tannic acid concentration. All other traits were not significant. Consistent with field
observations, leaf beetle fertility and survival were higher on flooded plants. Our data suggest that the
relative effects of topdown (predation) vs. bottomup forces (plant quality) vary along a water level
gradient and may further interact with abiotic conditions.
CAB International 2008
56
Abstracts: Theme 1 Ecology and Modelling in Biological Control of Weeds
An arthropod and a pathogen in combination as
biocontrol agents: how do they shape up?
L. Buccellato, E.T.F. Witkowski and M.J. Byrne
School of Animal, Plant and Environmental Sciences, University of the Witwatersrand,
Private Bag 3, Wits, Johannesburg 2050, South Africa
The South African biological control programme of Crofton weed (Ageratina adenophora) was initiated in 1984 with the release of a stem galling fly, Procecidochares utilis (Diptera: Tephritidae),
and a leaf spot pathogen, Phaeoramularia eupatorii-odorati (1987) (Hyphomycetes: Moniliales). A
postrelease evaluation at three geographically different sites shows that the pathogen is widespread
throughout the country, with up to 95% of plants being infected. However, the severity of infection is
relatively low, with less than 50% of leaves on individual stems infected. The fly is not widespread in
A. adenophora infestations. Where it occurs, 30% of the stems are galled and only 10% exhibit repeated galling. Parasitism and phenological asynchrony with the plant are shown to suppress fly numbers.
Neither the fly nor pathogen, individually or in combination, significantly affects vegetative growth of
the weed. However, low seedling densities at sites with the fly suggest it may reduce the reproductive
potential of A. adenophora. Mechanical removal is shown to reduce the density of adult plants but
promote recruitment of seedlings and is costly. Therefore, further agent selection is recommended for
the control of A. adenophora in South Africa.
Impact of invasive exotic knotweeds (Fallopia spp.)
on invertebrate communities
E. Gerber,1 U. Schaffner,1 C. Krebs,1 C. Murrell1 and M. Moretti2
CABI Bioscience Switzerland Centre, 2800 Delmont, Switzerland
Swiss Federal Research Institute WSL, Sottostazione Sud delle Alpi, 6504 Bellinzona, Switzerland
1
Exotic knotweeds (Fallopia spp.) are considered to be amongst the most serious invasive exotic weeds
in Europe, causing significant damage to native ecosystems. However, with the exception of competitive exclusion of native vegetation, their suggested ecological impact is poorly supported by experimental studies. We investigated the ecological impact of exotic knotweed species in selected areas of
France, Germany and Switzerland. Ten locations were selected along river courses with different levels
of knotweed infestations and invertebrate traps (pitfall and combi traps) were randomly established
in vegetation invaded by exotic knotweed, as well as in vegetation that can potentially be invaded by
knotweed (open and bush-dominated vegetation). Results indicate that invasion by exotic knotweeds
does not only have strong effects on native vegetation but also affects invertebrate communities. The
overall abundance, biomass and diversity of invertebrates in plots invaded by exotic knotweed were
strongly reduced compared with control plots. Exotic plant species are in general introduced without
their specific natural enemies and are also often less palatable to generalist herbivores. In accordance
with this, we found reduced diversity for herbivore invertebrates in knotweed patches. In addition,
a negative effect of exotic knotweeds could also be detected in other trophic groups (e.g. predators,
detritivores).
57
XII International Symposium on Biological Control of Weeds
An experimental test of the importance of climate
matching for biological control introductions
F.S. Grevstad,1 C.E. OCasey,2 M.L. Katz3 and K.H. Laukkenen1
Olympic Natural Resources Center, University of Washington, 2907 Pioneer Road,
Long Beach, WA 98631, USA
2
Washington State University Extension, 2907 Pioneer Road, Long Beach, WA 98631, USA
3
Section of Evolution and Ecology, University of California-Davis, One Shields Avenue,
Davis, CA 95616, USA
1
The best geographic source of biocontrol agents is often assumed to be the region with the best climate
match to the introduced range. However, in practice, it is often more convenient to collect from another
location. As an experimental test of the importance of climate matching, we compared the performance
of four North American sources of the planthopper, Prokelisia marginata, simultaneously introduced
into each of five replicate sites in Willapa Bay, WA, as biological control agents for Spartina alterniflora.
The sources were California, Georgia, Virginia and Rhode Island. The populations were monitored
for 3 years. All four source populations established at most of the sites, but they exhibited significant
differences in both reproduction during the summer and survival over winter. These differences were
largely consistent with predictions from climate matching. The four populations also differed in spring
emergence timing, adult wing morphology and age structure. Results of this study suggest that (1) the
geographic source of biocontrol agents can affect performance and should be carefully chosen and (2)
the use of multiple geographic sources of a biocontrol agent may improve overall performance in cases
where biocontrol is applied in large target region spanning a range of climates.
Effect of climate on biological control: a case study with
diffuse knapweed in British Columbia, Canada
C.A.R. Jackson, J.H. Myers, S.R. White and A.R.E. Sinclair
Department of Zoology, University of British Columbia, 6270 University Boulevard,
Vancouver, BC, Canada V6T 1Z4
With increasingly volatile and extreme weather patterns under global climate change, predicting how
species and ecosystems will respond to these patterns is emerging as an important and rich area of
ecological research. Climate mediates plantinsect interactions and consequently has the potential for
positive or negative effects on biological control systems and the spread of invasive weeds. Observational evidence suggests that a recent dramatic reduction in the density of an invasive knapweed,
Centaurea diffusa, in sites in the Okanagan Valley of British Columbia, Canada, is attributed to the
biological agent Larinus minutus, a seed head weevil. This dramatic decline took place the same year
in which the area experienced a severe late spring and summer drought. We have examined the effect
of increased precipitation on the strength of the LarinusCentaurea interaction. We conducted field
experiments using constructed rain shelters and weekly watering treatments to assess the effectiveness
of plant attack and seed destruction by Larinus minutus under moist and dry conditions. The results of
these experiments suggest that insect attack is greater under dry conditions.
58
Abstracts: Theme 1 Ecology and Modelling in Biological Control of Weeds
The IRA and getting the result you want
M.K. Kay
Ensis, 49 Sala Street, Rotorua 3012, New Zealand
The current quagmire of plant defence theories does little to assist the practitioners of classical biological control in the assessment, selection and interpretation of host feeding trials. The plethora of
defence theories probably arose from a reductionist experimental approach across a wide variety of
plant growth forms. Recent risk assessments of invasive species in forest ecosystems have offered
a macroecological explanation of insectplant interaction that is contrary to accepted plant defence
hypotheses. The Island Resource Allocation (IRA) hypothesis provides a strong insight into how ecosystems function and as such can successfully predict the palatability of plants within genera, the
metapopulation of species and plant life stages. The IRA offers some predictors for the likely outcome
of classical biological control of target species in different ecosystems.
Microclimate effects on biological control:
water hyacinth in South Africa
A.M. King,1 M.P. Hill,2 M. Robertson3 and M.J. Byrne1
School of Animal, Plant and Environmental Sciences, University of the Witwatersrand,
Johannesburg, South Africa
2
Department of Entomology, Rhodes University, Grahamstown, South Africa
3
Department of Zoology, University of Pretoria, Pretoria, South Africa
Low temperatures are repeatedly blamed for limiting the establishment of new biological control
agents. Where successful establishment does take place, adequate control is often not achieved because of insufficient thermal accumulation. This scenario prevails in most water hyacinth biological
control systems around South Africa. Microclimate was measured at 14 water hyacinth infestations
distributed throughout South Africas climatic regions. Water temperature, air temperature within the
weed canopy and ambient air temperature were recorded at half-hour intervals for 24 months. Correlated with monthly field measures encompassing a variety of plant and insect parameters, this high-data
resolution allowed for accurate descriptions of the seasonality and respective population dynamics
prevalent in the system. Current literature on the thermal physiology of both Neochetina weevils, however, did not match their developmental pattern recorded in the field. Different climatic regions were
found to be distinct in terms of both plant and insect phenology, population size, structure and growth
rate, insect damage and therefore the subsequent levels of biological control achieved. Microclimate
data provided a more accurate prediction of establishment that is otherwise too coarse when modelled
on more broad scale climatic data.
59
XII International Symposium on Biological Control of Weeds
Habitat analysis of the rush skeleton weed root moth,
Bradyrrhoa gilveolella (Lepidoptera: Pyralidae)
J.L. Littlefield,1 G.P. Markin,2 J. Kashefi3 and H.D. Prody4
1
Department of Land, Resources and Environmental Sciences, Montana State University,
Bozeman, MT 59715, USA
2
USDA Forest Service, Rocky Mountain Research Station, Forestry Sciences,
Montana State University, Bozeman, MT 59715, USA
3
USDAARS European Biological Control Laboratory, Thessaloniki, Greece
4
Formerly from the Department of Land, Resources and Environmental Sciences,
Montana State University, Bozeman, MT 59715, USA
The root feeding moth, Bradyrrhoa gilveolella (Treitschke) (Pyralidae) has been recently introduced
into western United States for the control of rush skeleton weed, Chondrilla juncea L. (Asteraceae).
Previous attempts to establish this moth in other countries, e.g. Australia and Argentina, have failed.
Based on life history studies of the moth and habitat types at collection sites in Europe, we hypothesize
that habitat will be a critical factor in successfully establishing the moth in North America. We surveyed 19 rush skeleton weed sites in northern Greece and southern Bulgaria, with and without populations of Bradyrrhoa. We compared these with release sites and potential release sites located in Idaho,
USA. Multivariate analysis of site characteristics, vegetation and soil properties was used to investigate similarities amongst sites. Because of the low number of sites with the presence of Bradyrrhoa,
it was difficult to discern distinct habitat differences. Soil texture appears to be the most important site
factor common with sites with moth populations.
Evaluating the performance of Episimus utilis
(Lepidoptera: Tortricidae) on the invasive Brazilian
peppertree in Florida
V. Manrique,1 J.P. Cuda,2 W.A. Overholt3 and D. Williams4
Indian River Research and Education Center, University of Florida, 2199 South Rock Road,
Fort Pierce, FL 34945, USA
2
Department of Entomology and Nematology, University of Florida, Building 970, Natural Area Drive,
Gainesville, FL 32614, USA
3
Indian River Research and Education Center, University of Florida, 2199 South Rock Road,
Fort Pierce, FL 34945, USA
4
Marine Genomics Laboratory, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA
1
Brazilian peppertree, Schinus terebinthifolius Raddi (Anacardiaceae), an introduced perennial tree
from South America, has become widely established throughout central and south Florida because
of its ability to invade a wide range of habitats from disturbed sites (e.g. along highways, canals) to
natural communities (e.g. pinelands, mangrove forests). Genetic studies have identified two chloroplast DNA haplotypes of Brazilian peppertree in Florida that come from two genetically differentiated
source populations in Brazil. Haplotype A is more common on the west coast of Florida, whereas haplotype B is more common on the east coast. In addition, hybridization between these two introduced
populations has occurred extensively in Florida. A leaf roller from Brazil, Episimus utilis Zimmerman
(Lepidoptera: Tortricidae), has been selected as a potential biological control agent against Brazilian
peppertree in Florida. The objectives of this study were to evaluate the performance of E. utilis on different Brazilian peppertree genotypes and plants subjected to different environmental conditions found
in Florida (e.g. saline vs. fresh environments, soil fertility and soil moisture content). The ecological
significance of the results is discussed in the context of predicting suitable sites for field releases to
increase the possibility of establishment and subsequent effectiveness of this candidate biological control agent.
60
Abstracts: Theme 1 Ecology and Modelling in Biological Control of Weeds
Successful biological control of diffuse knapweed in
British Columbia, Canada
J.H. Myers, H. Quinn, C.A.R. Jackson and S.R. White
Departments of Zoology and Agroecology, 6270 University Boulevard, University of British Columbia,
Vancouver, BC, Canada V6T 1Z4
In the year of the first Biological Control of Weeds Symposium, the first introductions of insects were
made in an attempt to control diffuse knapweed in Canada. In the next 23 years, 12 species of agents
were introduced into Canada on this weed. Finally, beginning in 2000 and after the last insect species
to be introduced, Larinus minutus, diffuse knapweed populations have declined dramatically in areas
with high Larinus beetle densities. To determine if this decline was simply the result of drought, we
have continued to monitor knapweed densities with the return of normal precipitation. Knapweed
densities have not rebounded in areas with L. minutus but have resurged in areas without beetles and
in the presence of other agents, Urophora spp. and Sphenoptera jugoslavica. To determine if attack by
L. minutus is sufficient to reduce knapweed densities without other species of control agents, we have
studied their impact in areas where knapweed has resurged after the eradication of all agents by wildfires. Cage experiments demonstrate the impacts of L. minutus on diffuse knapweed with and without
S. jugoslavica. Preliminary results indicate a strong impact of L. minutus and their rapid reinvasion to
newly infested knapweed sites.
An integrated approach to invasive plant management:
biocontrol and native plant interactions
J.G. Nachtrieb,1 M.J. Grodowitz,2 R.M. Smart3 and C.S. Owens4
University of North Texas, U.S. Army Engineer Research and Development Center, Environmental Laboratory,
Lewisville, TX, USA
2
US Army Engineer Research and Development Center, Environmental Laboratory, Vicksburg, MS, USA
3
US Army Engineer Research and Development Center, Environmental Laboratory, Lewisville, TX, USA
4
SpecPRo, Inc, US Army Engineer Research and Development Center, Environmental Laboratory,
Lewisville, TX, USA
Native aquatic macrophytes are known as effective competitors against invasive aquatic plants. Yet, the
invasive aquatic plants have an inherent competitive edge due to a lack of herbivores. Hence, the presence of diverse native plant assemblages coupled with sustained herbivory should hasten declines. To
test this hypothesis in the aquatic environment, several experiments were conducted with and without
plant competition and herbivory using insecticides to eliminate herbivores. In a 2-year study combining hydrilla biocontrol with native plant competition, overall tuber production was reduced 2-fold by
native plant competition and 1.3-fold by herbivory alone. However, even more substantial decreases
of more than fivefold were demonstrated when both were combined, stressing the importance of plant
competition when found in conjunction with herbivory. Throughout this and other studies, native
plants exhibited significant levels of damage due to invertebrate herbivory. However, little information is available that quantifies the impact of herbivores on native plants. In a second study examining
native herbivore impacts to submersed and floating leaf species, only the two floating-leaved species,
Potamogeton nodosus and Potamogeton illinoensis exhibited a 37% and 72% decrease in biomass in
the presence of herbivory, respectively. Hence, competitive edge in native plants is apparently curtailed
significantly because of herbivory.
61
XII International Symposium on Biological Control of Weeds
Impact of host-plant water stress on the interaction
between Mecinus janthinus and Linaria dalmatica
A.P. Norton
Department of Bioagricultural Sciences and Pest Management, Colorado State University,
Fort Collins, CO, USA
Mecinus janthinus (Coleoptera: Curculionidae) is a stem-boring weevil that has been introduced as
a biological control agent for Dalmatian toadflax (Linaria dalmatica L). We examined the impact of
host-plant water stress on plant performance, weevil choice, weevil oviposition rates and survival.
Weevils preferred well-watered plants and plants previously attacked by other females. The weevil had
greatest impact on toadflax fruit production in high-water treatment plants but had greatest impact on
stem production and plant weight in the intermediate water treatment. None of the plant performance
variables measured significantly responded to weevil feeding in the low-water treatment. However,
beetle survivorship increased with increasing water supplied to the plant. These results indicate that the
impact of this biological control agent on its host plant is context-dependent and varies with the amount
of water available to the plant. Mecinus prefers and performs better in plants that receive more water,
yet these plants are also less susceptible to weevil attack. These data suggest that studies of plant quality and herbivore impact need to include herbivore preference and performance to generate a complete
picture of the interaction.
Impact of insect herbivory on dispersal in
Hydrilla verticillata (L.f.) Royle
C.S. Owens,1 M.J. Grodowitz2 and R.M. Smart3
SpecPro, US Army Engineer Research and Development Center, Lewisville, TX, USA
US Army Corps of Engineers Environmental Research and Development Center, Vicksburg, MS, USA
3
US Army Engineers Research and Development Center, Lewisville, TX, USA
1
Hydrilla is an invasive aquatic plant that spreads through a variety of vegetative structures. Fragments,
long-distance dispersal and vegetative reproductive propagules, increase the ability of Hydrilla to colonize new and distant locations. Sustained levels of herbivory by leaf-mining flies (Hydrellia pakistanae
and Hydrellia balciunasi) can reduce biomass and impact the ability of hydrilla to photosynthesize. Impacts from herbivory apparently can increase fragmentation at points of feeding. If greater fragmentation occurs because of increased fly damage, how viable are these fragments? To answer this question,
viability of fragments with low (030%), medium (4060%) and high (70100%) leaf damage were
compared. Fragments with high levels of leaf damage produced three times less biomass when compared with hydrilla fragments with low leaf damage. Fragment establishment was also studied based
on settling, rooting and anchoring success of individual fragments. Results indicate that no highly damaged fragments settled, produced roots or anchored. However, more than 80% of the control fragments
produced roots and anchored, as compared with only 20% by low- and medium-damaged fragments.
Herbivory apparently has great impact, significantly reducing fragment viability.
62
Abstracts: Theme 1 Ecology and Modelling in Biological Control of Weeds
Dynamics of invasive plant monocultures after the
establishment of natural enemies: an example from the
Melaleuca quinquenervia system in Florida
M.B. Rayamajhi, P.D. Pratt, T.K. Van and T.D. Center
USDAARS, Invasive Plant Research Laboratory, Fort Lauderdale, FL 33314, USA
Melaleuca quinquenervia (melaleuca), a native tree of Australian origin, has become one of the most
invasive plants in Florida. Biological control was implemented as a long-term solution to melaleuca control in Florida. Now several natural enemies (insects and a pathogen) of melaleuca are wellestablished in Florida. Their impact on melaleuca populations is being monitored for several years.
Insectfungus integration in field conditions showed significant impact on stump regrowth control of
melaleuca. During the period when natural enemies prevailed, melaleuca stand density and basal area
declined at a greater rate across study sites. Myrtaceae (represented by melaleuca) and Cyperaceae
(represented by saw grass, Cladium jamaicense) were the first and second most important families during 1997, respectively. Their family importance values were reduced significantly during 2004/05, with
the increase of the importance values of other families in the stand. Overall plant richness and diversity
in melaleuca stands was higher (compared with 1997) in 2004/05. Within stands, plant species richness
and diversity indices were relatively higher at interior than at peripheral sections. Melaleuca-vacated
spaces in the stands were colonized mostly by native plant species. This phenomenon is predicted to
intensify as the natural enemies continue to impact melaleuca monocultures in Florida.
Modelling of Diorhabda elongata dispersal during
the initial stages of establishment for the
control of Tamarix spp.
J. Sanabria,1 C.J. DeLoach,2 J.L. Tracy2 and T.O. Robbins2
Texas A&M University, Blackland Research Center, 720 East Blackland Road, Temple, TX 76502, USA
2
USDAARS, 808 East Blackland Road, Temple, TX 76502, USA
Critical questions associated with use of Diorhabda for control of Tamarix are the following: How far
and how fast will the beetle defoliate salt cedar trees after the insects release? What are the factors
that affect the dispersal of the insect and the salt cedar defoliation? One effective way to answer these
questions is modelling dispersal of the beetle and the defoliation that it causes. Modelling strategies
depend on whether the Diorhabda is in the initial stages of establishment or in later stages when it is
already well-established. The diffusion model developed by Kovalev was fit to observed data at Big
Spring, TX, in 2005 and 2006. Kovalevs model was able to fit satisfactorily the data of larval density.
When Kovalevs waves are not symmetric, it is mainly because of unequal distances between transect
quadrants, which happens in areas of nonuniformly distributed salt cedar. The modelled larval dispersal front reached 100 m in transect 2 during 2005 and 510 m in 2006. The defoliation front, associated
with the high density of larvae, reached between 60 and 80 m in 2005 and about 240 m in 2006. Fitting
statistical models to data collected is in progress.
63
XII International Symposium on Biological Control of Weeds
Seed feeders: why do so few work and can we
improve our selection decisions?
R.D. van Klinken, R. Colasanti and G. Maywald
CSIRO Entomology, 120 Meiers Road, Indooroopilly, Brisbane, QLD 4068, Australia
Seed feeders have a mixed record in weed biological control. They are typically easy to test, rear,
release and establish and can be particularly useful when targeting beneficial plants. However, seed
predation rates are often relatively low. Also, evaluating the impacts of seed predation on the population dynamics of the plant, particularly on rates of spread, is difficult. In this paper, we present the
results from the field assessment and modelling of a seed-feeding bruchid (Penthobruchus germaini)
on the woody legume Parkinsonia aculeata. The seed feeder is typically abundant, and both the target
and the seed feeder occur across diverse climate zone (from the wetdry tropics to the arid interior)
and habitats (including upland and wetlands). In particular, we (1) determine seed predation rates and
identify factors that were limiting seed predation rates and impacts through national field surveys and
(2) estimate the effect of seed predation on rates of population increase (using DYMEX, a simulation
population model) and rates of spread (using a cellular automata model). The modelling work and field
evaluations both highlighted general factors that may limit the impact of seed feeders on plant populations. We discuss whether agent selection decisions involving seed-feeding insects can be improved to
identify potential agents that will have the greatest impact and avoid releasing ineffective agents.
64
Theme 2:
Benefit/RiskCost Analyses
Session Chair: Ernest (Del) Delfosse
65
This page intentionally left blank
Keynote Presenter
Return on investment: determining
the economic impact of biological
control programmes
R. McFadyen1
Summary
In >100 years of weed biological control, few economic impact assessments of biological control
programmes have been undertaken, and all were successes. Yet biological control is still largely paid
for by governments, who need proof of the return on their investment. Cost/benefit analyses can also
be used to rank biological control against other management methods. A recent economic impact assessment of all weed biological control undertaken in Australia since 1903, including both successes
and failures, demonstrated annual benefits of $95.3 million from an average annual investment of
$4.3 million (Aus$, 2005 values), a cost/benefit ratio of 23:1. Even with the enormous economic
impact of the prickly pear success excluded, the cost/benefit ratio of all other programmes was 12:1.
The benefit came from 17 successful programmes: two, which are usually considered failures, in
fact returned strongly positive benefits because small reductions in the weed problem nevertheless
resulted in considerable cost savings.
The scarcity of economic studies has many causes: long period from commencement to full field
results; difficulties in assigning monetary values to biodiversity and social impacts; and difficulties in
assessing impacts of biological control. The Australian study demonstrated the economic returns from
partial successes, where these reduce the costs of other management methods. It also demonstrated
the importance of obtaining baseline economic data before starting biological control and at intervals
during the agent release period. Seeking advice from economists at all stages of a biological control
programme must become as routine as consulting statisticians.
Keywords: weed biological control, cost/benefit, success rates, Australia.
Introduction
20 reported impacts at plant population level, but only
three analysed the economic impact.
The 1996 Symposium had a session on Evaluation
and Economics with three papers, of which only one
(Coombs et al., 1996) analysed the economic benefits,
while the 2000 Bozeman Symposium had no papers
on economic analysis. The 2003 Symposium session
on Risk Analysis treated this purely as biological risks
with no mention of economic risks. However, Stanley
and Fowler (2004) stated that economic analyses are
an important part of resolving conflicts of interests
and Decision-makers often find arguments couched
in monetary terms to be more convincing. The theme
Evaluation included a report of a long-term study
showing massive reduction in thistles in rangelands in
the western United States (Kok et al., 2004), and other
papers demonstrating impact on weed populations, but
A check of the proceedings of the last three international symposia, plus the main biological control journals, confirmed that although many evaluation studies
have been published, few included any economic data.
Doeleman (1990) on the biological control programme
against salvinia Salvinia molesta DS Mitchell and Jarvis et al. (2006) on scotch broom Cytisus scoparius
(L.) Link were among the few exceptions. Dhileepan
(2003) listed published evaluation studies from Australia, of which 32 reported impacts at plant level and
1
Cooperative Research Centre for Australian Weed Management, Block
B, 80 Meiers Road, Indooroopilly, QLD 4068, Australia <rachel.
[email protected]>.
CAB International 2008
67
XII International Symposium on Biological Control of Weeds
with no economic data. The only economic assessment
was of the control of the waterweed Azolla filiculoides
Lamarck in South Africa (McConnachie et al., 2003).
A January 2004 conference in the United States discussed benefits and risks of biological control in order
to address critical issues facing policy makers. However, no economic benefit/risk analyses or consideration of return on investment were included. There were
20 papers dealing with risks and how to predict and
reduce these, but only one on the economic framework
for decision-making in biocontrol (Jetter, 2005). In
summary, although there is an increasing recognition
of the need to demonstrate impact on weed populations
and native biodiversity (e.g. Coombs et al., 2004; Story
et al., 2006; Barton et al., 2007), economic impact
analyses remain very much the exception.
son, it is important to include failures in the economic
analysis, as only this can give real estimates of success
rates while also including benefits from partial successes, which otherwise may be seriously undervalued.
For weed biological control, there is therefore a need
to measure all research and research-related costs from
all programmes undertaken, and to assess this against
all benefits gained. A similar analysis was undertaken
for all programmes of the Australian Centre for International Agricultural Research (Raitzer and Lindner,
2005), and could be done for weed biological control
on a country or state basis. Only in this way can we
calculate the true probability of a positive return on investment. Selected case studies where failures are not
included (van Wilgen et al., 2004) are useful but do not
help assess the probability of failure.
This overall approach will always underestimate
benefits, because the costs are already realised and
are comparatively easy to determine, while benefits
can only be calculated where both the baseline data to
measure change has been collected and the impact assessments have been done. Benefits may also continue
to increase after the assessment period; hence, the approach is inherently conservative.
Value of economic impact assessment
Most weed biological control is funded by public money in one form or another and thus competes
for funding with other government programmes. The
demonstrated cost-effectiveness of a control method is
a reason for continuing or increasing both the use of
the method and the resources dedicated to it (Culliney,
2005; Lodge et al., 2006). For example, net economic
benefits from pre-introduction screening of plant imports is a justification both for major policy change in
the United States and for the maintenance of expensive
border controls in Australia (Keller et al., 2007), and
McConnachie et al. (2003) highlighted the use of cost/
benefit analyses to rank biological control against other
means of control.
Syrett, Briese and Hoffmann (2000), in their excellent overall summary of economic evaluation in weed
biological control, make several important and stillvalid statements. For example, it is important to know
how successful the technique is overall. If a high proportion of programmes are successful, then the likelihood of a new programme being successful is relatively
high too. Jetter (2005) pointed out that policy decisions
are based on economic criteria, which in turn depend
on the probability of success. Increasingly, risk assessment (or riskbenefitcost assessment) is being applied
worldwide to most activities. This requires identification of hazard and benefits, and then exposure analysis
(quantitative assessment of probabilities/likelihood)
(Sheppard et al., 2003). It is therefore essential to know
the probability that the programme will result in a positive economic outcome (i.e. economic return will exceed economic costs), which allows the calculation of
the risk-weighted return of an individual investment.
This can then be used to prioritise investments across
the entire portfolio. Note that this is not probability of
complete success but the probability of a positive return on investment.
In any risk assessment, probabilities are calculated
from historic data for similar activities. For this rea-
Economic impact assessment of
Australian weed biological
control programmes
This section is a summary of results from the publication
by Page and Lacey (2006) of the consulting firm AEC
Group Ltd. Their study aimed to determine the cost of
all biological control research undertaken in Australia
since the first prickly pear programme in 1903, and
compare this against the total benefits received. The
work was funded by the Cooperative Research Centre
for Australian Weed Management (Weeds CRC) as part
of our Biological Control Evaluation project. Data were
collected by Colin Wilson, with assistance from many
Weeds CRC staff. The direct cost of this project was
Aus$85,000 plus many uncosted hours of scientist time.
Methods
Only economic costs and benefits were considered:
other benefits were listed but not included in the analysis. Costs of current programmes were excluded where
it was still too early for field impacts to be realised.
However, completed programmes where no agents
were released or established were counted as failures
and the costs included in the analysis.
The first step was to assemble previously undertaken
economic analyses and convert these to current values.
All economic data throughout the study were converted
to and are cited as 2005 Aus$ values. Available ex-ante
costbenefit analyses were recalculated in the light of
actual results. The next step was to list all weed biolog68
Return on investment: determining the economic impact of biological control programmes
ical control programmes undertaken, including those
that did not lead to releases, and assemble cost data
wherever possible. Data on the costs of research and
releases were usually available from internal or published reports and records. Where data were unavailable or incomplete, the duration of the research in years
and the number of staff employed (usually known from
reports) was used to calculate costs, using a factor of
Aus$300,000 per scientist-year whether employed in
Australia or overseas.
The most difficult part proved to be locating data on
the economic impacts of the weeds prior to biological
control. Despite careful searches of the literature, including the original reports used to justify biological control programmes, all too often there was little or no data
on the value of losses due to the weed or on the cost of
control, and little quantitative information on the extent
of the weed or the rate of spread. However, all available
information was collated and converted to 2005 Aus$
values. Independent economic data from the Australian
Bureau of Statistics (ABS) Australian Farm Surveys
were used whenever available, but ABS surveys did
not include questions on the cost of weeds until 2006;
therefore, comparable national data was not available.
The biological control programmes against the three
floating water weeds salvinia, water hyacinth Eichhornia crassipes (Mart.) and water lettuce Pistia stratiotes
L. were considered as one programme, with both costs
and benefits summed into one amount. These weeds
occupy the same water surfaces and each is able, if not
controlled, to occupy 100% of water surfaces. Therefore, the benefits only accrue if all three are simultaneously controlled, which was achieved between 1975
and 1995.
The long and expensive programme on Noogoora
burr (Xanthium occidentale Bertol.), with overseas exploration from 1930 to 1970, resulted in the establishment of three insects but no impact on the weed. A rust
disease Puccinia xanthii Schw. was studied in France
but not introduced on the grounds that it was insufficiently host-specific. In 1974 the same rust turned up in
Brisbane and in north Queensland (Morin et al., 1996),
almost certainly illegally imported by landholders in
contact with the research group. The rust spread rapidly and, together with a moth released against parthenium weed (Parthenium hysterophorus L.), has given
almost complete control of Noogoora burr over most
of the affected lands. As the introduction of the rust,
although illegal, was a direct consequence of the biological control investigations, the benefits have been
included, as economically damaging consequences,
whether intended or not, have also been included in
all analyses.
Information on the biological control impacts was
obtained from publications, unpublished reports and
personal communications from scientists involved in
the projects. Proper evaluations had not been undertaken for most programmes, even for successful ones.
Where agents had established, information on spread
and impact was often very incomplete, as was information on impact on the weed population. Too often, there
was only expert opinion, expressed in percent reduction of the weed population, percent of the total area
of weed infestation affected, and percent of years with
maximal impact.
Results
Results are presented in Table 1 (summary of Table
3.1 in Page and Lacey, 2006). A total of 36 weed biological control programmes in Australia have been undertaken in the 100 years to 2004, but for three of these
(against crofton weed Ageratina adenophora (Sprengel), St Johns wort Hypericum perforatum L. and
docks Rumex spp.), although the biological control has
been largely successful, key data on costs or benefits
could not be obtained. There were existing economic
cost/benefit analyses for eight programmes (e.g. Adamson and Bray, 1999), though some (e.g. Nordblom
et al., 2002) were ex-ante studies, and actual realised
benefits had never been assessed.
Of the 33 programmes with economic data, no agents
were ever released or established for five of these. Economic benefits were zero in a further four where established agents have, to date, proven ineffective. All
remaining programmes returned economic benefits.
In three (alligator weed Alternanthera philoxeroides
(Martius), groundsel bush Baccharis halimifolia L. and
Sida spp.), although successful control of the weed was
achieved, the benefit/cost ratio was less than 1 because
the direct economic impact of the weed was small. For
some others where the benefit/cost ratio is <1, the benefits are still increasing and the assessments may be
premature.
Seventeen programmes out of the 33 assessed returned a positive return on investment; i.e. economic
benefits exceeded costs. Thirteen of these have resulted
in very large economic benefitsthe prickly pear
(Opuntia spp.) programme, but also those against ragweed Ambrosia artemisiifolia L., nodding thistle Carduus nutans L., skeleton weed Chondrilla juncea L.,
rubber vine Cryptostegia grandiflora R. Br., Patersons
curse Echium plantagineum L., harrisia cactus Harrisia martinii (Labouret), giant sensitive plant Mimosa
diplotricha C. Wright, Onopordum thistles, parthenium weed, the three water weeds, ragwort Senecio
jacobaea L., and Noogoora burr. Surprisingly, the
programmes against lantana Lantana camara L. and
blackberry Rubus fruticosus L. agg., both generally
considered to be still unsuccessful, returned positive
cost/benefit ratios because the economic losses from
these two weeds are so great that relatively small reductions in these losses are worth a great deal of money.
On the other hand, the successful control of bridal
creeper Asparagus asparagoides (L.) did not result
in large economic gains because it is almost entirely
69
XII International Symposium on Biological Control of Weeds
Table 1.
Total costs (converted to Aus$ million 2005 values), benefits (net present value, Aus$ million), and benefit/cost
ratios for 33 Australian weed biological control programmes.
Weed species
Senecio madagascariensis
(fireweed)
Senna obtusifolia (sicklepod)
Emex australis & E. spinosa
(spiny emex/double gee)
Heliotropium europaeum
(common heliotrope)
Lantana montevidensis
(creeping lantana)
Acacia nilotica (prickly
acacia)
Ageratina riparia (mistflower)
Parkinsonia aculeata
(parkinsonia)
Silybum marianum & Cirsium vulgare (variegated and
spear thistle)
Marrubium vulgare
(horehound)
Alternanthera philoxeroides
(alligator weed)
Chrysanthemoides monilifera
(bitou bush/bone seed)
Prosopis spp. (mesquite)
Sida acuta & S. rhombifolia
(Paddys lucerne)
Baccharis halimifolia
(groundsel bush)
Mimosa pigra (mimosa)
Asparagus asparagoides
(bridal creeper)
Rubus fruticosus agg.
(blackberry)
Xanthium occidentale
(Noogoora burr)
Lantana camara (lantana)
Carduus nutans
(nodding thistle)
Parthenium hysterophorus
(parthenium weed)
Onopordum spp. (Scotch,
stemless & Illyrian thistles)
Carduus pycnocephalus &
C. tenuiflora (slender thistles)
Mimosa diplotricha (giant
sensitive plant)
Harrisia spp. (Harrisia
cactus)
Salvinia molesta, Eichornia
crassipes, Pistia stratiotes
(waterweeds)
Senecio jacobaea (ragwort)
Echium plantagineum
(Patersons curse)
Yrs
19901994
Cost
0.4
Benefits NPV
0
Benefit /cost
0
19922000
19741978
0.7
2.0
0
0
0
0
No agents released
No agents established
19731991
4.4
No agents established
19962000
0.2
No agents established
1980
5.3
19862001
0.2
19831990
1.6
19882002
3.0
Agents established but
ineffective
Agent established but
ineffective
Agents established but
largely ineffective
Agents established but
ineffective
19892001
1.8
0.3
0.2
Benefits still increasing
19762004
1.4
0.3
0.4
Social benefits not costed
1990
7.1
2.7
0.5
19922004
19841999
2.3
4.2
0.8
1.1
0.5
0.5
Agents established but
largely ineffective
Benefits still increasing
19611998
9.6
2.1
0.7
1981 2004
1990 2004
21.6
7.3
6.1
8.2
0.8
2.0
19772004
4.9
6.1
2.5
19301975
10.1
23.4
19142004
13.6
3.0
5.6
81.3
6.9
Benefits largely from first
agents
Data from earlier study
Benefits still increasing
19862000
n/a
19772004
11.0
38.6
7.2
19882004
3.7
20.1
9.6
19871997
2.1
22.5
14.1
19821992
1.7
21.4
18
19591976
1.0
19.4
23.5
19741993
5.1
79.4
27.5
19772004
19722004
7.9
97.2
1,201.3
32.4
52.0
70
Comment
No agents released
Changed land use reduced
benefits
Benefits still increasing
Very large biodiversity
benefits
Benefits still increasing
Data from earlier study
Additional uncosted
biodiversity benefits
Benefits still increasing
Data from earlier study
Return on investment: determining the economic impact of biological control programmes
Table 1.
(Continued) Total costs (converted to Aus$ million 2005 values), benefits (net present value, Aus$ million), and
benefit/cost ratios for 33 Australian weed biological control programmes.
Weed species
Ambrosia artemisiifolia
(annual ragweed)
Cryptostegia grandiflora
(rubber vine)
Chondrilla juncea (skeleton
weed)
Opuntia spp. (prickly pears)
Yrs
Cost
Benefits NPV
Benefit /cost
19851991
0.6
52.5
103.7
19842004
3.6
234.6
108.8
19031987
21.1
1,425
112.1
3,110.3
312.3
Comment
Additional uncosted biodiversity benefits
Data from earlier study
Shaded data are from earlier studies, not included in overall benefit/cost ratio.
Three programmesAgeratina adenophora (crofton weed), Hypericum perforatum (St Johns wort), Rumex spp. (docks)were excluded for
lack of economic data, but all resulted in significant control of the weed.
a weed of natural ecosystems and the direct financial
costs of the weed are small.
Costs also varied greatly. Some programmes continued over decades, with years of overseas research
and the employment of several scientists; the most
expensive being that against Mimosa pigra L., with a
total cost of Aus$21.6 million. The cheapest successful
programme was against annual ragweed at a cost of
Aus$0.6 million. This cost was low because the successful agents were imported for the control of the closely
related parthenium weed, and the only additional costs
were for extra releases in ragweed areas. Two other
insects from the US were tested, but this was undertaken alongside another major project and costs were
minimal. However, median cost for the 17 successful
programmes (including bridal creeper, groundsel bush,
and blackberry) was Aus$7 million and the duration 14
to 27 years. It is unrealistic to expect good results for
smaller investments (cf Coombs et al., 2004).
As was expected, benefits from the control of the
prickly pears were enormous. For simplicity, only data
from the statistical area of the Darling Downs in southern Queensland were used in the calculations. This is
a large and rich agricultural area which was almost
totally unusable due to prickly pear until after the impact of cactoblastis in the early 1930s. Benefits from
infested land in central and north Queensland and in
New South Wales have not been included; their inclusion might double the measured economic return. In
the calculations, economic benefits from land under
cultivation were considered to have ceased in the 1960s
because the larger machines available from the 1950s
can handle even land densely infested with prickly pear.
However, in land used for grazing, there is still no other
economic control for prickly pears. Social and medical
benefits from reduction in injuries and infections due to
spines were not considered.
The overall benefit/cost ratio was 23.1 for the 28 programmes where data could be analysedan astonishing
result. Even if the iconic prickly pear success is excluded, the overall benefit/cost ratio is 12.3. Out of the
total 36 programmes (including those where economic
analysis was not possible), only nine were failures, with
few or no economic benefits.
Issues for the future
There are two major messages from this study: the
large benefits from even partial control of major and
widespread weeds such as lantana and blackberry, and
the overwhelming importance of documenting the economic costs of the target weed at the start of a biological control programme. The key issue is to quantify the
economic costs of the target weed at the start, so that
the benefits from any reduction in its abundance can in
turn be quantified.
Therefore, the first step, unfortunately often omitted,
is a good economic assessment of the economic losses
caused by the weed; e.g. for Melaleuca quinquenervia
(Cav.) in Florida (Diamond and Davies, 1991; Turner
et al., 1998) and for Tamarix spp. in the United States
[Zavaleta (2000) and reviews by Culliney (2005) and
Coombs et al. (2004)]. Ideally, this information should
be part of an ex-ante benefit/cost study prior to starting
any biological control programme. This is best done by
independent economists and made part of the decision
process (Greer and Sheppard, 1990; Jarvis et al., 2006).
Key principles are transparency; i.e. key assumptions,
data sources, and data treatment must be clearly described and explained; and analytical rigour in the use
of the data (Raitzer and Lindner, 2005). If the cost basis
used is made explicit, studies can be critiqued by others
and future updating undertaken. For example, a study
on the productivity benefits from parthenium biological control (Adamson and Bray, 1999) used a range of
80 to 120 cents per kg for the price of cattle. Within
three years, market prices more than doubled to 200 to
400 cents per kg, and the economic benefits increased
accordingly.
As part of an ex-ante analysis, the question must be
asked: If the target weed is removed, would yield losses/weed control costs be reduced? If the target weed
would be replaced by other weeds and control costs
would not fall, then a biological control programme
71
XII International Symposium on Biological Control of Weeds
should not be started. If, however, control of the replacement weeds would be cheaper or yield losses be
reduced, there would still be an economic benefit. For
example, in Indonesia and elsewhere, the weed chromolaena, Chromolaena odorata (L.), often replaced
other weeds including lantana, which then return once
chromolaena has been controlled. However, control
costs (or production losses) for chromolaena greatly
exceed those from lantana; hence, there is still a benefit from its successful biological control even if lantana subsequently re-invades. Other variables can also
influence costs; for example, chromolaena is a weed
in oil palm plantations in both Indonesia and Papua
New Guinea, but in Indonesia, the oil palms are much
more densely planted so that the weed is shaded out in
mature plantations and control is only needed for five
years. Costs are therefore much greater in Papua New
Guinea where control is needed for the full 25-year life
of the plantation. With weedy trees in South Africa,
70% of the losses are due to reductions in water flows
(van Wilgen et al., 2004), which would not apply in
other countries with different native floras or different
hydrology.
A related issue is the level of proof required. Detailed scientific studies can be used to demonstrate impact of agents in small-scale studies (e.g. Dhileepan,
2001) but this does not demonstrate landscape-scale
reduction in weed impacts. Alternatively, independent
economic measures, or properly conducted end-user
surveys (e.g. Ireson et al., 2007) may demonstrate cost
reductions over several years, but with only correlative
evidence that this is due to the impacts of biological
control. Ideally, there should be both: detailed in-field
evaluations using agent exclusion methods on plots
planted to obtain pre-control levels of the weed, demonstrating increased production and/or reduced control
costs, and supported by independent evidence of enduser cost savings.
However, it is equally acceptable, and much cheaper,
to start with end-user results, such as the reduction in
livestock deaths used by Coombs et al. (1996) in their
assessment of the impact of the ragwort biological control programme in Oregon. They used the dramatic fall
over a 20-year period in livestock deaths from pyrrolizidine alkaloid poisoning at local veterinary laboratories as a measure of benefits. Data for the estimate of
programme costs were largely available. They calculated a 15:1 benefit cost ratio at 7% discount, and an
annual benefit of $5 million. Correlative analysis based
on industry-wide data is widely accepted as scientifically valid for other policy decisions such as health
interventions (e.g. Productivity Commission, 2005),
and it is time to recognise that it is equally valid for
biological control.
Quick and dirty or scientific?
There is often a painful choice ineconomic ana
lysesone may choose only two among the three characteristics: fast, accurate and cheap! (T. Nordblom,
2006, unpublished paper). The decision has to be based
on the objective to produce scientific papers in the
highest-impact journals, or to convince policy makers
and national governments? A recent discussion on the
assessment of augmentative biological control (used
in greenhouse and other horticultural crops) deals with
exactly this problemwhether it is necessary to have
scientifically valid studies or whether reduced industry use of pesticides is sufficient (van Lenteren, 2006;
Collier and Steenwyk, 2006). To quote: (the issue is)
what represents convincing evidence, In our view,
the type of evidence that makes a convincing case for
efficacyis quantitative data from replicated field experiments with valid control plots. The data should also
be published in a peer-reviewed journal instead of the
grey literature and implementation [by farmers]
is not equivalent to quantitative data from scientifically
rigorous studies. (Collier and Steenwyk, 2006 p.120).
However, economic analyses for policy purposes
are rarely, if ever, published in peer-reviewed scientific
journals. The recent Stern Review The Economics of
Climate Change first appeared as a government document in 2006, then on their website with comments and
postscript reports, and in 2007 was published as hardcopy by Cambridge University Press. For a more relevant, less-prestigious example, the Raitzer and Lindner
(2005) review was published by ACIAR in their Impact
Assessment Series, put on their website, and is obtainable only from them.
Conclusion
If the overall objective of a weed biological control programme is to reduce the weeds harmful impact, then
Key Performance Indicators (KPI) will measure the degree to which this has been achieved, and assessment
protocols must be designed to measure these KPIs.
This means the initial state prior to the biological control programme must be adequately recorded. For most
weeds, even nonproduction impacts have an economic
aspectif control was cheap or easy, the community
would not permit weeds to take over environmental
areas. It is the economic cost that makes removal and
restoration unviable except for very small areas.
Economic impact is the sum of many factors: loss
of agricultural productivity; extent of infestation, actual and potential; spread rate; cost of removal; and frequency of recurrence. Measurement of these at the start,
even if only on a coarse scale, is essential to make future assessments possible. This is most easily achieved
by developing an ex-ante costbenefit study for each
programme (Coombs et al., 2004). Such analyses immediately clarify where data are not available, as well
as identifying the critical outcomes desired. In a long
programme, the analysis can be repeated after several
years, when more information is available (e.g. whether
72
Return on investment: determining the economic impact of biological control programmes
suitable agents can be found and established; spread rate
of weed). Comparison over several years may identify
significant gains even from unsuccessful programmes
(Hoffmann and Moran, in these proceedings). Ex-ante
studies, using realistic probabilities of success based
on historic rates for the country and type of weed, and
taking into account the full range of potential costs
and benefits, are powerful tools to convince funding
agencies. In time, ex-post analyses based on these data
will clarify the true probabilities of failure, and consequently, return-on-investment for weed biological control. Halfway benefit/cost analysis, when agent spread
and impact are known, can also help decide whether to
spend more resources on increased releases to achieve
faster results (Nordbloom et al., 2002).
The key messages from our study, therefore, were:
biological control is a very cost-effective method and
has given excellent returns-on-investment for Australian governments; indeed, the only better investment in
weed management is expenditure to prevent new incursions; and organizations and governments undertaking
biological control need to ensure adequate economic
data is collected at the start of the programmes and at
intervals throughout.
Oregon State University Press, Corvallis, USA , pp.122
126.
Culliney, T.W. (2005) Benefits of classical biological control
for managing invasive plants. Critical Reviews in Plant
Sciences 24,131150.
Dhileepan, K. (2001) Effectiveness of introduced biocontrol
insects on the weed Parthenium hysterophorus (Asteraceae) in Australia. Bulletin of Entomological Research 91,
167176.
Dhileepan, K. (2003) Evaluating the effectiveness of weed
biocontrol at the local scale. In: Spafford Jacob, H. and
Briese, D.T. (eds) Improving the Selection, Testing and
Evaluation of Weed Biological Control Agents. Technical
Series no. 7, CRC for Australian Weed Management, Adelaide, Australia, pp. 5160.
Diamond, C. and Davies, D. (1991) Economic impact statement: the addition of Melaleuca quinquenervia to the Florida prohibited aquatic plant list. In: Center, T.C, Doren,
R.D., Hofstetter, R.L., Myers, R.L. and Whiteaker, L.D.
(eds) Proceedings of the Symposium on Exotic Pest Plants.
National Parks Service, Denver, USA, pp. 87110.
Doeleman, J.A. (1990) Biological control of salvinia in Sri
Lanka: an assessment of cost and benefits. Economic Assessment Series. Australian Centre for International Agricultural Research, Canberra, Australia.
Greer, G. and Sheppard, R.L. (1990) The economic evaluation of the benefits of research into biological control of
Clematis vitalba. Research Report No 203. Agribusiness
and Economics Research Unit, Lincoln University, Lincoln, NZ.
Ireson, J. E, Davies, J.T., Friend, D.A., Holloway, R.J., Chatterton, W.S., van Putten, E.I. and McFadyen, R.E.C. (2007)
Weeds of Pastures and Field Crops in Tasmania: Economic Impacts and Biological Control. Technical Series
no.13, CRC for Australian Weed Management, Adelaide,
Australia. 78 p.
Jarvis, P.J., Fowler, S.V., Paynter, Q. and Syrett, P. (2006)
Predicting the economic benefits and costs of introducing new biological control agents for Scotch broom Cytisus scoparius into New Zealand. Biological Control 39,
135146.
Jetter, K. (2005) Economic framework for decision making in
biological control. Biological Control 35, 348357.
Keller, R.P., Lodge, D.M. and Finnoff, D.C. (2007) Risk assessment for invasive species produces net bioeconomic
benefits. Proceedings of the National Academy of Sciences
of the United States of America 104, 203207.
Kok, L. T., McAvoy, T.J. and Mays, W.T. (2004) Biological control of Carduus thistles in Virginiaa long-term
perspective, three decades after the release of two exotic
weevils. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lons
dale, W.M., Morin, L.and Scott, J.K. (eds) Proceedings
of the XI International Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia,
pp. 554559.
Lodge, D.M., Williams, S., MacIsaac, H.J., Hayes, K., Leung,
B. et al. (2006) Biological invasions: recommendations
for US policy and management. Ecological Applications
16, 20352054.
McConnachie, A.J., de Wit, M.P., Hill, M.P. and Byrne, M.J.
(2003) Economic evaluation of the successful biological
control of Azolla filiculoides in South Africa. Biological
Control 28, 2532.
Acknowledgements
My thanks to the organizing committee for inviting
me to write this review and to the CRC for Australian
Weed Management who commissioned the report on
which part of this paper is based. Thanks also to Ashley
Page and Kieron Lacey and my Weeds CRC colleagues
for their comments on earlier drafts.
References
Adamson, D.C. and Bray, S. (1999) The Economic Benefit
from Investing in Insect Biological Control of Parthenium
Weed (Parthenium hysterophorus). RDE Connections,
NRSM, University of Queensland, Australia. 44 p.
Barton, J., Fowler, S.V., Gianotti, A.F., Winks, C.J.,
de Beurs, M., Arnold, G.C. and Forrester, G. (2007) Successful biological control of mist flower (Ageratina riparia)
in New Zealand: Agent establishment, impact and benefits
to the native flora. Biological Control 40, 370385.
Collier, T. and van Steenwyk, R. (2006) How to make a convincing case for augmentative biological control. Biological Control 39, 119120.
Coombs, E.M., Radtke, H., Isaacson, D.L. and Snyder, S.P.
(1996) Economic and regional benefits from the biological control of tansy ragwort, Senecio jacobaea, in Oregon.
In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of
the IX International Symposium on Biological Control
of Weeds. University of Cape Town, South Africa, pp.
489494.
Coombs, E.M., Radtke, H. and Nordblom, T. (2004) Economic benefits of biological control. In: Coombs, E.M.,
Clark, J.K., Piper, G.L. and Cofrancesco, A.F. Jr. (eds)
Biological Control of Invasive Plants in the United States.
73
XII International Symposium on Biological Control of Weeds
Morin, L., Auld, B.A. and Smith, J.E. (1996) Rust epidemics, climate and control of Xanthium occidentale. In:
Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of
the IX International Symposium on Biological Control of
Weeds. University of Cape Town, South Africa, pp. 385
391.
Nordblom, T.L., Smyth, M.J., Swirepik, A., Sheppard, A.W.
and Briese, D.T. (2002) Benefit-cost analysis for biological control of Echium spp. (Patersons curse and related
species) in Australia, 19722050. In: Spafford Jacob, H.,
Dodd, J. and Moore, J.H. (eds) Proceedings of the 13th
Australian Weeds Conference. Plant Protection Society of
WA, Perth, Australia, pp. 753756.
Page, A.R. and Lacey, K.L. (2006) Economic Impact Assessment of Australian Weed Biological Control. Technical
Series no.10, CRC for Australian Weed Management, Adelaide, Australia. 151 p.
Productivity Commission (2005) Impacts of Advances in
Medical Technology in Australia, Research Report, Melbourne, Australia.
Raitzer, D.A. and Lindner, R. (2005) Review of the Returns
to ACIARs Bilateral R&D Investments. Impact Assessment Series Report No. 35, August 2005. Australian
Centre for International Agricultural Research, Canberra,
Australia.
Sheppard, A.W., Hill, R., DeClerke-Floate, R.A., McClay,
A., Olckers, T., Quimby, P.C. Jr. and Zimmermann, H.G.
(2003) A global review of risk-benefitcost analysis for
the introduction of classical biological control agents
against weeds: a crisis in the making? Biocontrol News
and Information 24, 91108.
Stanley, M.C. and Fowler, S.V. (2004) Conflicts of interest
associated with the biological control of weeds. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M.,
Morin, L. and Scott, J.K. (eds) Proceedings of the XI International Symposium on Biological Control of Weeds.
CSIRO Entomology, Canberra, Australia, pp. 322340.
Story, J.M., Callan, N.W., Corn, J.G. and White, L.J. (2006)
Decline of spotted knapweed density at two sites in western Montana with large populations of the introduced
root weevil Cyphocleonus achates (Fahraeus). Biological
Control 38, 227232.
Syrett, P., Briese, D.T. and Hoffmann, J.H. (2000) Success
in biological control of terrestrial weeds by arthropods.
In: Gurr, G. and Wratten, S. (eds) Biological Control:
Measures of Success. Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 189230.
Turner, C.E., Center, T.D., Burrows, D.W. and Buckingham,
G.R. (1998) Ecology and management of Melalueca
quinquenervia, an invader of wetlands in Florida, USA.
Wetlands Ecology and Management 5, 165178.
van Lenteren, J. (2006) How not to evaluate augmentative
biological control. Biological Control 39, 115118.
van Wilgen, B.W., de Wit, M.P., Anderson, H.J., Le Maitre,
D.C., Kotze, I.M., Ndala, S., Brown, B. and Rapholo,
M.B. (2004) Costs and benefits of biological control of invasive alien plants: case studies from South Africa. South
African Journal of Science 100, 113122.
Zavaleta, A. (2000) Valuing ecosystem services lost to Tamarix invasion in the United States. In: Mooney, H.A. and
Hobbs, R.J. (eds) Invasive Species in a Changing World.
Island Press, Washington DC, USA, pp. 261300.
74
Post-release non-target monitoring of
Mogulones cruciger, a biological control
agent released to control Cynoglossum
officinale in Canada
J.E. Andreas,1,2 M. Schwarzlnder,1 H. Ding1 and S.D. Eigenbrode1
Summary
Non-target effects of approved biological control agents have raised questions about the safety of
biological control of weeds and resulted in an increased emphasis on monitoring and reporting of
non-target effects as part of post-release assessments. This is particularly important in the case of the
root-mining weevil Mogulones cruciger (Herbst), which was approved in Canada to control houndstongue, Cynoglossum officinale L., but not in the United States because of concerns over its environmental safety. To address these concerns and the potential for non-target effects, we monitored
co-occurring confamilial Boraginaceae species at six M. cruciger release sites in Alberta and British
Columbia over two years. All four co-occurring species were attacked by the weevil to varying degrees although attack was inconsistent between years and sites, and non-targets were mostly attacked
to a lesser degree than houndstongue. There was a positive relationship between the probability of
non-target attack and houndstongue attack rate by M. cruciger indicating potential spillovers and
early evidence suggests non-target attack may be transitory. The comparison between the pre- and
post-release evaluations and preliminary plant volatile, electroantennogram, and host-choice behaviour data suggest that chemical ecology may provide an important tool in understanding an insects
host-choice selection in pre-release host-specificity assessments.
Keywords: houndstongue, non-target effects, host-choice behaviour, chemical ecology
Introduction
determine an insect species physiological and ecological host range, which is then used to predict an insects
realized host range once released in the invaded range
(Schaffner, 2001; van Klinken and Edwards, 2002).
Studies demonstrate that the physiological host range,
which can be reliably determined experimentally (e.g.
Papaj and Rausher, 1983; Szentesi and Jermy, 1990;
van Klinken, 2000; van Klinken and Heard, 2000), appears to be an effective criterion for identifying species
at risk of attack by introduced agents, since there is no
example of an insect agent attacking a plant outside its
physiological host range (Fowler et al., 2000; Pemberton, 2000; van Klinken and Edwards, 2002). A species
realized host range, however, is almost certainly narrower than its physiological host range, which is evident from the narrowing host range that is frequently
observed under increasingly natural testing conditions (i.e. multiple-choice tests and open-field tests).
However, adequate pre-release screening protocols to
Examples of non-target effects have created widespread
interest and concern through both the scientific and
public communities about the environmental safety of
biological weed control agents (Simberloff and Stiling,
1996; Louda et al., 1997, 2003 a, b; Strong, 1997; Thomas and Willis, 1998; Pemberton, 2000). The predictability of an agents host range is often at the center of the
debate. Conventional host-specificity tests are used to
Department of Plant, Soil, and Entomological Sciences, University of
Idaho, Moscow, ID 83844-2339, USA.
2
Washington State University, King County Extension, Suite 100, 200 Mill
Avenue South, Renton, WA 98057, USA.
Corresponding author: J.E. Andreas <jennifer.andreas@kingcounty.
gov>.
CAB International 2008
1
75
XII International Symposium on Biological Control of Weeds
predict this have not been developed (Van Driesche et
al., 2000; Schaffner, 2001; Hopper, 2001; and references
therein). Realized host ranges will be limited due to
phenological, ecological, and behavioural constraints.
While phenological and ecological factors likely differ
among continents and habitats, behavioural factors are
inherent to the agent and could be examined as part of
improved pre-release assessment protocols.
The host selection process of phytophagous insects
typically progresses from host finding (dependent upon
volatile and visual cues), through host examination
(dependent in addition on gustatory and tactile cues),
to host acceptance (oviposition or sustained feeding)
(Dethier, 1982; Miller and Strickler, 1984; Bernays
and Chapman, 1994). Each stage is dependent upon
completion of the previous stage. Thus, in nature, host
examination and acceptance, as evaluated in no-choice
bioassays, cannot occur unless the insect arrives at the
potential non-target host during the host-finding stage.
Therefore, knowledge of the chemical basis for hostfinding behaviour might improve our ability to assess
whether a non-target plant species that can support the
development of a biocontrol agent would be at risk of
attack in the field (Thomas and Willis, 1998; Heard,
2000; Schaffner, 2001).
The Mogulones cruciger Herbst, houndstongue
(Cynoglossum officinale L.), system provides a prime
example of the importance of such testing. Mogulones cruciger is a root-feeding weevil that has been
released in Canada to control the noxious rangeland
weed, houndstongue. Release in the United States,
however, has been denied because of concerns about
its environmental safety. Previous host-specificity testing has demonstrated that M. crucigers physiological
host range is fairly broad, but it has always shown a
strong preference for houndstongue (Jordan et al.,
1993; De Clerck-Floate et al., 1996; De Clerck-Floate
and Schwarzlaender, 2002; Andreas, 2004). A recent
study of six field sites in Canada found that M. cruciger was utilizing four confamilial species growing
sympatrically with houndstongue (Andreas, 2004). The
attacked Boraginaceae species are Cryptantha spiculifera (Eastw.) Payson, Hackelia floribunda (Lehm.)
I.M. Johnston, Lappula squarrosa (Retz.) Dumort and
Lithospermum ruderale Dougl. ex Lehm.. The first
three species are within the known physiological host
range of M. cruciger (De Clerck-Floate et al., 1996;
De Clerck-Floate and Schwarzlaender, 2002; Andreas,
2004) while the latter species has not been sufficiently
tested because its cultivation has not been successful.
Because M. cruciger is known to prefer houndstongue
(Jordan et al., 1993; De Clerck-Floate et al., 1996; De
Clerck-Floate and Schwarzlaender, 2002; Andreas,
2004), we hypothesize that these non-target attacks result from spillover, sensitization/central excitation
effects or both. In spillover, high population densities on
the target result in some insects colonizing non-targets
due to random dispersal especially after the target re76
source begins to be depleted (Strong, 1997). In sensitization/central excitation, acceptance thresholds for
non-targets are lowered after contact with the true host
or in the presence of ambient volatile organic compounds emitted by the true host (Marohasy, 1996, 1998;
Withers and Barton Browne, 1998). Regardless of the
specific mechanism, these recorded non-target attacks
indicate a need to determine the risk that M. cruciger
poses to native species that are within the physiological host range and within the potential range of the
weevils dispersal from target host populations. Therefore, we explored the early stages of host-selection
behaviour and its underlying phytochemical basis to
determine the plant species likely to be colonized by
M. cruciger.
Materials and methods
Study organisms
Houndstongue is a Eurasian herbaceous, facultative, short-lived perennial. Plants produce rosettes in
the first year and typically reproduce in the second or
third year (Wesselingh et al., 1997). After sexual reproduction, barbed nutlets are formed and dispersed
via epizoochory (De Clerck-Floate, 1997). In Europe,
the plant is found in sand dunes, roadsides, and open
woodlands (Tansley and Adamson, 1925; Tutin et al.,
1972; Hegi, 1975; Klinkhamer and de Jong, 1988). In
North America, this ruderal species colonizes disturbed
areas, rangelands, pastures, and forests (Macoun, 1884;
Upadhyaya and Cranston, 1991).
Mogulones cruciger is a root-feeding weevil native
to central Europe. In spring, after hibernation, adults
occur on above-ground plant parts, where they feed on
leaves, mate, and oviposit into leaf petioles (Schwarzlaender, 1997). Larvae hatch from eggs 7-10 days after
oviposition and begin to mine down into the root crown
where they feed primarily in the vascular cylinder. Mogulones cruciger has three instars and pupates in the
surrounding soil (Schwarzlaender, 1997). In late summer, adults emerge and begin feeding on houndstongue
rosettes between July and October. Oviposition begins
in late August until temperatures cool and oviposition
sites become unavailable. Adult weevils hibernate in
leaf litter (Schwarzlaender, 1997; De Clerk-Floate and
Schwarzlaender, 2002). Mogulones cruciger often has
overlapping generations. As a consequence, eggs and
larvae can be found in houndstongue roots and leaf
petioles at most times of the year.
The native North American Boraginaceae species
Hackelia venusta (Piper) St. John was chosen as a nontarget species in this study because it is listed as endangered by the United States Fish and Wildlife Service
(USFWS) and laboratory tests indicated that M. cruciger
is capable of partial larval development on this species
(Andreas, 2004). Its distribution is limited to one small
remaining population of approximately 150 individuals
Post-release non-target monitoring of Mogulones cruciger
in a ponderosa pine (Pinus ponderosa P. & C. Lawson)
and Douglas fir (Psuedotsuga menziesii (Mirbel) Franco)
clearing in Chelan County, Washington, just south of
the Canada/U.S. border (Center for Plant Conservation,
2004). Fire suppression has been an important factor
in the reduction of H. venusta populations (Center for
Plant Conservation, 2004). Cryptantha spiculifera and
Hackelia floribunda were selected for this experiment
because they co-occur with houndstongue at field sites
in Canada (Kartesz, 1999) and were monitored in a
post-release open field study (Andreas, 2004).
Olfactometer bioassay
Our methods for an olfactometer bioassay and for
electroantennogram (EAG) and GC/EAD have been
adapted from those proven effective for a close relative
of M. cruciger, the cabbage seed weevil, Ceutorhynchus obstrictus Marsham (C. assimilis Payk.) (Evans
and Allen-Williams, 1992; Evans and Bergeron, 1994).
The olfactometer was a four-arm configuration (Vet
et al., 1983) (Syntech, Hilversham, The Netherlands).
We conducted preliminary 3-way choice bioassays in
the olfactometer with air streams directed over intact
plants of H. floribunda, C. spiculifera, houndstongue,
and a blank (carrying humidified air). Intake air for all
treatments was prefiltered through activated charcoal.
Airflow was balanced at 300 ml/min through each arm
of the olfactometer.
Illumination was with overhead fluorescent fixtures,
diffused through a white, translucent plastic pail inverted
over the olfactometer. The lens of a video camera was
inserted through the top center of the pail to permit
continuous recording of weevil behaviour during the
bioassay. Twelve female weevils (having fed previously
on houndstongue but not having encountered other
hosts) were tested simultaneously in a single run. The
weevils were placed at the center of the olfactometer
and their positions and movements recorded on videotape for 30 minutes for later review and recording
using a computer program (Noldus Observer, Wageningen, The Netherlands). Relative attractiveness of each
source was quantified in terms of the proportion of time
spent in the respective quadrants of the olfactometer.
VOC analysis
Volatile organic compounds (VOCs) were trapped
from the headspace of houndstongue, H. venusta,
H. floribunda, and C. spiculifera with a volatile collection apparatus (Analytical Research Systems, Inc.,
Gainesville, FL) following methods modified from Eigenbrode et al. (2002). The headspace VOC profiles
were compared for similarity based on the number of
shared compounds detected and by calculating similarity coefficients based on occurrence and relative abundance of each component. We employed two binary
coefficients, Jaccards and Srensens, and one quanti77
tative coefficient, Bray-Curtis, commonly used for ecological studies (Southwood and Henderson, 2000).
EAG and GC/EAD/FID
EAG assay methods were modified from those of
Evans and Allen-Williams (1992). Recordings were
taken from single antennae on partially excised heads
using sharpened stainless steel electrodes coated with
an electrolyte gel. To measure antennal response to
the total blend of VOCs from each species, samples
of headspace volatile that had been standardized based
on plant dry weight were applied in solvent to a filter
paper strip and delivered to the antenna in a puff of prefiltered air. Antennal response to VOCs of each species
was standardized on the basis of a response to a single
concentration of linalool. Responses from ten weevils
(three males and seven females) were obtained from
H. floribunda.
Gas chromatography using EAG and flame ionization detectors (GC/EAD/FID) (Bjostad, 1998) can help
identify the components of potential host VOCs that
are most important for observed behavioural responses.
A 1-l sample of headspace VOCs dissolved in methylene chloride was injected onto a Shimadzu GC17-A
GC (Shimadzu Corp., Kyoto Japan) fitted with a column
splitter delivering approximately half the column effluent to a flame ionization detector and half via a heated
transfer line to the electroantennograph (Syntech, Hilversam, The Netherlands) with the insect preparation.
Temperature programme and column specifications
were conducted as described in Eigenbrode et al.
(2002). Antennal responses (expressed as mV of depo
larization) were standardized based on an external
sample of linalool injected through the GC/FID/EAD
between each injected sample of headspace VOCs. One
female was tested.
Results
Olfactometer bioassay
The results from the behavioural bioassay (Figure 1)
indicate that M. cruciger females responded to host
VOCs in the olfactometer. The weevils spent the smallest proportion of time in the H. floribunda quadrant of
the olfactometer, approximately 20% of their time in
C. spiculifera and the no-plant control quadrants, and
the greatest amount of time (approximately 40 %) in the
houndstongue quadrant (One-way ANOVA; F = 4.71,
P = 0.006).
VOC analysis
Of the 44 VOCs detected in headspace of the four
species, six were shared by all and six were unique to
houndstongue. Hackelia venusta headspace had the
most compounds in common with houndstongue, followed by C. spiculifera and then H. floribunda. All
XII International Symposium on Biological Control of Weeds
Control
C. spiculifera
H. floribunda
C. officinale
10
30
20
40
60
50
Average % time in quadrant
Figure 1.
Response of Mogulones cruciger females to volatile organic compounds (VOCs) from houndstongue (Cynoglossum officinale), two non-target plant species (Cryptantha spiculifera and Hackelia floribunda) and a blank (carrying humidified air) in a four-arm olfactometer. Values are percentage of time in extremity of the test arm with
stimulus (n=12).
three similarity coefficients follow a similar trend, with
H. venusta most similar to houndstongue, followed by
C. spiculifera and then H. floribunda (Table 1).
EAG and GC/FID/EAD
EAG: Antennal preparations exhibited depolarization (i.e. peaks in impulses indicate excitation to particular compounds) in response to puffs of extracted
headspace VOCs and this response tended to be
stronger for houndstongue compared to H. floribunda
Table 1.
(Figure 2) (2-way ANOVA P values = 0.08, 0.16 and
0.28 for species, sex and interaction, respectively). GC/
FID/EAD: The combined FID/EAD trace resulting from
chromatography of H. venusta volatiles showed that
an individual M. cruciger female antenna responded
to a subset of the VOCs present.
Discussion
Despite the recognized need to include host-choice
behaviour and chemical ecology in host-range inves-
Measures of similarity for headspace volatile profiles of Hackelia venusta, Cryptantha spiculifera and Hackelia
floribunda, as compared with houndstongue (Cynoglossum officinale) headspace volatiles.
Species compared to C. officinale
H. venusta
Shared components
21
Srensena
0.84
Jaccardb
0.72
Bray-Curtisc
0.26
C. spiculifera
11
0.41
0.26
0.23
H. floribunda
0.36
0.22
0.22
a
The Srensen coefficient is binary in that it is based on the presence/absence of each compound. It is calculated as: Cs= 2a/(2a + b + c) in
which a = the number of compounds held in common, b = the number of compounds unique to houndstongue and c = the number of compounds unique to the species compared to houndstongue. It ranges from zero for non-overlapping profiles to one for identical profiles.
b
Jaccard similarity is also binary. It is calculated as: Cj= a/(a + b + c) and also ranges from zero to one.
c
The Bray-Curtis coefficient is a quantitative index of similarity that takes into account abundance of the components. It is calculated as
Cn = 2jN/(aN + bN), in which jN = the sum of lesser values for those compounds shared between species and aN and bN are the sum of total
values for each species. The coefficient scales from near one for equivalent profiles to zero.
78
Post-release non-target monitoring of Mogulones cruciger
male
C. officinale
H. floribunda
female
C. officinale
H. floribunda
0
0.5
1.5
2.5
Standardized EAG response
Figure 2.
EAG responses by male and female Mogulones cruciger to puffs of total VOC from its host, houndstongue
(Cynoglossum officinale) and a non-target species, Hackelia floribunda (response standardized relative to linalool), (n= three males and seven females).
tigations of candidate weed biocontrol agents (Briese,
2005; Sheppard et al., 2005), few studies have attempted
this (Heard, 2000; Hopper, 2001; Schaffner, 2001). Our
data provide an illustration of the type of information
that can be obtained with behavioural and chemical
ecological experiments.
Our olfactometer results indicate greater responsiveness of M. cruciger to odour cues from houndstongue
than to the tested non-target species. We did not detect directed upwind movement by the weevils in this
bioassay (data not shown), indicating that M. cruciger
dispersal among potential hosts could be undirected
but that host VOCs can stimulate arrested or restricted
searching behaviour on or near the host. Our bioassay
has some limitations for fully understanding relevant
host-selection responses to target and non-target VOCs
in the field. We studied walking behaviour, but flying is
likely also important in host location. The three-odour
choices we offered may not represent the situation encountered from isolated non-target plants in the field.
Further testing is required to address these concerns.
Our focus on VOC isolates behaviour during early
stages of host selection. After contact with potential
hosts, appropriate tactile and gustatory cues are required for host acceptance. Further study of the socalled examination phase of host selection is needed
for a more complete understanding of factors determining the realized host ranges of candidate biological
control agents. For example, M. cruciger may respond
to detectable pyrrolizidine alkaloids on the plant surface. Although recent work failed to find that these
alkaloids could explain oviposition preference by another specialist insect, the cinnabar moth, Tyria jacobaeae L. (Lepidoptera: Arctiidae) (Macel and Vrieling,
79
2003), the response by M. cruciger to these alkaloids
and other gustatory cues of target and non-target plants
should be examined.
The basis of apparent M. cruciger discrimination
among odours of houndstongue and non-target species
remains unknown. Some insects apparently integrate
information from the VOC blend of their host plants
(Roseland et al., 1992), while others, including a close
relative of M. cruciger, the cabbage seedpod weevil,
C. obstrictus, use a few specific VOCs to orient to
potential hosts (Smart and Blight, 1997). The VOC profiles of H. floribunda, H. venusta and C. spiculifera differ from houndstongue. Hackelia floribunda VOCs are
the least similar to those of houndstongue. Among the
three tested non-target species, the VOCs of only two,
C. spiculifera and H. floribunda, were subjected to the
olfactometer bioassay, where H. floribunda was once
again less preferred. If specific compounds are required
for host acceptance, they may be present in houndstongue but lacking in H. floribunda. Moreover, since
H. floribunda appeared repellent in our tests, VOCs
present in its profile, but lacking in houndstongues
(i.e. benzaldehyde, undecane and dodecane) are candidate repellents. On the other hand, the weevils may integrate information from several cues during response
to the VOC blends of potential hosts. For determining
host-range tendencies in pre-release studies, it may not
be necessary to determine these mechanisms. Our result indicates that it is feasible to include responses to
cues used during host finding in assessing risks of colonization of non-targets as part of pre-release studies.
Our results with M. cruciger indicate that H. floribunda
may be at less risk than the other species tested here
because the weevils are not attracted and possibly even
XII International Symposium on Biological Control of Weeds
repelled by this plant, whereas H. venusta may be at
greater risk because of the similarity of its VOC pattern
to that of houndstongue.
The weevils response to several components in the
GC/FID/EAD test may help determine the basis of its
host selection behaviour, as these volatiles are candidate components whose presence could attract the weevil to a non-target host that produces them. However,
given that only one weevil species and one plant species was used, the results presented here simply revealed that M. cruciger can detect plant stimuli.
Our approach could help elucidate the risk M. cruciger poses to non-targets near (e.g. Andreas, 2004)
and remote from colonized houndstongue infestations. In the former setting, weevils dispersing from
houndstongue into the environment will be influenced
by their responsiveness to VOCs and other long-
distance cues from potential hosts. If there is little or no
response to non-target cues (as we may have found for
two non-targets), the risk of non-target attack should
be reduced. Spillover effects or central excitation/sensitization potentially alter this assessment. If weevils
are sufficiently abundant and mobile, regardless of
responses to VOC from potential hosts, they may encounter and colonize the non-targets (spillover). If prior
contact with hosts or host odours lowers acceptability
thresholds to subsequently encountered plants (central
excitation/sensitization), temporary non-target attack
could be facilitated. Longer-distance colonization of
non-targets is also potentially mediated by VOCs. Our
result suggests that, while walking, weevil colonization
of the two non-targets we tested are not increased due
to their released VOCs.
Our approach could help assess risks of attack of
other sensitive non-targets by M. cruciger. Three of the
five endangered Boraginaceae species in the United
States are within the weevils physiological host range.
Complete development was possible on Plagiobothrys
hirtus (Greene) I.M. Johnston and Amsinckia grandiflora (Kleeb. ex Gray) Kleeb. ex Greene and partial
development occurred on H. venusta (Andreas, 2004).
The two remaining confamilial species that are listed as
endangered in the United States, Cryptantha crassipes
I.M. Johnston and Plagiobothrys strictus (Greene) I.M.
Johnston were not tested. VOC profiles of these species
could be tested to assess weevil attraction. H. venusta
VOC profiles are more similar to houndstongue than
others we have tested and are potentially attractive or
arrestant for M. cruciger.
Comparisons of realized and predicted host ranges
from pre-release studies (Cullen, 1990; Briese et al.,
1995; Clement and Cristofaro, 1995; Briese, 1999) can
evaluate the soundness of prior risk-assessment procedures (Ewel et al., 1999; Hopper, 2001). Selection behaviour and its phytochemical basis could, in contrast,
greatly improve predictions of eventual host ranges of
released agents (Marohasy, 1998; Thomas and Willis,
1998; Heard, 2000; Schaffner, 2001). By using meth80
ods similar to those presented here, behavioural and
electrophysiological bioassays could form the basis
of risk-assessment studies. Specifically, the release of
M. cruciger into North America provides an opportunity
to compare its realized host range with host-selection
predictions based on phytochemical and behavioural
studies.
Acknowledgements
The authors thank Brad Harmon for lab assistance, the
Idaho Agricultural Experiment Station, Terry Miller
for the use of the quarantine at Washington State
University and the editors for manuscript revisions.
Dr. Sarah Reichard kindly provided the H. venusta
plant material. Travel funds for the presenting author
kindly provided by the United States Forest Service
and Washington State University.
References
Andreas, J.E. (2004) Non-target effects of Mogulones cruciger Herbst (Coleoptera: Curculionidae), a biocontrol
agent released to control houndstongue in Canada. MSc
thesis, Department of Plant, Soil and Entomological Sciences, University of Idaho, Moscow, ID.
Bernays, E.A. and Chapman, R.F. (1994) Host plant selection by phytophagous insects. Chapman and Hall, New
York, NY.
Bjostad, L.B. (1998) Electrophysiological methods. In: Millar,
J. and Phelan L. (eds) Methods in Chemical Ecology. Kluwer Academic, Dordrecht, The Netherlands, pp. 339376.
Briese, D.T., Sheppard, A.W.and Reifenberg, J.M. (1995)
Open-field host specificity testing for potential biological
control agents of Onopordum thistles. Biological Control
5, 158166.
Briese, D.T. (1999) Open field host-specificity tests: is natural good enough for risk assessment? In: Wither, T.M.,
Barton Browne, L. and Stanley, J. (eds) Host specificity
testing in Australasia: towards improved assays for biological control. CRC for Tropical Pest Management, Brisbane, Australia, pp. 4459.
Briese, D.T. (2005) Translating host-specificity test results
into the real world: The need to harmonize the yin and
yang of current testing procedures. Biological Control 35,
208214.
Center for Plant Conservation (2004). CPC National collection plant profile. http://ridgwaydb.mobot.org/cpcweb/
CPC_ViewProfile.asp?CPCNum=2109
Clement, S.L. and Cristofaro, M. (1995) Open-field tests in
host-specificity determination of insects for biological
control of weeds. Biocontrol Science and Technology 5,
395406.
Cullen, J.M. (1990) Current problems in host specificity screening In: Delfosse, E.S. (ed.) Proceedings of the VII International Symposium on Biological Control of Weeds. Ministry
for Agriculture and Forestry, Rome, Italy, pp. 2736.
De Clerck-Floate, R., Schroeder, D., and Schwarzlaender, M.
(1996) Supplemental information to the petition (Can-93-4
and TAG 93-06) to release Ceutorhynchus (Mogulones)
cruciger for the biological control of hounds-tongue
Post-release non-target monitoring of Mogulones cruciger
(Cynoglossum officinale, Boraginaceae) in Canada. Agriculture and Agri-Food Canada report.
De Clerck-Floate, R. (1997) Cattle as dispersers of houndstongue on rangelands in southeastern British Columbia.
Journal of Range Management 50, 239243.
De Clerck-Floate, R. and Schwarzlnder, M. (2002) Host
specificity of Mogulones cruciger (Coleoptera: Curculionidae), a biocontrol agent for houndstongue (Cynoglossum officinale), with emphasis on testing of native North
American Boraginaceae. Biocontrol Science and Technology 12, 293306.
Dethier, V.G. (1982) Mechanisms of host-plant recognition.
Entomologia experimentalis et applicata 31, 4956.
Eigenbrode, S.D., Ding, H., Shiel, P. and Berger, P.H. (2002)
Volatiles from potato plants infected with potato leafroll
virus attract and arrest the virus vector, Myzus persicae
(Homoptera: Aphididae). Proceedings of the Royal Society, Series B 269, pp. 455460.
Evans, K.A. and Allen-Williams, L.J. (1992) Electorantennogram responses of the cabbage seed weevil, Ceutorhynchus
assimilis, to oilseed rape, Brassica napus ssp. oleifera,
volatiles. Journal of Chemical Ecology 18, 16411659.
Evans, K.A. and Bergerson, J. (1994) Behavioral and electrophysiological response of cabbage seed weevil (Ceutorhynchus assimilis) to conspecific odors. Journal of
Chemical Ecology 20, 979989.
Ewel, J.J., ODowd, D.J., Bergelson, J., Daehler, C.C.,
DAntonio, C.M., Gmez, L.D. Gordon, D.R., Hobbs,
R.J., Holt, A., Hopper, K.R., Hughes, C.E., LaHart, M.,
Leakey, R.R.B., Lee, W.G., Loope, L.L., Lorence, D.H.,
Louda, S.M., Lugo, A.E. McEvoy, P.B., Richardson, D.M.
and Vitousek, P.M. (1999) Deliberate introductions of
species: research needs. BioScience 49 (8), 619630.
Fowler, S.V., Syrett, P. and Hill, R.L. (2000) Success and
safety in the biological control of environmental weeds in
New Zealand. Austral Ecology 25, 553562.
Heard, T.A. (2000) Concepts in insect host-plant selection behavior and their application to host-specificity testing. In:
Van Driesche, R., Heard, T., McClay, A. and Reardon, R.
(eds) Proceedings of Session: Host Specificity Testing of
Exotic Arthropod Biological Control Agents The Biological Basis for Improvement in Safety, Proceedings of the X
International Symposium on Biological Control of Weeds.
US Forest Service, Forest Health Technology Enterprise
Team, Report FHTET-99-1, Morgantown, WV, pp. 110.
Hegi, G. (1975) Illustrated Flora of Central Europe, vol. 5,
part 3, Pirolaceae-Verbenaceae. Paul Parey, Hamburg and
Berlin, Germany.
Hopper, K.R. (2001) Research needs concerning non-target
impacts of biological control introductions. In: Wajnberg,
E., Scott, J.K. and Quimby, P.C. (eds) Evaluating Indirect
Ecological Effects of Biological Control. CABI Publishing, Oxon, UK, pp.3955.
Jordan, T., Schwarzlnder, M., Tosevski, I. and Freese, A.
(1993) Ceutorhynchus cruciger Herbst (Coleoptera,
Curculionidae): a candidate for the biological control of
hounds-tongue (Cynoglossum officinale L., Boraginaceae) in Canada. Unpublished Final Report. International
Institute of Biological Control.
Kartesz, J.T. (1999) A synonymized checklist and atlas with
biological attributes for the vascular flora of the United
States, Canada and Greenland. First Edition. In: Kartesz,
J.T. and Meacham, C.A. (eds) Synthesis of the North
81
American Flora, version 1.0. North Carolina Botanical
Garden, Chapel Hill, NC.
van Klinken, R.D. (2000) Host-specificity constraints evolutionary host change in the psyllid Prosopidopsylla flava.
Ecological Entomology 25, 413422.
van Klinken, R.D. and Heard, T.A. (2000) Estimating fundamental host range: a host-specificity study of a biocontrol
agent for Prosopis species (Leguminosae). Biocontrol
Science and Technology 10, 331342.
van Klinken, R.D. and Edwards, O.R. (2002) Is host-specificity
of weed biological control agents likely to evolve rapidly
following establishment? Ecology Letters 5, 590596.
Klinkhamer, P.G.L. and de Jong, T.J. (1988) The importance
of small-scale disturbance for seedling establishment in
Cirsium vulgare and Cynoglossum officinale. Journal of
Ecology 76, 383392.
Louda, S.M., Kendall, D., Connor, J. and Simberloff, D.
(1997) Ecological effects on an insect introduced for the
biological control of weeds. Science 277, 10881089.
Louda, S.M., Pemberton, R.W., Johnson, M.T. and Follett,
P.A. (2003a) Non-Target effectsthe Achilles heel of biological control? Retrospective analyses to reduce risk associated with biocontrol introductions. Annual Review of
Entomology 48, 365399.
Louda, S.M., Arnett, A.E., Rand, T.A. and Russell, F.L.
(2003b) Invasiveness of some biological control insects
and adequacy of their ecological risk assessment and regulation. Conservation Biology 17, 7382.
Macel, M. and Vrieling, K. (2003) Pyrrolizidine alkaloids as
oviposition stimulants for the cinnabar moth. Journal of
Chemical Ecology 29, 14351446.
Macoun, J. (1884) Catalogue of Canadian plants. Part 2, Gamopetalae, p. 334. Dawson Brothers, Montreal, Canada.
Marohasy, J. (1996) Host shifts in biological weed control:
real problems, semantic difficulties or poor science? International Journal of Pest Management 42, 7175.
Marohasy, J. (1998) The design and interpretation of hostspecificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News and
Information 19(1), 13N-20N.
Miller, J.R. and Strickler, K.L. (1984) Finding and accepting
host plants, In: Bell, W.J. and Card, R.T. (eds) Chemical Ecology of Insects. Chapman and Hall, London,
pp. 128157.
Papaj, D.R. and Rausher, M.D. (1983) Individual variation
in host location by phytophagous insects. In: Ahmed, S.
(ed.) Herbivorous Insects: Host Seeking Behaviour and
Mechanisms, Academic Press, New York, pp. 77124.
Pemberton, R.W. (2000) Predictable risk to native plants in
weed biological control. Oecologia 125, 489494.
Roseland, C.R., Bates, M.B., Carlson, R.B. and Oseto, C.Y.
(1992) Discrimination of sunflower volatiles by the red
sunflower seed weevil. Entomologia Experimentalis et
Applicata 62, 99106.
Schaffner, U. (2001) Host range testing of insects for biological weed control: how can it be better interpreted? BioScience 51, 19.
Schwarzlaender, M. (1997) Bionomics of Mogulones cruciger (Coleoptera: Curculionidae), a below-ground herbivore for the biological control of hounds-tongue. Environmental Entomology 26, 357365.
Sheppard, A.W., van Klinken, R.D. and Heard, T.A. (2005)
Scientific advances in the analysis of direct risks of weed
XII International Symposium on Biological Control of Weeds
biological control agents to non-target plants. Biological
Control 35, 215226.
Simberloff, D. and Stiling, P. (1996) How risky is biological
control? Ecology 77, 19651974.
Smart, L.E. and Blight, M.M. (1997) Field discrimination of
oilseed rape, Brassica napus, volatiles by cabbage seed
weevil, Ceutorhynchus assimilis. Journal of Chemical
Ecology 23, 25552567.
Southwood, T.R.E. and Henderson, P.A. (2000) Ecological
methods, 4th edn. Blackwell Science, London.
Strong, D.R. (1997) Fear no weevil? Science 277, 10581059.
Szentesi, A. and Jermy, T. (1990) The role of experience in
host choice by phytophagous insects. In: Bernays E.A.
(ed.) InsectPlant Interactions, Volume 2. CRC Press,
Boca Raton, FL, pp. 3974.
Tansley, A.G. and Adamson, R.S. (1925). Studies of the
vegetation of the English chalk. Journal of Ecology 13,
117223.
Thomas, M. B. and Willis, A.J. (1998). Biocontrolrisky but
necessary? Trends in Ecology and Evolution 13, 325329.
Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M. and Webb, D.A. (ed.) (1972)
Flora Europaea. Volume 3, Diapensiaceae to Myoporaceae. Cambridge University Press.
82
Upadhyaya, M.K., Cranston, R.S. (1991) Distribution, biology, and control of hounds-tongue in British Columbia.
Rangelands 13, 103106.
Van Driesche, R.G., Heard, T., McClay, A., and Reardon,
R. (2000). Proceedings of Session: Host Specificity Testing of Exotic Arthropod Biological Control AgentsThe
Biological Basis for Improvements in Safety. Proceedings
of the X International Symposium on Biological Control
of Weeds, Bozeman Montana. U.S. Forest Service, Forest
Health Technology Enterprise Team, Report FHTET-99-1,
Morgantown, WV.
Vet, L.E.M., Van Lenteren, J.C. Heymans, M. and Meelis, E.
(1983) An airflow olfactometer for measuring olfaction of
hymenopterous parasitoids and other small insects. Physiological Entomology 8, 97106.
Wesselingh, R.A., Klinkhamer, P.G.L., de Jong, T.J. and
Boorman, L.A. (1997) Threshold size for flowering in
different habitats: effects of size-dependent growth and
survival. Ecology 78, 21182132.
Withers, T. and Barton Browne, L. (1998) Possible causes of
apparently indiscriminate oviposition in host-specificity
tests using phytophagous insects. In: Sixth Australasian
Applied Entomological Research Conference. University
of Queensland, Brisbane, Australia, pp. 565571.
Assessing indirect impacts of biological
control agents on native biodiversity:
a community-level approach
L.G. Carvalheiro,1 Y.M. Buckley,2,3 R. Ventim1 and J. Memmott1
Summary
The safety of biological control methods is a subject that has received considerable attention for
a long time. However, apparent competition (competition due to shared natural enemies) has been
neglected when considering possible impacts of biological control agents. One of the reasons for the
lack of studies in this area is the difficulty in assessing and predicting indirect effects due to apparent
competition. In this paper we outline a methodology to predict and measure non-target impacts of
biological control agents due to apparent competition.
Keywords: biological control, methodology.
Underlying rationale
Invasive species are one of the main threats to global
biodiversity (Schmitz and Simberloff, 1997). Classical
biological control involves the deliberate introduction
of an alien species and it is viewed as a sustainable, environmentally friendly form of pest control. The safety
of biological control is a subject that has received much
attention, with particular concerns about the interactions between biological control agents and non-target
species (Pemberton and Strong, 2000; Thomas and
Willis, 1998; Boettner et al., 2000; Louda et al., 1997).
Non-target species can be affected directly, if an agent
attacks a non-target host, or indirectly, for instance,
when the agent shares natural enemies with native species (apparent competition, reviewed by Holt and Lawton, 1994). One of the main criteria for a certain species
to be considered a safe biological control agent is its
high host specificity, reducing its likelihood to directly
affect native species. However, a successfully established biological control agent is an abundant resource
for natural enemies present in the target ecosystem,
School of Biological Sciences, Woodland Road, Bristol BS8 1UG, UK.
School of Integrative Biology, University of Queensland, St. Lucia,
QLD 4072, Australia.
3
CSIRO Sustainable Ecosystems, Queensland Bioscience Precinct,
306 Carmody Road, St. Lucia, QLD 4067, Australia.
Corresponding author: L.G. Carvalheiro <luisa.gigantecarvalheiro@
bristol.ac.uk>
CAB International 2008
1
83
such as parasitoids, parasites and pathogens, which
can frequently be oligophagous or polyphagous (e.g.
Hawkins and Goeden, 1984; Memmott et al., 1994).
Therefore, if these natural enemies include such an
abundant food resource in their diet, their population
abundance can in turn increase, creating a potential for
apparent competition.
Several studies have shown that apparent competition can have strong impacts on population dynamics,
either due to shared parasites (Tompkins et al., 2000),
predators (Muller and Godfray, 1997) or parasitoids
(Morris et al., 2001), as well as on community structure (e.g. herbivorous communities in Morris et al.,
2004, 2005; aphid-parasitoid communities in Muller
and Godfray, 1999; Muller et al., 1999). However,
non-target impacts of an introduced biological control
agent on native species through apparent competition
is a subject that has not received much attention (Willis
and Memmott, 2005).
If a biological control agent is effective in reducing
weed abundance to low levels, then non-target impacts
due to apparent competition can be minimal. However, very few pre-release studies have predicted the
effectiveness of potential biological control agents in
reducing target weed abundance (e.g. Buckley et al.,
2005; Wirfl, 2006). If an introduced agent remains at
high abundance over a long period of time, the probability of non-target effects due to apparent competition is enhanced. Furthermore, non-target impacts are
of particular concern for endemic species whose distribution range overlaps completely with the range of
XII International Symposium on Biological Control of Weeds
the weed/biological control agent, as they are the most
likely species to suffer irreversible damage that may
potentially lead to their extinction.
community, which can then be tested using regression
models.
Suggested methodology
Community level approach
Plantinsect interaction systems can be extremely
complex, involving dense webs of interactions (e.g.
Waser et al., 1996; Memmott, 1999; Muller et al., 1999;
Bascompte et al., 2003). Thus, to fully assess the potential indirect effects of biological control, community-
level surveys are necessary. Food webs have been
suggested as the appropriate way of analysing possible
non-target interactions in biological control (Henneman
and Memmott, 2001; Strong, 1997), since food webs
enable us to ask how a biological control agent can
influence native communities (Memmott et al., 2007).
Using food webs as predictive tools in conservation
biology has, until recently, been considered an unattainable goal, as at first sight they appear very labor
intensive to make and statistically difficult to analyse
(Memmott et al., 2007). However, community-level
ecology has developed to a stage where we are capable
of sampling, visualizing and analysing complex food
web interactions at community-level scale (Memmott
et al., 2004; Dunne et al., 2002; Sole et al., 2001; Bersier, et al., 2002; Banasek-Richter et al., 2004; Cattin
et al., 2004).
Some studies have already used a community-level
approach to look for non-target effects of biological
control agents. For example, Louda et al., (1997) used
this approach to highlight the ability of biological control agents to disrupt communities. They demonstrated
that an exotic seed-feeding biological control agent was
displacing native seed feeders associated with non-
target plants. Henneman and Memmott (2001) used this
approach to show that in a remote area of Hawaii, 83%
of parasitoids reared from native moths were biological
control agents. Nowadays, this type of non-target impact (due to lack of host specificity) is avoided by using the current safety regulations governing biological
control (e.g. Fowler et al., 2000; Sheppard et al., 2005).
However, indirect non-target impacts are much harder
to predict and avoid. Willis and Memmott (2005) revealed that the biological control agent, Mesoclanis polana (Munro) (Diptera: Tephritidae) had the potential
to disrupt the native food web structure due to apparent competition, mediated by shared native parasitoids,
whose population abundances exponentially increased
following the population outbreak of M. polana. However, this study did not clearly test for impacts of the weed
and the biological control agent on abundance and/or
species richness of native communities. To test for such
effects, repeated sampling in sites with different abundances of weed and biological control agent is needed.
In this paper we propose that food webs provide a
protocol that can quantify the impact of both the alien
plant and its biological control agent upon the natural
84
For a correct assessment of the impacts of the abundance of the weed and its biological control agents,
two components need to be included in a post-release
impact assessment programme. The first component is
descriptive, involving the construction of food webs describing the patterns of trophic linkages between plants,
herbivores and parasitoids in communities invaded by
weeds. The second component involves statistical testing of the effects of the weed and the biological control
agent abundance on native communities abundance
and species richness.
Sampling
Selection of ten to 20 plots covering all habitats that
are threatened by the weed, and covering a gradient
of abundance of the weed and the biological control
agent, is required. Plot size should be selected in order
to include the maximum number of plant species of the
field site (a suggested size of the plot is 40 40 m), and
the plots should be at least 500 m apart, so they can be
considered independent.
Ideally, all ecological niches would be studied, but
it is more practical to focus on the most likely ecological niche to be affected, this being the one that includes
the biological control agent in focus (e.g. seed predators, leaf miners, herbivores) and its parasitoids. Furthermore, assessing parasitism has another advantage,
since it may also be highly relevant to the success and
impact of the biological control programme.
Community-level sampling requires a high amount
of effort. Based on a pilot project, we estimate that
it will take approximately four weeks to sample 20
field sites, with two full-time people. Repeated sampling over time is needed during the seasons of higher
abundances of the biological control agent to include
the maximum number of species. The plots should be
sampled for plants, herbivores and parasitoids monthly.
The sampling and rearing methods have been described
in previous literature: seed predators and their parasitoids (Memmott and Godfray, 1994); leaf herbivores
and their parasitoids (Memmott et al., 1994; Lewis et
al., 2002); and aphids and their parasitoids (Muller and
Godfray, 1997; Muller et al., 1999). Rearing time can
vary with the biology and geographical region of the
species involved. As an example, a pilot study with
seed predators in Australia involved ten weeks of rearing after samples were collected.
Determining species links
It is relatively straightforward to determine trophic
links between herbivorous insects and plant species.
Assessing indirect impacts of biological control agents on native biodiversity: a community-level approach
Determining the parasitoids of most herbivore species can be also straightforward. Immature stages of
the host insect are reared in isolation until either adult
hosts or parasitoids emerge (Memmott, 1999; Memmott and Godfray, 1994). Determining parasitoids of a
given endophagous herbivore species (e.g. seed predator) is not as simple, as the seed predators themselves
develop inside the seed. However, for many plant species there are only a few pre-dispersal seed predators
and information in the literature on the food habits of
the parasitoid species may be enough to identify the
host. Plantherbivoreparasitoid webs are taxonomically complex; therefore, taxonomic input is essential,
although it can be time consuming and costly.
pattern, allowing the assessment of the total magnitude of the effects of the biological control agent.
The approach presented here has recently been applied by the authors to test for indirect impacts due
to apparent competition of a highly specific biological control agent, Mesoclanis polana Munro, recently
introduced in Australia (1996) to control an invasive
weed, Chrysanthemoides monilifera (L.) T. Norlindh,
spp rotundata (Carvalheiro et al., 2008).
Acknowledgements
We would like to thank Kate Henson for her comments
on the manuscript and Fundao para a Cincia e Tecnologia for funding.
Testing for apparent competition
References
Effects of the weed and biological control agent
on native communities of herbivores/seed predators,
parasitoids and plants can be tested by using generalized linear models (GLMs) where all possible combinations of the relevant variables (e.g. habitat, latitude,
weed abundance, biological control agent abundance)
will be tested. By ranking all possible models, the best
model can be selected (Zuur et al., 2007). If the effect
of the biocontrol agent is strong enough, a significant
effect will be detected over and above the effect of the
weed abundance. This allows differentiating which
native community patterns are significantly related to
the weed abundance and/or to the biological control
agent abundance. For example, if a model including
the biological control agent abundance is selected as
the best model (e.g. native herbivores species richness
~ habitat*biological control agent abundance), and if
the contribution of the biological control agent abundance is significant to the fit of the model, we can conclude that the analysed variable is being affected by the
agent.
Banasek-Richter, C., Cattin, M.F. and Bersier, L.F. (2004)
Sampling effects and the robustness of quantitative and
qualitative food-web descriptors. Journal of Theoretical
Biology 226, 2332.
Bascompte, J., Jordano, P., Melian, C.J. and Olesen, J.M.
(2003) The nested assembly of plantanimal mutualistic
networks. Proceedings of the National Academy of Sciences of the United States of America 100, 93839387.
Bersier, L.F., Banasek-Richter, C. and Cattin, M.F. (2002)
Quantitative descriptors of food-web matrices. Ecology
83, 23942407.
Boettner, G.H., Elkinton, J.S. and Boettner, C.J. (2000) Effects of a biological control introduction on three nontarget native species of saturniid moths. Conservation Biology 14, 17981806.
Buckley, Y.M., Rees, M., Sheppard, A.W. and Smyth, M.J.
(2005) Stable coexistence of an invasive plant and biological control agent: a parameterised coupled plant
herbivore model. Journal of Applied Ecology 42, 7079.
Carvalheiro, L.G., Buckley, Y.M., Ventim, R., Fowler, S.V. and
Memmott, J. (2008) Apparent competition can compromise
the safety of highly specific biocontrol agents. Ecology Letters 11, 690700.
Cattin, M.F., Bersier, L.F., Banasek-Richter, C., Baltensperger,
R. and Gabriel, J.P. (2004) Phylogenetic constraints and adaptation explain food-web structure. Nature 427, 835839.
Dunne, J.A., Williams, R.J. and Martinez, N.D. (2002) Network structure and biodiversity loss in food webs: robustness increases with connectance. Ecology Letters 5,
558567.
Fowler, S.V., Syrett, P. and Hill, R.L. (2000) Success and
safety in the biological control of environmental weeds in
New Zealand. Austral Ecology 25, 553562.
Hawkins, B.A. and Goeden, R.D. (1984) Organization of
a Parasitoid Community Associated with a Complex of
Galls on Atriplex Spp in Southern-California. Ecological
Entomology 9, 271292.
Henneman, M.L. and Memmott, J. (2001) Infiltration of a
Hawaiian community by introduced biological control
agents. Science 293, 13141316.
Holt, R.D. and Lawton, J.H. (1994) The Ecological Consequences of Shared Natural Enemies. Annual Review of
Ecology and Systematics 25, 495520.
Conclusions
Insects form numerous key links with other species,
leading to complex networks of interactions. To fully
assess the post-release impacts of an introduced biological control agent, community-level studies involving
quantitative data are needed. The recent practical and
theoretical advances made in food-web construction
and analysis allows wider applications in the field of
conservation biology, such as the assessment of biological control impacts. The food-web approach suggested
in this work will provide a post-release impact assessment in an understandable, applicable form for both
biological control practitioners and site managers. In
addition, although the methodology proposed here allows the assessment of post-release impacts, it is advisable that community-level studies are also done before
the release of the biological control agents. This would
reveal the native communitys pre- and post-invasion
85
XII International Symposium on Biological Control of Weeds
Lewis, O.T., Memmott, J., Lasalle, J., Lyal, C.H.C., Whitefoord, C., Godfray, H.C.J. (2002) Structure of a Diverse
Tropical Forest InsectParasitoid Community. The Journal of Animal Ecology 71 (5), 855873.
Louda, S.M., Kendall, D., Connor, J. and Simberloff, D.
(1997) Ecological effects of an insect introduced for the
biological control of weeds. Science 277, 10881090.
Memmott, J. (1999) The structure of a plant-pollinator food
web. Ecology Letters 2, 276280.
Memmott, J. and Godfray, H.C.J. (1994) The use and construction of parasitoid webs. In: Hawkins, B.A. and Sheehan, W. (eds) Parasitoid Community Ecology. Oxford
University Press, Oxford, pp. 300318.
Memmott, J., Godfray, H.C.J. and Gauld, I.D. (1994) The
Structure of a Tropical Host Parasitoid Community. Journal of Animal Ecology 63, 521540.
Memmott, J., Waser, N.M. and Price, M.V. (2004) Tolerance
of pollination networks to species extinctions. Proceedings of the Royal Society of London Series B-Biological
Sciences 271, 26052611.
Memmott, J., Gibson, R.H., Carvalheiro, L.G., Heleno, R.,
Henson, K.S.E., Lopezaraiza, M.E. and Pearce, S. (2007)
The Conservation of Ecological Interactions. In: Stewart, A.J.A., Lewis, O.T. and New, T.R. (eds) Insect Conservation Biology. CABI Publishing, Wallingford, UK,
pp. 226244.
Morris, R.J., Muller, C.B. and Godfray, H.C.J. (2001) Field
experiments testing for apparent competition between
primary parasitoids mediated by secondary parasitoids.
Journal of Animal Ecology 70, 301309.
Morris, R.J., Lewis, O.T. and Godfray, H.C.J. (2004) Experimental evidence for apparent competition in a a tropical
forest food web. Nature 428, 310313.
Morris, R.J., Lewis, O.T. and Godfray, H.C.J. (2005) Apparent
competition and insect community structure: towards a spatial perspective. Annales Zoologici Fennici 42, 449462.
Muller, C.B. and Godfray, H.C.J. (1997) Apparent competition between two aphid species. Journal of Animal Ecology 66, 5764.
Muller, C.B. and Godfray, H.C.J. (1999) Indirect interactions
in aphidparasitoid communities. Researches on Population Ecology 41, 93106.
86
Muller, C.B., Adriaanse, I.C.T., Belshaw, R. and Godfray,
H.C.J. (1999) The structure of an aphidparasitoid community. Journal of Animal Ecology 68, 346370.
Pemberton, R.W. and Strong, D.R. (2000) Safety data crucial
for biological control insect agents. Science 290, 1896
1897.
Schmitz, D.C. and Simberloff, D. (1997) Biological invasions: A growing threat. Issues Science and Technology
13, 3340.
Sheppard, A.W., Shaw, R.H., and Sforza, R. (2005) Top 20
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 93117.
Sole, R.V. and Montoya, J.M. (2001) Complexity and fragility in ecological networks. Proceedings of the Royal
Society of London Series B-Biological Sciences 268, 2039
2045.
Strong, D.R. (1997) EcologyFear no weevil? Science 277,
10581059.
Strong, D.R., Lawton, J.H., and Southwood, T.R.E. (1984)
Insects on Plants. Blackwell Scientific Publications, Oxford, 313 pp.
Thomas, M.B. and Willis, A.J. (1998) Biocontrolrisky but
necessary? Trends in Ecology & Evolution 13, 325329.
Tompkins, D.M., Draycott, R.A.H. and Hudson, P.J. (2000)
Field evidence for apparent competition mediated via the
shared parasites of two gamebird species. Ecology Letters
3, 1014.
Waser, N.M., Chittka, L., Price, M.V., Williams, N.M. and
Ollerton, J. (1996) Generalization in pollination systems,
and why it matters. Ecology 77, 10431060.
Willis, A.J. and Memmott, J. (2005) The potential for indirect
effects between a weed, one of its biocontrol agents and
native herbivores: A food web approach. Biological Control 35, 299306.
Wirf, L. (2006) Using simulated herbivory to predict the
efficacy of a biocontrol agent: the effect of manual defoliation and Macaria pallidata Warren (Lepidoptera:
Geometridae) herbivory on Mimosa pigra seedlings Australian Journal of Entomology 45, 324326.
Zuur, A.F., Ieno, E.N. and Smith, G.M. (2007). Analysing
Ecological Data. Springer, 680 pp.
Factors affecting oviposition rate
in the weevil Rhinocyllus conicus on
non-target Carduus spp. in New Zealand
R. Groenteman,1,2 D. Kelly,1 S.V. Fowler2 and G.W. Bourdt3
Summary
Adults of the nodding thistle receptacle weevil, Rhinocyllus conicus (Froehlich) (Coleoptera: Curculionoidae), oviposit on developing thistle flower buds. Larval feeding on the receptacle prevents seed
development. The weevil is known to attack several thistle species, but has clear preference for nodding thistle, Carduus nutans L. The effects of plant characteristics on oviposition preference and/or
the size of emerging adult weevils were examined on winged and slender-winged thistles (Carduus
tenuiflorus Curtis and C. pycnocephalus L., respectively). The results indicate that larger, higher seed
heads on larger plants are preferred for oviposition. Larger seed heads supported the development of
larger adults. This paper is part of a study looking at ecological aspects of non-target effects of thistle biological control in New Zealand. Nodding thistle flowers over an extended period of time but
the two winged thistle species offer additional oviposition opportunities three to four weeks before
nodding thistle flowers. The adults emerging from the winged thistle species are likely to establish a
second generation, enabling this normally univoltine weevil to sustain seasonally prolonged attack on
nodding thistle. Thus, proximity in space, combined with separation in time of closely related weed
species, potentially enhances the performance of the oligophagous R. conicus as a biocontrol agent
of all three thistle species.
Keywords: Carduus nutans; C. tenuiflorus; C. pycnocephalus; beneficial non-target
effects.
Introduction
In New Zealand, there are no native plants in the
tribe Carduae, and the only crop plant belonging to the
tribe, globe artichoke, is of minor economic importance
(Paynter et al., 2004). Thus, any thistle here if not
a weed already could potentially become one in the
future; and any impact a biological control agent such
as R. conicus may have on any thistle here is, therefore,
desirable.
We hypothesized that since R. conicus is highly attracted to C. nutans and is not likely to leave patches of
C. nutans, then less preferred (non-target) thistle species
are more likely to be attacked by the weevil in patches
they share with C. nutans, than in patches from which
C. nutans is absent. We selected two species that are
closely related to C. nutans as non-target species; they
were Carduus tenuiflorus Curtis and C. pycnocephalus
L., and we tested the level to which they were attacked
by R. conicus when C. nutans was present in close proximity, versus when it was absent. We also examined
what plant characteristics may affect R. conicus oviposition rate among these species, in an attempt to separate this effect from that of C. nutans proximity.
Rhinocyllus conicus (Froehlich) (Coleoptera: Curculionoidae), nodding (musk) thistle receptacle weevil, is
known to attack different thistle species but displays
a clear preference for nodding thistle, Carduus nutans
L. (Zwlfer and Harris, 1984). Eggs are laid on the developing flower buds, and the larvae burrow their way
into the receptacle, where their feeding prevents seed
formation (Zwlfer and Harris, 1984). The weevil, introduced to New Zealand in 1972 (Jessep, 1989), was
the first agent deployed here for nodding thistle biological control.
School of Biological Sciences, University of Canterbury, Private Bag
4800, Christchurch, New Zealand.
2
Landcare Research, Gerald St, PO Box 40, Lincoln, 7640, New Zealand.
3
AgResearch Lincoln, PO Box 4749, Christchurch 8140, New Zealand.
Corresponding author: R. Groenteman <[email protected]>
CAB International 2008
1
87
XII International Symposium on Biological Control of Weeds
Materials and methods
Carduus pycnocephalus, C. tenuiflorus, and C. nutans
plants were grown in the University of Canterbury
glasshouses from seeds collected in summer 2005
around Canterbury. In late autumn 2005 they were
transplanted as small rosettes to the experimental site
at the Landcare Research Lincoln campus.
At the experimental site, non-target species rosettes
were organized into plots, each containing five C. tenui
florus plants, five C. pycnocephalus plants and five Cir
sium vulgare plants (not discussed further) in a three x
five randomised arrangement, with 0.2 m between plants
and 3 m between plots, for a total of 24 plots. Plots
were grouped in blocks of four plots each, with at least
5 metres between blocks. Two plots in each block were
randomly assigned a C. nutans present treatment,
and in these, ten C. nutans rosettes were planted 0.2 m
apart, in two rows, 0.4 m away from the non-target
plants. The experiment was designed with six blocks
(replicates), with four treatments per block (C. nutans
present/absent and cage present/absent, to control for
weevil density), but the cage treatment was effectively
cancelled due to difficulty obtaining sufficient weevils
early enough in the season, when the non-target plants
were already bolting. Weevils populated the experiment
independently from the surrounding environment, and
thus, R. conicus densities were not controlled.
The longest leaf of each rosette was measured at
bolting, and plant height, number of clusters, their position, and number of capitula, were recorded fortnightly.
From each plot, one plant of each non-target species
was randomly selected, from which all the ripe capitula
were collected fortnightly and placed individually into
paper bags. The height and number of capitula in the
cluster were recorded for these individually collected
capitula. The receptacle diameter was measured for
each capitulum and any sign of attack by R. conicus was
recorded. For each non-target species, the difference in
R. conicus attack signs per capitulum (log transformed)
were compared between treatments (C. nutans present
vs. absent) in mixed models with Poisson distribution
using R (R Development Core Team, 2006); all different plant and capitulum variables were included in
the models as covariates, and non-significant variables
were excluded in backwards selection using the Chisquared test to compare between models.
Emerging R. conicus adults were sexed and elytra
length was measured. The length was then compared be-
Table 1.
tween sexes and between thistle species using an F test
(which included a comparison to adults that emerged
from early C. nutans capitula in the experiment).
Results and discussion
Rhinocyllus conicus attack (number of attack signs per
capitulum) on the non-target species did not differ be
tween plots with and without C. nutans (Table 1). C.
pycnocephalus was attacked more than C. tenuiflorus
(Table 1).
On both C. tenuiflorus and C. pycnocephalus, R.
conicus preferred larger capitula on larger plants for
oviposition, and attack on these species was stronger
early in the spring, decreasing towards summer (equations Ac.t and Ac.p below).
The equations describing the effects of the measured
covariates on R. conicus attack (number of signs per
capitulum) for C. tenuiflorus (Ac.t ) and C. pycnocepha
lus (Ac.p) are, respectively:
Ac.t = e(3.6 1.01*day 0.48*c. height + 0.46*p. height + 0.35*diameter + 0.09*
cl. size + 0.002*cap)
Ac.p = e(2.24 2.97*tertiary 0.90*secondary + 0.52*diameter 0.44*day + 0.29*
p. height 0.11*c. height + 0.006*cap)
;
where day is day of the year on which the capitulum
was collected, c. height is height of the individual capitulum (mm), p. height is maximum height of the
plant at the time of collection (mm), diameter is external diameter of the individual capitulum (mm), cl.
size is cluster size (number of capitula in the cluster
from which the individual capitulum was collected),
cap is total number of capitula on the plant, tertiary
is for a capitulum collected from a tertiary cluster, and
secondary is for a capitulum collected from a secondary cluster. Similar response of R. conicus to plant and
capitulum size was found on C. nutans (Sheppard et
al., 1994; Groenteman et al., 2007; but see McNeill
and Fletcher, 2005).
In New Zealand, R. conicus adults emerge from
overwintering around November (spring), when C.
tenuiflorus and C. pycnocephalus are bolting, but three
to four weeks before the first C. nutans capitula are
formed (Jessep, 1989). In those mid-spring weeks, R.
conicus attack on the non-targets peaked at 35.7 1.3%
of capitula being attacked. Interestingly, even when
C. nutans became available, oviposition on the non-
targets did not cease completely, and 3.4 0.8% of the
Number of Rhinocyllus conicus attack signs per capitulum (back-transformed from log-scale means with 95% CI).
Thistle species
Treatment
Carduus nutans present
Carduus nutans absent
Carduus tenuiflorus
0.025 (0.0150.042)
0.028 (0.0170.05)
88
Carduus pycnocephalus
0.11 (0.070.17)
0.10 (0.060.16)
Factors affecting oviposition preference in the weevil Rhinocyllus conicus on non-target Carduus spp. in New Zealand
capitula still were attacked. In contrast, 100% of early
C. nutans capitula were attacked. Carduus tenuiflorus
and C. pycnocephalus capitula are considerably smaller
than C. nutans capitula (diameters 5.66 0.018 mm,
7.23 0.019 mm, and 19.73 0.098 mm, respectively),
and they received up to 4 and 6 eggs per capitulum,
respectively (with means of 0.17 0.007 and 0.25
0.009 for C. tenuiflorus and C. pycnocephalus, respectively). In Southern California, a biotype of R. conicus
specialized on C. pycnocephalus was reported to make
up to 17 (mean 1.7) penetration holes per capitulum
on that plant (Goeden and Ricker, 1985). Early C. nu
tans capitula are known to receive many dozens of
eggs each (Jessep, 1989; Woodburn, 1996). In accordance with oviposition levels and capitulum size, C.
tenuiflorus and C. pycnocephalus were usually found
to support no more than one adult per capitulum (occasionally two, and rarely three, adults per capitulum),
which is comparable to the C. pycnocephalus specialized biotype (with mean 0.9 adults per capitulum; Goeden and Ricker, 1985), whereas C. nutans capitula can
support the successful development of dozens of adults
(R. Groenteman, personal observations). Capitulum
size appeared also to affect the size of adults, with C.
tenuiflorus producing the smallest adults, followed by
C. pycnocephalus, and C. nutans with the largest individuals (elytra length 3.13 0.05 mm, 3.21 0.04
mm, and 3.49 0.02 mm respectively, F2,1058 = 38.63,
P < 0.0001).
Although in this experiment the non-targets C. ten
uiflorus and C. pycnocephalus did not appear to be
attacked more in the presence of C. nutans, this was
probably due to the spatial scale of the experiment
which, considering the distance flown by the weevils
to reach the site, was negligible. Therefore, the plots
could not be truly considered far enough apart to create
C. nutans-present and C. nutans-absent treatments.
Legal constraints regarding C. nutans propagation prevent any large scale manipulative experimentation on
the species in New Zealand. Rhinocyllus conicus has
not been considered a long distance disperser, mainly
due to lack of knowledge (Sezen, 2007), and hence this
was not considered a problem at the time the experiment was designed.
A field survey on a larger scale has revealed that
non-targets are attacked by R. conicus more in the presence of C. nutans than in its absence (R. Groenteman,
unpublished data). Furthermore, at times closer to the
introduction of the weevil to New Zealand, C. tenui
florus and C. pycnocephalus were commonly growing
close to C. nutans (Jessep, 1989). Currently, a population of C. nutans in sympatry with either C. tenuiflorus
or C. pycnocephalus is hard to find (R. Groenteman,
personal observation). It may be that the biological control agent has, in the many years since its introduction,
successfully reduced populations of the non-targets
where they were adjacent to C. nutans. The non-targets
89
are still abundant in New Zealand; only, not so in proximity to C. nutans.
To conclude, although we were unable to show
that the non-target species were attacked more in the
presence of the target species, we have found that
R. conicus effectively detects patches of the non-target
species, and perhaps does not leave them quickly if
C. nutans is in close proximity. Thus, the separation in
time combined with the proximity in space is possibly
improving biocontrol for all of these species: The nontargets are attacked by the less-likely-to-leave weevils
early in the season, thus producing a second generation
of R. conicus. This second generation attacks C. nutans
later in the season, after the first generation has already
completed its life cycle, and when C. nutans is in its
peak flowering in New Zealand.
Acknowledgements
We thank Morgan Coleman, Emry Dolev, Caroline
Thomson, Alex Groenteman, Raviv Carasuk, Nicolette LeCren, Helen Parish, Lindsay Smith, Eyal Twig,
and Amit Yigal for field and lab assistance; and Dave
Saville and Richard Duncan for statistical advice. The
project was funded by the New Zealand Foundation for
Research Science and Technology.
References
Goeden, R.D. and Ricker, D.W. (1985) Seasonal asynchrony
of Italian thistle, Carduus pycnocephalus, and the weevil,
Rhinocyllus conicus (Coleoptera, Curculionidae), introduced for biological control in southern California. Envi
ronmental Entomology 14, 433436.
Groenteman, R., Kelly, D., Fowler, S.V. and Bourdt, G.W.
(2007) Interactions between nodding thistle seed predators. New Zealand Plant Protection 60, 152157.
Jessep, C.T. (1989) Carduus nutans L., nodding thistle
(Asteraceae). In: Cameron, P.J., Hill, R.L., Bain, J. and
Thomas, W.P. (eds) A Review of Biological Control of In
vertebrate Pests and Weeds in New Zealand 1874 to 1987.
CAB International, Wallingford, UK, pp. 339342.
McNeill, M.R. and Fletcher, C.J. (2005) Interspecific competition between Rhinocyllus conicus and Urophora solsti
tialis L. on nodding thistle in Canterbury? New Zealand
Plant Protection 58, 140147.
Paynter, Q.E., Fowler, S.V., Gourlay, A.H., Haines, M.L.,
Harman, H.M., Hona, S.R., Peterson, P.J., Smith, L.A.,
Wilson-Davey, J.R.A., Winks, C.J. and Withers, T.M.
(2004) Safety in New Zealand weed biocontrol: a nationwide survey for impacts on non-target plants. New Zea
land Plant Protection 57, 102107.
R Development Core Team (2006) R: A language and envi
ronment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
Sezen, Z. (2007). Interactions of the invasive thistle Carduus
nutans and its biocontrol agent Rhinocyllus conicus in
heterogeneous environments. PhD thesis. Pennsylvania
State University, University Park, PA.
XII International Symposium on Biological Control of Weeds
the IX International Symposium on Biological Control of
Weeds. Stellenbosch, South Africa, pp. 409415.
Zwlfer, H. and Harris, P. (1984) Biology and host specificity of Rhinocyllus conicus (Froel) (Col, Curculionidae), a
successful agent for biocontrol of the thistle, Carduus nu
tans L. Zeitschrift Fur Angewandte Entomologie-Journal
of Applied Entomology 97, 3662.
Sheppard, A.W., Cullen, J.M. and Aeschlimann, J.P. (1994)
Predispersal seed predation on Carduus nutans (Asteraceae) in southern Europe. Acta Oecologica 15, 529541.
Woodburn, T.L. (1996) Interspecific competition between
Rhinocyllus conicus and Urophora solstitialis, two biocontrol agents released in Australia against Carduus nutans.
In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of
90
Fortieth anniversary review of the
CSIRO European Laboratory:
does native range research provide
good return on investment?
A.W. Sheppard,1 D.T. Briese,1 J.M. Cullen,1 R.H. Groves,2 M.H. Julien,3
W.M. Lonsdale,1 J.K. Scott4 and A.J. Wapshere5
Summary
CSIRO established its first overseas research laboratory on biological control at Montpellier in late
1966 to start a programme on skeleton weed, Chondrilla juncea L.). The laboratory was set up to
develop the science to underpin effective biological control, by parallel studies in the native and
introduced range of Australias pests. Since establishment within a French research agency (CNRS),
the facility moved in 1994 from rented facilities into a purpose-built CSIRO-owned facility, with
support from Australian industry bodies and the French government. This facility has been CSIROs
largest long-term overseas investment in research. The core focus on biological control of weeds has
been increasingly supplemented by other research activities that are not otherwise possible within
Australia. We present an economic and scientific review of the laboratory on its 40th anniversary. The
facility cost on average Aus$1.3 million (2006 $$) per year (67% on direct research activities and
33% of infrastructure and administration) and generated at least $27 benefit for Australia for every
$1 invested. Staff produced 279 publications of which 159 are in journals that are currently ISI rated
(average citation rate in 2007 was 14.8 per ISI journal paper).
Keywords: biological control, cost/benefit, foreign exploration, historical review.
Introduction
Classical biological control aims to suppress invasive
exotic pest populations by releasing specialist natural enemies, termed biocontrol agents, selected from
the native range of the pest, while generating no or acceptably low non-target impacts (see Briese, 2000a).
In this context, the native range of the target pest is the
source of most biocontrol agents. These agents need
to be found, identified, and any risks they may pose
following introduction assessed by exposing them to
native and commercially important species using a
centrifugal phylogenetic approach (Wapshere, 1974;
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia.
CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601, Australia.
3
CSIRO European Laboratory, Campus de Baillarguet, 34980 Montferrier-
sur-Lez, France.
4
CSIRO Entomology, Private Bag 5, PO Wembley, WA 6913, Australia.
5
Deceased November 2007.
Corresponding author. A.W. Sheppard <[email protected]>.
CAB International 2008
1
2
91
Briese, 2005) prior to importation into the invaded region. Countries, and their research agencies, around the
world that practise classical biological control of exotic
pests, tend to achieve this in one of three ways:
1. Scientific staff select potential biocontrol agents
overseas during visits to the pests native range and
import them to a quarantine facility for detailed assessment. The Plant Protection Research Institute in
South Africa has largely adopted this approach.
2. Contracted or collaborative research arrangements
are set up with a research-provider agency to conduct the native range aspects of the research, including exploration and risk assessment. Agriculture
Canada, certain US States, Landcare Research New
Zealand, Queensland (Australia) and many developing countries have often adopted this approach with
CABI as the dominant research provider.
a. Research agencies set up their own overseas research facilities in the native range of the pest
so that they can manage the whole biological
control programme and carry out risk assess-
XII International Symposium on Biological Control of Weeds
ment and efficacy evaluation overseas prior to,
or in conjunction with, agent importation and
release. The Australian government agency,
CSIRO, and USDA-ARS have largely followed
this approach. Australia has or has had its own
facilities in Europe, South Africa and Central
America and USDA-ARS has its own facilities
in Europe, South America, Asia and Australia.
The USDA-ARS Australian Biological Control
Research Laboratory is actually now run by
CSIRO. Similarly, CSIRO also contracts survey
work to the USDA South American Biological
Control Laboratory.
These three options represent a progression in the
scale of investment in native range research, the payoff for which has been historically argued as a) improving the chance of a successful biological control
programme (Gurr and Wratten, 2000) through singleagency management of risk and efficacy assessment,
release and establishment; b) greater understanding of
the target through comparative ecological and genetic
studies between the invaded and native range; c) greater
understanding of the ecology of potential biocontrol
agents and interactions with the target; and d) opportunities for enhancing the general scientific-basis of both
the agent selection based on prediction of efficacy, and
agent risk assessment components of biological control
programmes. Additionally, the location of an agency
outpost in another country offers compelling opportunities for increased international collaboration.
No previous review has evaluated any of these three
options, either individually or collectively, for either
a) economic return on investment, or b) scientific performance and outputs. In November 2006, the CSIRO European Laboratory (CSIRO-EL) celebrated its 40th year
in Montpellier, France. This paper reviews the economic returns and scientific performance of this facility
over the 40 years, including a short history of CSIROEL during that period.
Research activities
Weed biological control projects undertaken at the facility in its 40 years include 30 targets across 27 genera
of weeds (Table 1). Seven projects focusing on genera
(Rubus, Onopordum, Fumaria, Reseda, Sonchus, Vulpia and Convolvulus). At least 73 species of potential
agents, including 11 plant pathogens, were tested. Forty-
two species of agents were released in Australia, including four plant pathogens. These were the rusts Puccinia
chondrillina Bubak & Sydenham (four strains) for control of Chondrilla juncea L.; Phragmidium violacium
(Shultz) Winter (nine strains) for Rubus spp.; Uromyces heliotropii Sred. for Heliotropium europaeum L.;
and Puccinia cardui-pycnocephali Sydow for Carduus
pycnocephalus L. and Carduus tenuiflorus Curtis.
The chronology of these releases is given in Figure 1.
92
Three biocontrol projects targeting insects and two
targeting snails also led to releases of three insect
biocontrol agents [two against Sitona weevil and one
against Mediterranean snails, Theba pisana (Mller)]
in Australia. Work at the facility also contributed to
Australias highly successful dung beetle project that is
widely accepted to have both increased carbon and nitrogen cycling and to have lessened the public nuisance
from bush flies (Edwards, 2007). Six European dung
beetle species were shipped to Australia, of which Onthophagus taurus Schreber, Euoniticellus fulvus Goeze,
E. pallipes (Fabricius), Geotrupes spiniger Marsham,
and Copris hispanus L. established. Research has also
been conducted, pre-emptively, against two insect pests
that threaten but have yet to arrive in Australia: Russian
wheat aphid, Diuraphis noxia (Mordvilko), and Asian
gypsy moth, Lymantria dispar L.
Project benefits and costs
Of the 33 classical biological control research programmes undertaken at the Montpellier facility, 11 are
considered to have led to some level of success in measurable economic terms (Table 1). Most successful biological control programmes worldwide have not been
subjected to economic benefit/cost analysis. Fortunately, many of the biological programmes undertaken at
the facility were included in a recent economic analysis
of weed classical biological control programmes across
Australia (Page and Lacey, 2006). This report can be
criticized for often using very limited or subjective
data, but for most programmes these analyses are the
best available. Certain biological control programmes
had relatively in-depth economic assessments conducted prior to the report, particularly the programmes
against C. juncea (Cullen, 1976; Marsden et al., 1980),
Echium plantagineum L. (IAC, 1985), Carduus nutans
L. (Young and Woodburn, 2002) and Onopordum thistles (Meat & Livestock Australia, 1993, unpublished
data). In most cases, Page and Lacey (2006) took these
assessments into account, thereby ensuring more rigorous analysis and sounder conclusions. Certain biocontrol programmes based at the facility have never
been economically evaluated, however, despite widespread agreement of realized benefits. These include
programmes targeting Rumex pulcher L. and Hypericum perforatum L., and other programmes not targeting weeds, i.e. the dung beetle programme (Edwards,
2007).
CSIRO European Laboratory
benefit/cost analysis
Only the available benefits published in Page and
Lacey (2006) could be used for a benefit assessment
of the research conducted at CSIRO European Laboratory. These overall published programme benefits were
Table 1.
Projects at CSIRO-EL between 1967 and 2007 arranged by type and chronological order, giving the years of activity, the biological control agents studied and released, and the estimated total programme (not just CSIRO-EL
component) benefits and costs (from Page and Lacey, 2006) where known.
Target pest
Weeds biocontrol
Chondrilla juncea
Echium plantagineum
Senecio jacobaea
Hypericum perforatum
Heliotropium europaeum
Rubus spp.
Dates of project
at CSIRO-EL
(approx.)
No. of biocontrol
agents tested
No. of biocontrol
agents released
19671999
19731983a
19871996
19781987
19781989
19781982
19911993
19801984
19992007
19811990
19851990
19861990
19861995
19861995
4
10
Rumex pulcher
Asphodelus fistulosus
Emex spinosa
Carduus nutans
Carduus tenuiflorus/
C. pycnocephalus
Cirsium arvense
19861990
Convolvulus spp.
19891990
Onopordum spp.
19872000
Cytisus scoparius
19892002
Cirsium vulgare
19901995
Marrubium vulgare
19911995
Silybum marianum
19921994
Carthamus lanatus
19921999
Reseda spp.
19951999
Vulpia spp.
19961999
Lepidium latifoliume
19992000
Raphanus raphanistrum
19992002
Genista monspessulana
19992007
Fumaria spp.
20032005
Sonchus spp.
20032005
Ulex europaeus
20032007
Phyla canescens
2006
???/
Xanthium strumarium
Invertebrate biocontrol
Sitona spp.
19751986
(Lucerne weevils)
Cochlicella acuta (snail)
19891994
Theba pisana (snail)
19891994
Nezara viridulae
2006
(green vegetable bug)
Maconellicoccus hirsutuse
20002002
(pink hibiscus mealybug)
Other biocontrol projects
Dungd
19771984
Invertebrate risk assessment
Diuraphis noxia (Russian
19891991
wheat aphid)
2001
Lymantria dispar g
19992004
(Asian gypsy moth)
Project generated a benefit?
Benefit
M$
4
7
Yes
Yes
1425.7
1201.0
12.7
23.0
5
6+1b
4
4
6+1
2
Yes
Yes
No
97.2
Nd
0
3
3
1.9
1c
Yes
6.1
2.4
2
1
1
5
2
1
0
1
3
2
Yes
No
No
Yes
Yesd
Nd
0
0
81.3
22.5
1.3
0.2
0.2
11.9
1.6
1
1
8
8
3
3
2
1
1
0
0
2
3
1
0
0b
0
1
0
0
7
3
3
2
0
0
0
0
0
0
1c
0
0
0
0
0
No
No
Yes
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
0
0
25.0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1
0.1
7.0
7.0
1.5
1.2
0.5
0.4
0.2
0.5
0.2
0.2
3.0
0.1
0.1
0.1
0.
0.1
No
1.0
3
3
0
1
1
0
0
0
n/a
0.3
0.3
0.1
No
No
Evaluation
research
Evaluation
research
n/a
0.1
6+
Yes
Nd
0.6
n/a
n/a
n/a
0.6
n/a
n/a
Pre-emptive
research
Pre-emptive
research
n/a
0.3
Cost
M$
Court injunction caused gap in research.
Built on previous research in 1950s.
c
Accidental releases.
d
Work moved to Cordoba in Spain from 19841987; although Page and Lacey analyse a net benefit from this project, most of the authors
dispute any real magnitude to the claimed success.
e
U.S. target contracted research.
f
Australian weed of S. American origin, but ecological and genetic studies are being undertaken using invasive populations in France.
g
For Australia and New Zealand.
a
93
XII International Symposium on Biological Control of Weeds
16
pathogen
# Agents released in AU
14
insect
12
10
8
Planned
6
4
2
0
19701974
Figure 1.
19751979
19801984
19851989
19901994
19951999
20002004
20052009
The number of weed biological control agents released into Australia from CSIROEL by 5-year period (data from Julien and Griffiths, 1998).
During the 40 years, the facility has had eight officers-
in-charge (OIC), 18 other scientists, 39 project officers,
7 administration staff, 14 casual staff, 34 undergraduate students, 21 postgraduate students and 17 scientific
visitors (based at the facility for at least several months).
Annual expenditures of the facility by CSIRO (set at
2006 Aus$ values using published historic consumer
price index data) are presented in Figure 2. These
expenditures from 198687 onwards were calculated
halved to reflect the contribution made by native range
research at the facility; the remainder considered to result from the Australian-based activities. Estimates of
the benefits to Australia from the research conducted
at the facility are therefore conservative. Costs for the
facility were based on estimates of the direct costs of
the facility and full research project costs, rather on the
project costs associated with each of the relevant programmes as used in Page and Lacey (2006).
4.0
External research funding
3.5
CSIRO research expenditure
2006 $A millions
3.0
Administration expenditure
2.5
2.0
1.5
1.0
Figure 2.
2006-07
2005-06
2004-05
2003-04
2002-03
2001-02
2000-01
1999-00
1998-99
1997-98
1996-97
1995-96
1994-95
1993-94
1992-93
1991-92
1990-91
1989-90
1988-89
1987-88
1986-87
1985-86
1984-85
1983-84
1982-83
1981-82
1980-81
1979-80
1978-79
1977-78
1976-77
1975-76
1974-75
1973-74
1972-73
1971-72
1970-71
1969-70
1968-69
1967-68
0.0
1966-67
0.5
Estimated annual running costs of CSIRO-EL for a) external funding sources; b) CSIRO research
project expenditure; and c) administration/buildings costs (including external support for the new
laboratory constructed from 1992 to 1993) in 2006 Aus$.
94
Fortieth anniversary review of the CSIRO European Laboratory
based on financial records at the facility produced by
the OIC. Prior to that, financial data were only available
for the 198081 financial year. These data presented
detailed and research costs and salaries per Australian
and local project staff and per admin staff. These data
were used to estimate costs in intervening and earlier
years based on the quarterly records of the number of
staff employed at the facility.
Data on external funding from primary industry research and development corporations (RDCs) were also
available (Figure 2) and, where necessary, extrapolated
to years for which data were not available (19661980
and 19831986); although prior to 1985, apart from
some early funding from the Wheat Industry Research
Council, external funding was limited. The construction of the new CSIRO European Laboratory on the
Baillarguet campus between 1992 and 1993 resulted in
significant additional once-off costs of c.Aus$2 million
(equivalent to Aus$2.9 million in 2006 Aus$); 55%
came from Australian RDCs (facility construction)
and 30% from the Languedoc-Roussillon region (subsidies to servicing) and district of Montpellier (glasshouse construction).
The total conservative benefits of the research at
CSIRO European laboratory are presented in Figure 3
next to the total costs of the facility over its 40-year
life. The facility cost an average of Aus$1.33 million
(2006 Aus$) per annum to run per year, 67% of which
was spent on direct research activities and 33% on in-
frastructure and administrative costs. Research over the
40 years provided a Aus$1.43 billion benefit to Australian primary industries. This represents a benefit/cost
ratio of 27:1, a result that is similar to the overall 23:1
benefit/cost ratio from all weed biological control research in Australia as calculated by Page and Lacey
(2006). The CSIRO-European Laboratory has been
an effective research investment in the native range of
pests as part of Australias overall strategy for weed biological control. In addition, it has provided the basis for
successful programmes in other countries, e.g. Sitona
weevil in New Zealand.
Science performance
All publications from the CSIRO European Laboratory
over the 40-year period were collated, including papers
written by staff while at the facility and papers written
based on research carried out at the facility. The list
includes 279 publications of which 159 are in journals
that are currently ISI rated (Figure 4). Of these, 197
were research papers that addressed agent surveys and
taxonomy (11%), biology and host specificity (37%),
release and evaluation (8%), agent ecology (18%)
agenttarget interactions (8%), target ecology, genetics
and evolutionary biology (12%) and ecological theory
(3%). The remainder consisted of 51 reviews, 20 technical notes and 10 book chapters. These publications
were produced by 21 CSIRO staff at the facility and by
$1,600
$1,400
ragwort
thistles
2006 $A millions
$1,200
$1,000
Paterson's
curse
$800
$600
$400
skeleton
weed
$200
non-CSIRO
CSIRO
$0
Total Benefit
Figure 3.
Total Lab Cost
Total calculated benefits available from CSIRO-EL research in 2006 Aus$ divided by
target weed and total facility costs to CSIRO, divided into direct Australian Federal
Government and Primary Industry (+ French) funded components. The benefit/cost
ratio is 27:1.
95
XII International Symposium on Biological Control of Weeds
30
25
Other
ISI Journal papers
Scientists
4
15
3
# scientists
# Publications
5
20
10
2
5
19
6
19 9
7
19 0
7
19 1
72
19
7
19 3
7
19 4
7
19 5
7
19 6
77
19
7
19 8
7
19 9
8
19 0
8
19 1
8
19 2
8
19 3
8
19 4
8
19 5
8
19 6
87
19
8
19 8
8
19 9
9
19 0
9
19 1
9
19 2
9
19 3
9
19 4
9
19 5
9
19 6
9
19 7
9
19 8
99
20
0
20 0
0
20 1
0
20 2
0
20 3
0
20 4
0
20 5
06
Figure 4.
The number of publications by scientists based at CSIRO-EL or based on work predominantly done
at the facility since 1966. One hundred ten scientist years (FTEs per year 40) at the facility has
generated a total of 279 publications (seven per year), 156 (four per year) in journals that are currently ISI rated.
control research based on the research activities at the
Montpellier facility.
A benchmark paper on centrifugal phylogenetic approach to agent risk assessment (Wapshere, 1974) and his
highly cited paper on global plant invasions (Lonsdale,
1999) were written while their authors were Officers-
in-Charge. The facility produced the first successful
programme in classical biological control using a plant
pathogen (Cullen et al., 1973; Cullen, 1978, Burdon et
al., 1981). This innovative result opened the door to using plant pathogens as biological control agents around
the world (Cullen and Hasan, 1988; Barton, 2004; Morin
et al., 2006; Fisher et al., 2007). Some of the first work
on the genetic interactions between pathogen/insect and
hostplant genotypes in natural systems took place at
the facility (Michalakis et al., 1993; Chaboudez and
Burdon, 1995; Briese et al., 1996; Espiau et al., 1998).
Work at the facility led to many key papers in characterizing insect herbivore communities (Briese et al., 1994)
and the population ecology of insectplant interactions
(Sheppard et al., 1994; Briese, 1996, 2000b). Research
over many years comparing the ecology (Paynter et al.,
1998; Grigulis et al., 2001; Jongejans et al., 2006) and
evolution (OHanlon et al., 1999) of plants in their native European range with parallel work as invaders in
Australia was also carried out through the facility. Such
studies are now seen as a key approach to understanding invasion and biological control processes (Hinz
and Schwarzlaender, 2004; Hierro et al., 2005). Collaboration between the CSIRO European Laboratory,
CABI and Imperial College in the U.K. also led to an
integration of ecological modelling to better understand
such plant invasions (Rees and Paynter, 1997; Rees et
al., 1999) and their interactions with biological control
five visitors from other agencies. Up until 2007, these
papers had been cited 2,915 times on the ISI-cited reference database (since citation records began in 1985).
The journal papers from journals currently ISI rated
have an average citation rate of 14.8 per paper in comparison with the CSIRO-wide average of 7.9 for Agricultural Sciences and 11.6 for Ecology and Environment
(CSIRO, 2007). The CSIRO-European Laboratory has
also successfully produced four PhD, 11 MSc and at
least 34 undergraduate project dissertations.
Discussion
After 40 years, the CSIRO European Laboratory has
had a very significant impact on the control of invasive species of European origin in Australia as well
as a widely accepted significant impact to nutrient
recycling by way of its contribution to activities on
dung beetles (Edwards, 2007). The economic benefits
achieved from research conducted at the facility have
been twenty-seven times its total costs and the scientific
performance of the research in terms of research publications and citations is better than the relevant average
for CSIRO (CSIRO, 2007). Without similar reviews of
other less costly investment models used by biological
control research agencies for native range studies outlined in the Introduction, it is hard to evaluate whether
the higher costs of an overseas facility yield greater net
benefits or increased scientific impact. It would be hard
to argue that CSIRO has had a higher success rate in
biological control programmes than in countries adopting these other models, such as South Africa, Canada
and New Zealand. However, CSIRO has built a strong
track record and international reputation in biological
96
Fortieth anniversary review of the CSIRO European Laboratory
agents (Buckley et al., 2005). In summary, the scientific
returns to CSIRO of having a research facility in France
have been substantial and have heralded many collaborative projects with USDA-ARS, CABI, French and
European agencies and universities and the home agencies of visiting scientists that spent time at the facility.
Montpellier, Montferrier-sur-Lez and the LanguedocRoussillon region, USDA-ARS and many other agencies that have supported CSIRO-EL.
Future directions
Barton, J. (2004) How good are we at predicting the field
host-range of fungal pathogens used for classical biological control of weeds? Biological Control 31, 99122.
Briese, D.T. (1996) Potential impact of the stem-boring weevil Lixus cardui on the growth and reproductive capacity
of Onopordum spp. thistles. Biocontrol Science & Technology 6, 251261.
Briese, D.T. (2000a) Classical biological control. In: Sindel,
B. (ed.) Australian Weed Management Systems. Richardson Publications, Melbourne, Australia, pp. 161192.
Briese, D.T. (2000b) Impact of the Onopordum capitulum
weevil Larinus latus on seed production by its host plant.
Journal of Applied Ecology 37, 238246.
Briese, D.T. (2005) Translating host-specificity test results
into the real world: The need to harmonize the yin and
yang of current testing procedures. Biological Control 35,
208214.
Briese, D.T., Sheppard, A.W., Zwlfer, H., and Boldt, P.E.
(1994) The phytophagous insect fauna of Onopordum
thistles in the northern Mediterranean basin. Biological
Journal of the Linnean Society 53, 231253.
Briese, D.T., Espiau, C. and Pouchot-Lermans, A. (1996) Micro-
evolution in the weevil genus Larinus: host-race formation and speciation. Molecular Ecology 5, 531545.
Burdon, J.J., Groves, R.H. and Cullen, J.M. (1981) The impact of biological control on the distribution and abundance of Chondrilla juncea in southeastern Australia
Journal of Applied Ecology 18, 957966.
Buckley, Y.M., Rees, M., Sheppard, A.W. and Smyth, M.
(2005) Stable coexistence of an invasive plant and biocontrol agent: a coupled plantherbivore model. Journal
of Applied Ecology 42, 7079.
Chaboudez, P. and Burdon, J.J. (1995) Frequency-dependent
selection in a wild plantpathogen system, Oecologia
102, 490493.
CSIRO (2007) Science Health Report for 200506. CSIRO
Australia. 27 p.
Cullen, J.M. (1978) Evaluating the success of the programme
for the biological control of Chondrilla juncea L. In: Feeman, T.E. (ed.) Proceedings of the IV International Symposium on Biological Control of Weeds. University of
Florida, Gainesville, FL, pp. 117121.
Cullen, J.M. and Hasan, S. (1988) Pathogens for the control
of weeds. Philosophical Transactions of the Royal Society
of London 318, 213224.
Cullen, J.M., Kable, P.F. and Catt, M. (1973) Epidemic
spread of a rust imported for biological control. Nature
244, 462464.
Edwards, P. (2007) Introduced dung beetles in Australia
19672007: current status and future directions. Landcare
Australia, unpublished report. http://www.landcareonline.
com/resource.asp?rcID=9.
Espiau, C., Rivire, D., Burdon, J.J., Gartner, S., Daclinat,
B., Hasan, S. and Chaboudez, P. (1998) Hostpathogen
diversity in a wild system: Chondrilla juncea-Puccinia
chondrillina. Oecologia 113, 133139.
References
The data presented in Figures 1, 2 and 4 show a decline
in activity at the facility since a peak in the early 1990s.
A number of factors contributed to this decline. First,
Europe and the surrounding Old World countries are
declining in importance as a source of invasive pests for
Australia. Continents such as Asia, South America and
Africa are the contemporary sources of many of Australias weeds and pests. Second, the funding streams
for projects that require overseas research activities are
increasingly hard to attract, as governments and Rural
Development Corporations seek short-term, sometimes
politically motivated, measurable impacts and returns.
A funding crisis throughout the 1990s for biological
control research based on its inherent risk and longterm horizons has been widely recognized (Sheppard
et al., 2003), including by the Australian Weeds Committee, and has yet to be resolved. Third, the costs of
overseas facilities have, along with the costs of scientific research generally, increased enormously, making
it harder to provide sustainable project funding for a
small laboratory restricted to research not possible in
Australia. Increasing stringent occupational health and
safety standards make increasingly expensive purposebuilt research facilities a far more rational option for
todays scientific needs than the rented premises used
by the CSIRO Biological Control Unit for 13 years
through the 1970s to 1980s. Finally, direct research
collaboration between international research agencies
is now the norm, through staff exchanges and the sharing of research facilities. Permanent overseas facilities
are often considered too inflexible to accommodate the
ever-changing international collaborative interests of
research scientists. Agencies with their own overseas
facilities appear increasingly isolationist to the modern
scientific community. Nonetheless, the management of
exotic weeds and pests and preparing for the increasing biosecurity threats associated with increased world
trade ensures the maintenance of an overseas capability
to undertake research on pest species before they arrive.
In this latter-day context, CSIRO European Laboratory
should increasingly represent a keystone to Australias
future biosecurity strategy.
Acknowledgements
CSIRO would like to thank the Australian Federal and
State governments, Grains Research and Development
Corporation, Meat & Livestock Australia, Australian
Wool Innovation, Australian Weed Management CRC,
97
XII International Symposium on Biological Control of Weeds
Morin, L., Evans, K.J. and Sheppard, A.W. (2006) Selection
of pathogen agents in weed biological control: critical issues and peculiarities in relation to arthropod agents. Australian Journal of Entomology 45, 349365.
OHanlon, P.C., Peakall, R. and Briese, D.T. (1999) AFLP
reveals introgression in weedy Onopordum thistles: hybridisation and invasion. Molecular Ecology 8, 12391246.
Page, A.R. and Lacey, K.L. (2006) Economic Impact Assessment of Australian Weed Biological Control. CRC for
Australian Weed Management technical series #10, University of Adelaide, Australia, 151 p. (http://www.weeds.
crc.org.au/documents/tech_series_10.pdf)
Paynter, Q., Fowler, S.V., Memmott, J. and Sheppard, A.W.
(1998) Factors affecting the establishment of Cytisus scoparius in southern France: implications for its control.
Journal of Applied Ecology 35, 582595.
Rees, M. and Paynter, Q. (1997) Biological control of Scotch
broom: modelling the determinants of abundance and the
potential impact of introduced insect herbivores. Journal
of Applied Ecology 34, 12031221.
Rees, M., Sheppard, A.W., Briese. D.T. and Mangel, M.
(1999) Evolution of size dependent flowering in Onopordum illyricum: a quantitative assessment of the role of
stochastic selection processes. American Naturalist 154,
628651.
Sheppard, A.W., Cullen, J.M. and Aeschlimann, J.-P. (1994)
Predispersal seed predation on Carduus nutans (Asteraceae) in southern Europe. Acta Oecologica 15, 529541.
Sheppard, A.W., Hill, R., DeClerck-Floate, R.A., McClay, A.,
Olckers, T., Quimby, P.C. and Zimmermann, H.G. (2003)
A global review of riskbenefitcost analysis for the introduction of classical biological control agents against
weeds: a crisis in the making? Biocontrol News & Information 24, 91N-108N.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
Young, R. and Woodburn, T. (2002) Evaluation of research on
the biological control of nodding thistle (C. nutans), CSIRO
Division of Entomology, unpublished report, 14 p.
Fisher, A.J., Woods, D.M., Smith, L. and Bruckart, W.L.
(2007). Developing an optimal release strategy for the
rust fungus Puccinia jaceae var. solstitialis for biological control of Centaurea solstitialis (yellow starthistle).
Biological Control 42, 161171.
Grigulis, K., Sheppard, A.W., Ash, J.E. and Groves, R.H.
(2001) The comparative demography of the pasture weed
Echium plantagineum between its native and invaded
ranges Journal of Applied Ecology 38, 281290.
Gurr, G. and Wratten, S. (eds) (2000) Biological Control:
Measures of Success. Kluwer Academic Publishers, Dordecht, The Netherlands.
Hierro, J.L., Maron, J.L. and Callaway, R.M. (2005) A biogeographical approach to plant invasions: the importance
of studying exotics in their introduced and native range.
Journal of Ecology 93, 515.
Hinz, H.L. and Schwarzlaender, M. (2004) Comparing invasive plants from their native and exotic range: what can
we learn for biological control? Weed Technology 18,
15331541.
IAC (Industries Assistance Commission) (1985) Biological
control of Echium species (including Patersons curse/
Salvation Jane). Canberra, ACT; Australian Government
Publishing Service, IAC Report No. 371.
Jongejans, E., Sheppard, A.W. and Shea, K. (2006) Predispersal seed predation controls the native population dynamics of the invasive thistle Carduus nutans. Journal of
Applied Ecology 43, 877886.
Lonsdale, W.M. (1999) Concepts: global patterns of plant
invasions, and the concept of invasibility. Ecology 80,
15221536.
Marsden, J.S., Martin, G.E., Parham, D.J., Ridsdill Smith, T.J.
and Johnston, B.J. (1980) Returns on Australian agricultural research, The Joint IAC-CSIRO Benefit Cost Study of
the CSIRO Division of Entomology. CSIRO, Melbourne,
Australia, pp. 8493.
Michalakis, Y., Sheppard, A.W., Noel, V. and Olivieri, I.
(1993) Population structure of an insect herbivore and
its host plant on a microgeographic scale. Evolution 47,
16111616.
98
Fortieth anniversary review of the CSIRO European Laboratory
Appendix:
CSIRO European Laboratory
a potted history
In 1965 the Wheat Industry Research Council funded
CSIRO to work on management options for skeleton
weed, Chondrilla juncea. The CSIRO Entomology
(CEnto) and Plant Industry (CPI) combined efforts to
initiate European research activities. Doug Waterhouse
(Chief CEnto) and Milton Moore (CPI), through connections with Louis Emberger (Director of the CNRS
Ecology Lab in Montpellier), selected Montpellier
as a climatically similar base for European studies.
Tony Wapshere (CEnto) and newly appointed Richard
Groves (CPI) drove to Montpellier from the Australian
Embassy in Paris in late 1966 and surveyed skeleton
weed populations in SW France and SE Spain.
Tony stayed in France and in November 1966 set up
the CSIRO Biological Control Unit as its first Officer-
in-Charge (OIC), initially employing a small team including Jeanine Lamora (later Mrs Bronner) as admin
officer, using space in the CNRS Centre dEtudes Phytosociologiques et Ecologiques. This started the first
native invasive range comparative weed ecology research (with CPI in Australia) anywhere in the world to
find potential biological control agents.
By 1968 the unit of seven staff included the plant
pathologist, Siraj Hasan, based at St Christol, and Louis
Caresche as entomologist. After ecological studies had
shown that the rust Puccinia chondrillina damaged
infestations of C. juncea, Sirajs discovery of a virulent
strain IT32 at Vieste in Italy, aided by morphological
matching of leaf shape (by CPI), led to the worlds first
successful weed biological control programme using a
plant pathogen in 1971. From 1970 studies of C. juncea insects, mites and pathogens extended to the eastern Mediterranean centred in Thessaloniki in Greece.
In 1971 the unit moved into half of a rented duplex
building at 335 Ave Abb Paul Parguel, near the experimental land and glasshouses maintained at CNRS.
In 1973 the unit hosted the III International Symposium on Biological Control of Weeds (25 participants).
Tony wrote his seminal paper (Wapshere, 1974) on the
phylogenetic centrifugal specificity testing system and
initiated preliminary surveys on several other potential weed targets, particularly Echium plantagineum
in the western Mediterranean. In 1975 Jean-Paul Aeschlimann joined as entomologist on Hymenoptera to
initiate an insect biological control programme against
Sitona weevil. Over the next three years staff increased
from 10 to 15 and Tony surveyed C. juncea agents in
Iran. In 1978 a research outpost there involving Siraj
99
and entomologist John Huber also initiated studies on
Heliotropium europaeum. Tony also sent a population of the ragwort, Senecio jacobeaea L., flea beetle
to Australia and re-initiated 1950s biocontrol work on
St Johns wort, Hypericum perforatum. In 1979 Siraj
returned to the unit and Alan Kirk joined the laboratory and initiated the European arm of Australias celebrated dung beetle programme in collaboration with
Jean-Pierre Lumaret at the University of Montpellier.
Alan developed an egg collection protocol and over a
five-year period sent eggs of five species that successfully established in Australia.
From 1980, prospecting for plant pathogens started
in earnest. Phragmidium rust became a candidate for
blackberry, Rubus spp., and El Bruzzese joined the unit
on secondment from the Victorian Department of Lands
to work as the first State Department use of the facility.
Janine Vitou joined the unit in 1981 and Isabelle Olivieri (Univ. Montpellier II) completed the units first
French PhD thesis on thistlepathogen interactions.
After a government review of the facility in the
early 1980s, Jim Cullen took over as OIC in 1983 (on
a shorter rotation cycle) and initiated biological control programmes on Carduus thistles while continuing work on Hypericum, Senecio and Heliotropium.
John Scott worked in the unit from 198184 on Rumex
pulcher for the Western Australia Department of Agriculture before leaving to join CSIRO. Pierre Chaboudezs electrophoresis PhD studies on C. juncea and
P. chondrillina in 1985 led to work in eastern Turkey
and collaboration near Ankara on a trap garden for
several years. Jos Serin joined the unit as glasshouse
manager. In 1986, Andy Sheppard joined the unit as a
post doc to expand work started by Jim on the population dynamics of target weeds in the native range.
In 1987 David Briese became OIC and initiated
a programme on Onopordum thistles. Carey Smith,
from Australia, also spent two years studying the ecology of E. plantagineum while a court injunction prevented work in Australia. Yvette Mas became admin
officer and Mireille Jourdan joined the pathology team
in 1988. In 1989, Max Whitten (Chief Cento) visited
and decided to support funding for a permanent facility starting a search for a suitable site. James Coupland
joined the unit to initiate a programme against Mediterranean snails, Theba pisana, and Genevieve Martinelle
joined to undertake pre-emptive research on Russian
wheat aphid (RWA), Diuraphis noxiaboth serious
XII International Symposium on Biological Control of Weeds
pests to the Australian cereal industries. Thierry Thomann joined the unit, pushing staff numbers past 20.
During 199091, Jean-Paul and David worked on
the new laboratory, identifying a prime 2-ha site on the
proposed Agropolis Baillarguet Campus at Montferrier-
sur-Lez north of Montpellier, linked to Agropolis
discussions of a biocontrol campus there including
French, European (never built), Australian, and American laboratories. Jim Cullen approved the site and
CSIRO bought the land in 1991. Building started in
mid-1992 with significant support from Australian
Primary Industries and the local and regional French
administration. With steady project growth, the unit
reached its peak size of 25 staff with new projects on
Marrubium vulgare L. and Cytisus scoparius (L.) Link.
Many students from European universities undertook
internships at the laboratory, providing valuable input
into projects, including Yannis Michalakis who completed a PhD on Onopordum.
In late 1992 Richard Groves (CPI) returned as OIC
and with Jean-Paul oversaw the construction and completion of the new CSIRO European laboratory and
glasshouses. In late 1993 staff moved there, vacating
glasshouses and experimental land kindly provided
by CNRS since the 1960s. Collaboration with CABI
on Cytisus started with the secondment of Quentin
Paynter to the facility. On 10 October 1994, CSIROEL was formally opened by Barry Jones (Australian
Science Minister) in the presence of Georges Frche
(Mayor of Montpellier) and Jacques Blanc (Head of
Languedoc-Roussillon Region). The Institut National
de la Recherche Agronomique (INRA) Centre de Biologie et de Gestion des Populations (CBGP) and the
USDA-ARS-European Biological Control Laboratory
(EBCL) started construction on adjacent sites in the
late 1990s. Meanwhile, the INRA biocontrol and pest
management team rented space in CSIRO-EL until
CBGP was completed in 2002. The three laboratories
associated under the Complex Internationale de Lutte
Biologique Agropolis (CILBA), the largest group of
biocontrol researchers in Europe, with shared library
facilities (Le Centre Commun de Resources Documentaires) at CBGP.
In 1996 Mark Lonsdale joined as OIC and initiated
studies on the ecology of annual grasses and a landmark analysis of global plant invasions (Lonsdale,
1999), and the 1st CILBA/IOBC Montpellier biocontrol
conference on Technology Transfer was organized.
Reduced project funding led to a sharp decline in staff
numbers to below ten and Drs Aeschlimann and Hasan
left CSIRO after long careers. A McMaster Fellowship
at CSIRO-EL by Mark Rees, Imperial College London,
led to a landmark invasions modelling paper (Rees and
Paynter, 1997).
In 1998 John Scott returned as OIC. A new collaboration and joint molecular laboratory with USDA-ARSEBCL assisted joint projects on Raphanus raphanistrum L., Lepidium spp., Genista monspessulana (L.)
L. Johnson, and pink hibiscus mealybug, Maconellicoccus hirsutus (Green), and re-initiated work on Rubus
spp. Biosecurity assessment of the Asian gypsy moth,
Lymantria dispar, during visits by Nod Kay from New
Zealand and Mamoru Matuski (CEnto), was assisted
by the extensive eucalypt plantings in the grounds. In
2000 the 2nd CILBA/IOBC Montpellier biocontrol conference on Non-target and Indirect Effects took place
and Steve Novak, Boise State University, ID, spent a
sabbatical working on invasive annual grasses. The following year, the INRA team moved out and the Centre
de Coopration Internationale en Recherche Agronomique pour le Dveloppement (CIRAD) Entomological team and two EBCL staff moved into CSIRO-EL.
In 2002 Andy Sheppard returned as OIC and projects were initiated on Ulex europaeus L. and re-initiated
on screening cereal varieties against worldwide populations of RWA. The 3rd CILBA/IOBC Montpellier
biocontrol conference on Genetics and Evolution
also took place. In 2006 Mic Julien became OIC and
projects were initiated with a revisit by Steve Novak
on lippia, Phyla canescens (Kunth) Greene, an invasive
in Australia and France, and with EBCL, a project on
the molecular evaluation of biological control agents.
In 2007, the XII International Symposium on Biological Control of Weeds was held near Montpellier at La
Grande Motte, organized by CSIRO-EL and USDAEBCL (c. 250 participants) (these proceedings).
100
Abstracts: Theme 2 Benefit/RiskCost Analyses
F1 sterility: a novel approach for risk assessment
of biocontrol agents in open-field trials
J.E. Carpenter and C.D. Tate
USDA-Crop Protection and Management Research Laboratory,
2747 Davis Road Building 1, Tifton, GA 31793-0748, USA
Because of the growing concern of the potential risk of non-target effects, more stringent hostspecificity testing is required to import and release exotic biological control agents. Appropriate hostspecificity testing beyond quarantine conditions could reduce the risks of releasing biological control
agents that cause negative ecological effects, and also reduce the risk that a valuable and safe biological
control agent would not be approved for release. The use of reproductively inactivated insects could allow in-field host-specificity and geographical-range testing to assess the safety of exotic lepidopterans
being considered as biological control agents against invasive weeds. The outstanding control of invasive cacti by Cactoblastis cactorum (Berg) (Lepidoptera: Pyralidae), is a classic example of successful
biological control. However, C. cactorum became an invasive pest after its recent unintentional arrival
in Florida, and currently a major effort is being developed to mitigate its negative impact. Nevertheless, the presence of C. cactorum in the United States and its status as both a beneficial insect and pest
species provided us a unique model system to conduct proof-of-concept studies on the use of inherited
(F1) sterility as a new risk management tool for assessing the safety of exotic lepidopterans being
considered as biological control agents for invasive weeds.
Impact of biocontrol agents on native biodiversity:
the case of Mesoclanis polana
L.G. Carvalheiro, Y.M. Buckley and J. Memmott
School Biological Sciences, University of Bristol, Bristol BS8 1UG, UK
The safety of biocontrol is a contentious issue, with particular concerns about the interactions between
biocontrol agents and non-target species. Such interactions can occur either directly, if an agent attacks a non-target host, or indirectly, when the agent affects non-target species via shared natural enemies. While there are some data on direct effects, there are very little data on indirect effects. In this
talk we ask how a native food web is affected by a recently introduced biocontrol agent, Mesoclanis
polana (Diptera: Tephritidae). While this agent will not directly affect native species (it feeds only on
the target weed), it can potentially affect food-web structure indirectly via native parasitoids shared
with native herbivores. Nearly 9000 seed predators were reared at the 18 field sites; 17 parasitoid species were reared from herbivores in Chrysantomoides monilifera seeds, with Mesoclanis polana the
most probable host. Using a food-web approach, we ask: How does M. polana influence the native
plantherbivoreparasitoid community?
CAB International 2008
101
XII International Symposium on Biological Control of Weeds
A look at host range, host specificity and
non-target safety from the perspective of a
plant virus as a weed-biocontrol agent
R. Charudattan, M. Elliott, E. Hiebert and J. Horrell
Plant Pathology Department, University of Florida/IFAS, Gainesville, FL 32611-0680, USA
Safety of weed biocontrol agents to non-target plants is determined primarily by means of host range/
host specificity testing, often relying on the Centrifugal-Phylogenetic scheme to guide the selection
of test plants. Recently, we have had a unique opportunity to examine the host range/host specificity of
a Tobamovirus, namely Tobacco mild green mosaic tobamovirus (TMGMV), in the context of its use
as a bioherbicide for Solanum viarum. A host range study of more than 400 plant species, subspecies or
varieties in 58 plant families revealed that the virus is a pathogen adapted to plants in the Solanaceae
and to a few outliers in unrelated families. The theory that host specialization follows a centrifugalphylogenetic design appears to hold true for TMGMV at the family level. Within the family, host
specificity is distinctly determined at the genus and species levels by a single gene or a few genes, with
the host reaction ranging from immunity to resistance and susceptibility even within a species. The
lethal hypersensitive resistance reaction seen in S. viarum is also different in that it is rarely observed
in plantvirus interactions. Our results provide a framework to analyse the non-target risk in using
TMGMV as a weed control agent.
Novel approaches for risk assessment: feasibility studies
on temporary reversible releases of biocontrol agents
J.P. Cuda,1 O.E. Moeri,1 W.A. Overholt,2 V. Manrique,2 S. Bloem,3
J.E. Carpenter,4 J.C. Medal1 and J.H. Pedrosa-Macedo5
University of Florida, Department of Entomology and Nematology, Building 970 Natural Area Drive,
Gainesville, FL 32611-0620, USA
2
University of Florida, Biological Control Research and Containment Laboratory,
2199 South Rock Road, Fort Pierce, FL 34945-3138, USA
3
USDA-APHIS-PPQ-CPHST, 1730 Varsity Drive Suite 300 Third Floor, Raleigh, NC 27606-5202, USA
4
USDA-Crop Protection and Management Research Laboratory,
2747 Davis Road Building 1, Tifton, GA 31793-0748, USA
5
Lab. Neotropical de Cont. Biol. de Plantas, SCA-Universidade Federal do Paran,
Rua Bom Jesus 650, Curitiba PR 80.035-010, Brazil
1
In accordance with a 1999 Executive Order adopted by the US government, federal agencies are mandated not to promote any environmental actions, e.g. biological control, unless the agencies determine
that the benefits outweigh the risks and that measures will be taken to minimize potential harm. Recent
case studies have shown that the risks associated with classical biological control are high because
(a) host and habitat specificity are difficult to predict, and (b) natural enemy releases are normally
permanent and irreversible. If the biological control agent is found not to be entirely host specific
post-release, it can spread and attack non-target species in perpetuity. The potential for negative environmental impacts from biological control releases can be minimized or eliminated by adopting a precautionary approach that, in effect, makes experimental releases of candidate biological control agents
temporary and reversible. The advantage of this approach is that biological control can proceed on an
experimental basis before full-scale implementation. We illustrate the feasibility of this approach with
two different natural enemies of Brazilian peppertree, Schinus terebinthifolius Raddi (Anacardiaceae),
by proposing single sex releases of the defoliating sawfly Heteroperreyia hubrichi Malaise (Hymenoptera: Pergidae) and sterilizing the leafroller Episimus utilis Zimmerman (Lepidoptera: Tortricidae)
with gamma radiation.
102
Abstracts: Theme 2 Benefit/RiskCost Analyses
A wolf in sheeps clothing: potential dangers of
using indigenous herbivores as biocontrol agents
J. Ding1,2 and B. Blossey2
Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan 430074, China
Department of Natural Resources, Cornell University, Ithaca, NY 14850, USA
1,2
Concerns about non-target effects of introduced natural enemies on native species and the existence
of indigenous natural enemies attacking invasive species stimulate an interest in using indigenous
herbivores for control of invasive plants. According to proponents of this strategy, using indigenous
species as biocontrol agents should receive priority over introductions of foreign natural enemies.
Such an approach is considered safe with low risk to native species. In contrast, we are concerned that
biological control using augmentation of indigenous herbivores may lead to more serious non-target
effect on native species. Indigenous natural enemies are never host specific (they have incorporated a
novel host into their diet!) and often they prefer their ancestral hosts over the novel invasive ones. Even
if a population or host race derived from an indigenous herbivorous insect prefers its novel invasive
host, it must also be of sufficient impact to control the target invasive plant. We examined a North
American indigenous herbivore, the leaf beetle Galerucella nymphaeae, for its potential as biological
control agent of water chestnut (Trapa natans), in particular for its potential non-target effect on native
plant species. Although speciation or host race formation often occurs in the genus Galerucella, the
North American G. nymphaeae preferred its ancestral host yellow water lily (Nuphar lutea) over water
chestnut. A competition study under different herbivory levels further indicated that mass-rearing and
releasing this indigenous leaf beetle will result in increased damage to non-target native species, promoting more vigorous growth of the invasive target weed. Introduction of foreign natural enemies for
biological control is not risk-free, but mass releases of indigenous herbivores may pose a more serious
threat to native species than generally acknowledged.
Impact of biological control of Salvinia molesta in
temperate climates on biodiversity conservation
B.R. Hennecke1,2 and K. French2
Centre of Plant and Food Science, University of Western Sydney, Locked Bag 1797,
South Penrith DC, NSW 1797, Australia
2
School of Biological Sciences, University of Wollongong, Wollongong, NSW 2522, Australia
1
Salvinia molesta is a Weed of National Significance in Australia, invading freshwater rivers and lakes
and resulting in loss of biodiversity. Biological control of salvinia has been successful in tropical
areas but has not yet shown an impact in temperate regions. Thus, salvinia is still considered one of
the worlds worst aquatic weeds and is currently spreading in many parts of the world. In Australia
increased efforts have been undertaken in recent years to distribute and establish the biological control
agent (Cytrobagous salviniae) in temperate regions. We investigated the potential long-term impact of
biological control on biodiversity conservation in the Hawkesbury-Nepean River in Sydney, Australia.
Differences in plant and microbe communities in salvinia infested and non-infested areas were recorded and analysed and related to water nutrient levels. The project highlights conservation priorities
for revegetation and restoration to maximize species diversity following the introduction of biological
control of salvinia.
103
XII International Symposium on Biological Control of Weeds
Opening Pandoras box? Surveys for attack
on non-target plants in New Zealand
Q. Paynter,1 S.V. Fowler,2 A.H. Gourlay,2 M.L. Haines,3 S.R. Hona,1 P.G. Peterson,4
L.A. Smith,2 J.R.A. Wilson-Davey,2 C.J. Winks1 and T.M. Withers5
Landcare Research Ltd, Private Bag 92170, Auckland, New Zealand
2
Landcare Research, P.O. Box 40, Lincoln, New Zealand
3
Lincoln University, P.O. Box 84, Canterbury, New Zealand
4
Landcare Research, Private Bag 11052, Palmerston North, New Zealand
5
Ensis, Private Bag 3020, Rotorua, New Zealand
1
The environmental safety record of weed biocontrol has recently been questioned when examples of
damage to non-target plants were reported overseas. We review previous records of non-target attack
and present the results of recently conducted systematic surveys to look for additional evidence of nontarget damage caused by weed biological control agents that became established in New Zealand between 1929 and 2001. Our findings are discussed to determine the reliability of host-specificity testing
and overall safety record of weed biological control in New Zealand and to ascertain whether lessons
can be learnt that will enhance the safety of future weed biocontrol programmes.
New biological control agents for Cytisus scoparius
(Scotch broom) in New Zealand: dealing with the
birds and the bees and predicted non-target attack
to a fodder crop
Q. Paynter,1 A.H. Gourlay,2 P.G. Peterson,3 J.R.A. Wilson-Davey,2 J.V. Myers,2
S.R. Hona1 and S.V. Fowler2
Landcare Research Ltd, Private Bag 92170, Auckland, New Zealand
2
Landcare Research, P.O. Box 40, Lincoln, New Zealand
3
Landcare Research, Private Bag 11052, Palmerston North, New Zealand
1
The invasive European shrub Scotch broom has detrimental impacts on farming, forestry and conservation in New Zealand. The current suite of biological control agents does not damage plants sufficiently over the entire growing season to have a major impact on broom populations and the release
of additional agents: a chrysomelid leaf-feeding beetle Gonioctena olivacea Frster and an oecophorid
stem-tying moth Agonopterix assimilella Treitschke was proposed. Both are narrowly oligophagous,
with the potential to develop on a few closely related plant species within the tribe Genisteae. New
Zealand has no native Genisteae but an exotic species, tagasaste (tree lucerne, Cytisus proliferus L.f.
var palmensis H. Christ) is closely related to Scotch broom and may be affected by non-target damage.
This is undesirable because tagasaste is planted to stabilize soil on marginal hill country and is a minor
fodder crop in New Zealand. Furthermore, broom is valued as a pollen source by the beekeeping industry and its pods form a seasonally important food source for kerer (an endemic pigeon Hemiphaga
novaeseelandiae). We describe a benefit/cost analysis and ecological studies performed that paved the
way for the release of G. olivacea and A. assimilella in New Zealand despite objections to their use.
104
Abstracts: Theme 2 Benefit/RiskCost Analyses
Predicting risk and benefit a priori in weed biological
control: a systems modelling approach
S. Raghu,1 K. Dhileepan2 and J. Scanlan3
Illinois Natural History Survey & University of Illinois, Champaign, IL, USA
Cooperative Research Centre for Australian Weed Management, Alan Fletcher Research Station,
Department of Natural Resources & Water, Sherwood, QLD, Australia
3
Robert Wicks Pest Animal Research Centre, Department of Natural Resources & Water,
Toowoomba, QLD, Australia
1
We developed a simulation model to predict risks and benefits a priori for Charidotis auroguttata, a
potential biocontrol agent for Macfadyena unguis-cati in Australia. Preliminary host-specificity testing
of this herbivore indicated that there was limited feeding on a non-target plant, although the non-target
was only able to sustain some transitions of the life cycle of the herbivore. The model incorporated herbivore, target and non-target life history, and spillover dynamics of populations of this herbivore from
the target to the non-target under a variety of scenarios. Data from greenhouse and quarantine studies
were used to parameterize the model and predict the relative risks and benefits of this herbivore when
the target and non-target plants co-occur. Key model outputs include population dynamics on target
(apparent benefit) and non-target (apparent risk) and fitness consequences to the target (actual benefit)
and non-target plant (actual risk) of herbivore damage. The model predicted that non-target risk became unacceptable (i.e. significant negative effects on fitness) when the ratio of target to non-target in
a given patch ranged from 1:1 to 3:2. By comparing the current known distribution of the non-target
and the predicted distribution of the target we were able to identify regions in Australia where the agent
may pose an unacceptable risk.
Comparative risk assessment of Linaria dalmatica
and L. vulgaris biological control
S.E. Sing and R.K. Peterson
Department of Land Resources and Environmental Sciences, Montana State University,
P.O. Box 173120, Bozeman, M 59717-3120, USA
Dalmatian (Linaria dalmatica) and yellow (Linaria vulgaris) toadflax (Scrophulariaceae) are shortlived perennial herbs of Mediterranean origin. Both species were imported to North America for
horticultural purposes but have since become naturalized through multiple introductions. Toadflax
diminishes floral diversity because it displaces desirable and/or native species in rangeland and forest
habitats. Toadflax infestations reduce effective available grazing land because the unpalatable foliage
is avoided by cattle. Herbicide treatment of toadflax is expensive due to the large acreages affected
and must be repeated at impractical frequencies. Classical biological control of toadflax was initiated in the late 1960s. To date, six exotic agent species targeting the flowers, stems, foliage or roots
of toadflax have established in North America. Documentation of non-target impacts in a number
of classical biological control releases against other weed species suggests that a retrospective risk
assessment of toadflax biological control examining ecosystem impacts of previous introductions
would be well-advised at this time. Preliminary results and longer term plans to accomplish this goal
will be discussed.
105
This page intentionally left blank
Theme 3:
Target and Agent Selection
Session Chair: Ren Sforza
107
This page intentionally left blank
Keynote Presenter
Latin American weed biological
control science at the crossroads
R.W. Barreto1
Summary
Latin America is the centre of origin of many of the invasive alien weeds threatening natural and
agricultural ecosystems throughout the world. As a result, it has been an important destination for
expeditions in search of natural enemies for their control. Unfortunately, the role of local scientists
has been mainly that of contracted explorers, cooperating on projects aimed at exploration for classical biological control agents. This is changing as the need to confront the growing threat from alien
weeds in Latin America gathers pace. Nevertheless, with limited funding and a continuing ignorance
by both the general public and the decision makers about the scale of the invasive weed problem
in Latin America, target selection will be critical since this will determine the long-term viability
of biological control in the region. In the proactive, new role to develop biological control in Latin
America, should easy targets be selected, for which there has been success on other continents,
or instead, should targets be more challenging, potentially confrontational, such as African grasses
which threaten not only the stability of unique ecosystems but which could also have global consequences? These issues will be discussed based on experiences gained from past and present collaborative projects.
Keywords: target selection; agent selection; classical biological control; bioherbicides.
Latin American weed biological
control: historical background
Latin America, including the Caribbean in this paper, is
the centre of origin of many of the invasive alien weed
threatening systems throughout the world. For instance,
59 of the 209 worst weeds on a worldwide scale are native to Latin America (Cronk and Fuller, 1995). They
include aquatic weeds such as water hyacinth, Eichhornia crassipes (Mart.) Solms; alligator weed, Alternanthera philoxeroides (Mart.) Griseb.; capybara grass,
Hymenachne aplexicaulis (Rudge) Nees; water lettuce,
Pistia stratiotes L.; arrowhead, Sagittaria motevidensis Cham. and Schlecht.; and salvinia, Salvinia molesta
D.S. Mitchell; and terrestrial weeds such as mistflower,
Ageratina riparia (Regel) King and Robinson; Siam
Departamento de Fitopatologia, Universidade Federal de Viosa, Viosa/
MG, 36570-000, Brazil <[email protected]>.
CAB International 2008
1
weed; Chromolaena odorata (L.) King and Robinson;
lantana; Lantana camara L.; mile-a-minute Mikania micrantha H.B.K.; sensitive plant, Mimosa spp.;
prickly pear, Opuntia spp.; strawberry guava, Psidium
cattleianum Sabine; and Brazilian pepper tree, Schinus
terebinthifolius Raddi.
Latin America has played a major role in weed biological control since its inception at the beginning of
the 20th century. Two early pioneering projects were
involved in transcontinental transfers of natural enemies aimed at L. camara and Opuntia vulgaris Miller.
Lantana camara
The first explorations for natural enemies of a weed
for biological control were conducted in Mexico by the
Hawaii Department of Agriculture against L. camara.
Insects were introduced into Hawaii in 1902 (Perkins
and Swezey, 1924). Eight of 33 insect species that were
released in Hawaii from 1902 to 1970 were established
(Waterhouse and Norris, 1987). Although the accounts
109
XII International Symposium on Biological Control of Weeds
of the impact of these insects are somewhat vague, they
are generally regarded as having contributed to partially
controlling the weed (Goeden, 1978). L. camara is
of worldwide importance, and interest in its biological control has been maintained to this date. In 1992,
the fungus Septoria was introduced to combat lantana
(Davis et al., 1992) with excellent results (Trujillo,
2005). The case of L. camara is remarkable as it was
the first target for biological control, and there have
been around 30 projects worldwide (Broughton, 2000).
The most recent introduction was a rust fungus Prospodium tuberculatum (Speg.) Arthur into Australia in
2001 (Ellison et al., 2006). Unfortunately, no agent or
combination of agents has proved sufficient to control
this important weed species, and it is likely that new
agents will be required. Fortunately, a highly diverse
list of parasites and arthropods attack it, and new potential agents are still being found (Barreto et al., 1995;
Pereira and Barreto, 2000).
Opuntia stricta
The control of prickly pear, Opuntia stricta (Haw.)
Haw., in Australia, was also based on collections made
in Latin America. In 1925, the moth Cactoblastis cactorum (Bergroth) was introduced from Argentina. In
1933, complete control was achieved over 24 million
hectares of valuable land (McFadyen and Willson,
1997). This was the first example of a silver-bullet
effect in weed biological control, but the contribution
of other arthropods and even pathogens may also have
been relevant. Twelve other species of Opuntia spp.
have been targeted by classical biological control projects using Latin American arthropods, mostly from Argentina and Mexico (Julien and Griffiths, 1998).
The first weed biological control project
targeting a weed in Latin America
The first deliberate introduction against a weed in
Latin America took place in Chile in 1952 using the
beetles Chrysolina hyperici (Forster) and Chrysolina
quadrigemina (Suffrian) (Chrysomelidae) against St.
Johns wort, Hypericum perforatum L. This successful project (Norambuena and Ormeo, 1991) piggybacked on the successful project carried out in 1947 in
the USA. Unfortunately, these introductions remained
the sole examples of classical introductions into Latin
America for the next 20 years.
Pioneering work of Latin American plant
pathologists in classical and inundative
weed biological control
Edgardo Oehrens Bertossi, Professor of plant pathology of the Universidad Austral de Chile and often
regarded as father of plant pathology in Chile, under-
took two pioneering introductions of fungal pathogens
against weeds in Latin America. The rust fungus Phragmidium violaceum (Schultz) Winter was introduced
from Europe into Chile against blackberry, Rubus spp.,
in 1973 (Oehrens, 1977; Oehrens and Gonzales, 1974),
and Uromyces galegae (Opiz) Sacc. was introduced,
also from Europe, against goats rue, Galega officinalis
L. (Oehrens and Gonzales, 1974). Phragmidium violaceum provided effective control of Rubus constrictus
Lefvre and P.J. Mll., but no control resulted for Rubus
ulmifolius Schott. (Oehrens and Gonzales, 1977; Medal,
2003). Uromyces galegae established but did not have
any impact on goats rue (Medal, 2003). It is interesting that these introductions were taking place almost at
the same time as the rust fungus Puccinina chondrillina
Bubak and Sydenham was being used for the first time
in Australia for the biological control of skeleton weed,
Chondrilla juncea L. (Cullen, 1974). Bertossi was
ahead of his time for Latin America biological control
science. He conjectured the use of rust fungi in weed
biological control as early as 1963 (Oehrens, 1963), before Wilsons (1969) seminal publication, making it one
of the earliest records of this kind of consideration from
a plant pathologist. It is very unfortunate that Chilean
pathologists have never followed Bertossis example.
The only other account of the deliberate introduction
of a pathogen as a classical biological control agent in
Latin America is that of a failed attempt in Argentina to
use of P. chondrillina as a classical biological control
agent for C. juncea (Julien and Griffiths, 1998).
The inundative strategy involving the use of endemic
fungal pathogens against invasive weeds in Latin America was explored by several research groups after being
pioneered by Jos Tadashi Yorinori, a leading Brazilian
soybean plant pathologist of Empresa Brasileira de Pesquisa Agropecuria (EMBRAPA-Soja). With collaborators, he evaluated the fungus Bipolaris euphorbiae
(Hansford) Muchovej as a mycoherbicide against wild
poinsettia, Euphorbia heterophylla L. (Yorinori, 1985,
1987; Yorinori and Gazziero, 1989). This work was interrupted in the 1990s due to changed research priorities
in EMBRAPA-Soja and to the discovery of common
biotypes of the weed that appeared to be resistant to
B. euphorbiae. Research on this fungus as a potential
mycoherbicide continues in Brazil (Marchiori et al.,
2001; Nechet et al., 2006; Barreto and Evans, 1998).
Continuation of searches for biological
control agents by foreign scientists
Most work in Latin America continued to be limited
to surveying for arthropods as potential agents for use
in other continents. In the late 1950s, US Department
of Agriculture (USDA)-Agricultural Research Service
(ARS) scientists surveyed South America for natural
enemies of A. philoxeroides and E. crassipes. Instead of
short expeditions that had previously been used, USDA-
110
Latin American weed biological control science at the crossroads
ARS chose to establish a base from which longer and
more frequent surveys could be made and supported.
The USDA-ARS South American Biological Control
Lab (SABCL) was inaugurated in 1962 and continues its activities with excellent results until this date
(Table 1). Similarly, Australian scientists from Commonwealth Scientific and Industrial Research Organisation (CSIRO) set up base in Curitiba, Brazil, later
(1984) moved to Acapulco Mexico and then in 1987
to the current station in Vera Cruz, Mexico. These stations often hosted researchers from other institutions
(Segura and Heard, 2004). Such strategy adopted by
US and Australian scientists yielded agents that resulted in some of the most spectacular cases of success in
weed biological control such as those that followed the
introduction into Australia of the weevil Cyrtobagous
salviniae Calder and Sands against S. molesta (Room
et al., 1981); the introduction of Agasicles hygrophila
Seleman and Vogt against A. philoxeroides into the
US (Spencer and Coulson, 1976), the weevil Neohydronomus affinis Hustach introduced against P. stratiotes in Australia (Harley et al., 1984) and the moth
Niphograpta albiguttalis (Warren) and the weevil Neochetina eichhornia Warner in the US and Neochetina
bruchi Hustache in Australia against E. crassipes. Such
successes were later replicated many times in different
parts of the world with the same agents (e.g. Center,
1982; Julien and Griffiths, 1998).
In the last part of the 20th century, Hawaii-based
entomologists such as C.J. Davis and R. Burkhart and
the plant pathologist E. Trujillo introduced insects from
Latin America (mainly the Caribbean) against Kosters
curse, Clidemia hirta (L.) D. Don, as well as one fungus [Colletotrichum gloeosporioides (Penzig) Penzig
and Sacc. (Julien and Griffiths, 1998)]. Although the
fungus and a thrips Liothrips urichi Karny were established and Trujillo (2005) claims control levels to be
adequate after repeated spraying with suspension of the
fungus conidia, the weed is still a cause for concern in
forest habitats (Cronk and Fuller, 1995).
Other weeds from Latin America that were targeted
in Hawaiian projects were: mistflower, Ageratina riparia (Regel) R. King and H. Robinson, from Mexico
which was spectacularly controlled with a white smut
fungus Entyloma ageratinae Barreto and Evans (Trujillo, 2005); bananapoka, Passiflora tarminiana Coopens, Barney, Jrgensen and MacDugal (=Passiflora
mollissima, Passiflora tripartita), against which insects
and a fungus were released. The fungus Septoria passiflorae Sydenham caused significant decline of banana
poka biomass in forest areas (Trujillo, 2005).
Scientists from South Africa (Plant Protection Research Institute) have also surveyed Latin America for
natural enemies of native plants that became serious
weeds in South Africa. Of 31 weed species listed, 15
are from Latin America or have Latin America as part
of their native range (Olckers and Hill, 1999). Some
projects piggy-backed on previous studies, such as
those against, L. camara, E. crassipes and P. stratiotes;
others were initiated by South Africans. Among the recent success stories are: red water fern, Azolla filiculoides Lamarck, using the weevil Stenopelmus rufinasus
Gyllenhal collected from the US, Argentina and Paraguay (Hill, 1999; Hill and Cilliers, 1999).
Intensive searches have also been made in Latin
America by scientists from CAB International for biological control of pantropical weeds such as C. odorata,
L. camara, M. micrantha, Mimosa pigra L. Parthenium
hysterophorus L., and others. A recent example of work
by CABI is the introduction of Puccinia spegazzini de
Toni from Latin America to India against M. micrantha
(Sankaran et al.,2008).
Latin American weed biological
control: the present
Targeting weeds in Latin America
restarted
Biological control activity restarted in Latin America in Chile (INIA-Centro Regional de Investigacin
Carillanca), with a programme in the 1980s against
gorse, Ulex europaeus L., using the seed feeder Exapion ulicis Forster, an agent already introduced with
some success from Europe into New Zealand (Norambuena et al., 1986; Norambuena and Piper, 2000). The
gorse spider mite, Tetranychus lintearius Dufour, was
also introduced later from Hawaii and Portugal (Norambuena et al., 2007).
Interactions between foreign weed
biological control scientists and
Latin American scientists
Very positive actions for weed biological control
science in Latin America have been the efforts by Australia-, New Zealand-, South Africa-, European- and
US-based scientists to encourage active involvement
of Latin American entomologists and plant pathologists in weed biological control programmes (see Table
1). Some of these involve interactions by scientists and
research groups from more than two Latin American
countries such as the projects on Brazilian pepper tree,
S. terebinthifolius, and tropical soda apple, Solanum
viarum Dunal (Gandolfo et al., 2007; Medal et al.,
2002).
Many scientists from Latin America were trained
in weed biological control in Europe, and US. Further,
after J. Medal and D Gandolfo took part of an intensive biological control of weeds training course in Australia, they also organized a series of three courses in
2002, 2004 and 2006 in Nicaragua with attendees from
numerous Latin American countries.
111
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Research groups involved with weed biological control in Latin America, their projects and status of activities.
112
Organization/research leader
Status of
activities
Centre for Agriculture and
Biosciences International,
Caribbean Latin American
Station
Centro Agronmico Tropical
de Investigacin y Enseanza
Interrupted
Location:
city/country
Principal or recent
target weeds
Approacha
Status of project
selected results
Selected publications
Curepe/Trinidad
Tobago
Chromolaena odorata
Opuntia spp.
ce
ci
Concluded
Concluded
Elango et al. (1993)
Interrupted
since 1999
Turrialba/Costa
Rica
Rottboellia cochinchinensis
(Lour.) W.D. Clayton
ci
Reeder and Ellison (1999),
Snchez-Garita (1999)
Centro de Recursos Naturales Renovables de la Zona
Semirida (CERZOS)
and Departamento de
Agronomia/Universidad
Nacional del Sur
F. Anderson, R. Delhey,
M. Kiehr, G. Traversa
CSIRO Entomology Mexico
Field StationT. Heard,
R. Segura
Ongoing
Bahia Blanca/Argentina
Cabomba caroliniana
A. Gray
Nassella spp.
Phyla canescens (Kunth)
Greene
ce
Interrupted due
political and
administrative
problems
Ongoing
ce
ce
Ongoing
Ongoing
Ongoing
Veracruz/Mexico
Mimosa pigra L.
Sida acuta Burman f.
ce
ce
Concluded
Concluded
Ostermeyer and Grace
(2007), Lonsdale et al.
(1995), Forno et al. (1992).
Departamento de Biologia
Aplicada Agropecuria
(DBAA)-Universidade
Estadual Paulista Jlio
de Mesquita (UNESPJaboticabal)R.A. Pitelli
Ongoing
Jaboticabal/Brazil
Eichhornia crassipes
Egeria densa Planch.
Senna obtusifolia (L.) Irwin
and Barney
in
in
in
Ongoing
Ongoing
Ongoing
Pitelli et al.(2007), vila
and Pitelli (2005), Borges
Neto and Pitelli (2004)
Sosa et al. (2008)
Observations
So far, one lost opportunity for a project with
great potential for Latin
America
Work with Nassella spp.
challenged by difficulties with rust life-cycle
(Anonym, 2006)
Until now 16 agents
evaluated in this
lab and released against
four weeds in Australia
Projects mainly concentrated on endemic
aquatic weeds
XII International Symposium on Biological Control of Weeds
Table 1.
Departamento de Fitopatologia (DFP)-Universidade
Federal de Viosa
R.W. Barreto
Ongoing
Viosa/Brazil
Fundao Universidade
Regional de Blumenau
M.D. Vitorino
in
in
in
ce
ci
Interrupted
Ongoing
Ongoing
Ongoing
Ongoing
ce
ce
ce
Ongoing
Interrupted
Ongoing
ce
Pereskia aculeata Miller
Pistia stratiotes
Psidium cattleianum
Saggitaria montevidensis
Cham. and Schlecht.
Schinus terebinthifolius
ce
ce
ce
in
One agent introduced and established in Hawaii
and Tahiti
Ongoing
Finished
Ongoing
Ongoing
Tradescantia fluminensis
Vell.
Cyperus rotundus
Senna obtusifolia
ce
in
in
ce
Suspended
Braslia/Brazil
Interrupted
Londrina/Brazil
Euphorbia heterophylla
in
Ongoing
Blumenau/Brazil
Psidium cattleianum
Schinus terebinthifolius
Tecoma stans (L.) Juss.
ex Kunth.
ce
ce
in
See site indicated below
for a complete list of publications, most representing
surveys of the mycobiota
of weeds in Brazil.b
Projects encompassing
a range of native and
introduced weeds in
Brazil and concentrated
on the study of fungal
pathogens as biological
control agents and also
some studies on plant
pathogenic nematodes.
Priorities within
EMBRAPA
changed to other
areas
vila et al. (2005), vila
et al. (2000), Borges Neto
et al. (2000)
Scientists in the team
still interested in returning to the field
Priorities within
EMBRAPA
changed to other
areas
Ongoing
Ongoing
Ongoing
Yorinori (1985, 1987),
Yorinori and Gazziero
(1989)
Also problems with
weed resistance to the
fungus
Cuda et al. (2005), Hight et
al. (2003), Vitorino
et al. (2000). Wikler and
Vitorino (2007)
Additional projects in
cooperation with USDA
and Plant Protection Research Institute (South
Africa)
One agent being
tested in quarantine
in Florida
Ongoing
Latin American weed biological control science at the crossroads
113
EMBRAPA-Centro Nacional
de Pesquisa de Recursos
Genticos e Biotecnologia
(CENARGEN)Sueli
Mello
EMBRAPA-Soja
J.T. Yorinori
Commelina benghalensis L.
Cyperus rotundus L.
Euphorbia heterophylla
Eichhornia crassipes
Hedychium coronarium J.
Koenig
Ipomoea carnea Jacq.
Lantana camara
Macfadyena unguis-cati
(L.) Gentry
Miconia calvescens DC
Table 1.
(Continued) Research groups involved with weed biological control in Latin America, their projects and status of activities.
Status of
activities
Instituto de Investigaciones,
Agropecuaria (INIA)-Centro
Regional de Investigacin
(CRI) Carillanca
H. Norambuena
Instituto Mexicano de
Tecnologia del Agua
M.M. Jimnez
Universidade Estadual do
Centro-Oeste do Paran
(UNICENTRO)C. Wikler
Ongoing
114
Principal or recent
target weeds
Approacha
Status of project
selected results
Selected publications
Observations
Temuco/Chile
Ulex europaeus
ci
Ongoing
Norambuena et al. (2007),
Norambuena and Piper
(2000), Norambuena et al.
(1986)
Almost a single-man
operation
continuation at risk.
Ongoing
Jiutepec/Mexico
Eichhornia crassipes and
other aquatic weeds
in
Ongoing
Irat/Brazil
ce
ce
ce
ce
Jimenez and Lopez (2001),
Jimenez and Charudattan
(1998)
Hight et al. (2003), Wikler
et al. (1996)
Universidad de Costa
RicaP. Hanson
Final stages
San Jos/Costa
Rica
Cabomba caroliniana
Psidium cattleianum
Schinus terebinthifolius
Tibouchina herbacea (DC)
Cogn.
Miconia calvescens
Fungi and insects
actually being used
in the field
Agents under
evaluation in quarantine in Hawaii
and Florida
Burckhardt et al. (2005)
Universidad Austral de
ChileE.B. Oehrens
Ended
Valdivia/Chile
Rubus spp.
Galega officinalis
ci
ci
Agents under
evaluation in quarantine in Hawaii
Partial success
No control
Universidade Federal do
ParanH. Pedrosa-Macedo
Ongoing
Curitiba/Brazil
Psidium cattleianum
Schinus terebinthifolius
Tradescantia fluminensis
ce
ce
ce
USDA-ARS SABCL South
American Biological Control
LaboratoryJ. Briano,
W. Cabrera Walsh, F.
McKay, C. Hernandez,
A. Sosa
Ongoing
Hurlingham/
Argentina
Eichhornia crassipes, Alternanthera philoxeroides,
Solanum viarum, Prosopis
spp., Cabomba caroliniana
A. Gray, Schinus terebenthifolius, Cardiospermum
grandiflorum Sw., Campuloclinium macrocephalum (Less.) DC, Pereskia
aculeata
ceall
projects
a
Location:
city/country
ce
Oehrens (1977), Oehrens
and Gonzales (1974, 1975,
1977)
Agents under
Pedrosa-Macedo et al.
evaluation in Brazil (2007a, 2007b), Pedrosaand in quarantine
Macedo (2000), Medal
in Hawaii and
et al. (1999)
Florida
Several successful Numerous publications by
introductions in the a series of leading scienUS and elsewhere. tists such as H. Cordo,
Several ongoing
D. Gandolfo, Willie
projects.
Cabrera Walsh,
Cristina Fernandez,
Fernando McKay,
Alejandro Sosa.
Continuation of activities unlikely after end of
project.
No disciples left behind
after a brilliant start.
Activity in this lab led
to the formation of new
groups (UNICENTRO
and FURB)
This has been by far the
largest and most productive team of biological
control scientists in
Latin America but dealing almost exclusively
with arthropods
ce, Classical biological exploration of local agents for introduction outside Latin America; ci, classical biological introduction of agents against alien weeds in Latin America; in, inundative/bioherbicide.
For a complete list of publications, see: http://lattes.cnpq.br/4191011304306773.
b
XII International Symposium on Biological Control of Weeds
Organization/research leader
Latin American weed biological control science at the crossroads
Present status of research groups and
research activities in weed biological
control in Latin America
An assessment of the status of research activities
in Latin America was undertaken through a search of
the literature and personal contacts (Table 1). Sixteen
researchers or groups have involvement with weed biological control in Latin America. Only six of 23 countries have scientists working in weed biological control:
Argentina, Brazil, Chile, Costa Rica, Mexico and Trinidad Tobago. Six have been involved solely dealing with
exploration for natural enemies to be used elsewhere.
Three labs deal mostly with the inundative/bioherbicide
approach utilizing endemic pathogens. Three labs have
been involved solely with classical introductions of
agents into Latin America. Unfortunately, only one of
these remains active (INIA-CRI Carillanca). Additionally, four labs had activities in more than one approach.
Six dealt solely with pathogens, five with arthropods
and four with both. This is surprising, considering the
much longer history of the use of insects in weed biological control and the great number of entomologists
involved in weed biological control.
Work at DFP/UFV (Brazil)
The Departamento de Fitopatologia, DFP/Universidade Federal de Viosa (UFV) is one of the largest
Plant Pathology departments of any university in Latin
America. Weed biological control activity began there
after 1994, funded by Brazilian agencies, such as Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq) and Fundao de Amparo Pesquisa do
Estado de Minas Gerais (FAPEMIG), and foreign organizations, such as the University of Hawaii and Landcare Research, New Zealand. Twelve MSc and PhD
students have studied weed biological control classical
(inoculative) and mycoherbicide (inundative) strategies. Four serious agricultural weeds in Brazil have
been selected for mycoherbicide development; wandering jew, Commelina benghalensis L., purple nutsedge,
Cyperus rotundus L., wild poinsettia, E. heterophylla
and arrowhead, Saggitaria montevidensis Cham. and
Schlecht. Work is advanced on the use of the fungus
Lewia chlamidosporiformans B.S. Vieira and Barreto
(Vieira and Barreto, 2005). Demonstrations of its commercial viability are presently under way.
Surveys to discover fungal pathogens attacking selected weeds in Brazil have been conducted. Recently,
surveyed were: Hedychium coronarium J. Koenig,
Ipomoea carnea Jacq., L. camara, Macfadyena unguiscati (L.) Gentry, Miconia calvescens D.C. (Seixas et
al., 2007), Mitracarpus hirtus (L.) DC (Pereira and
Barreto, 2005), Pereskia aculeata Miller (Pereira et al.,
2007), P. cattleianum (Pereira and Barreto, 2007) and
S. montevidensis Cham. and Schlecht. Publications describe the Brazilian mycobiota of 13 plant species, and
others are in preparation. These provide contributions
to the field of mycology and about potential biological
control agents for use in Brazil or abroad. Two of these
fungi have been used: Colletotrichum gloeosporioides
(Penz.) Sacc. f. sp. Miconiae, a pathogen of M. calvescens in Hawaii (Barreto et al., 2001) and P. tuberculatum
in Australia for the control of L. camara (Ellison et al.,
2006). A new species of Septoria is being evaluated for
S. terebinthifolius in quarantine in Florida. Preliminary
results of ongoing work at DFP/UFV on other weeds
are presented in these proceedings (Faria et al., 2008;
Macedo et al., 2008; Nechet et al., 2008; Pereira et al.,
2008; Soares and Barreto, 2008; Vieira et al., 2008).
Other scientists of this department are becoming involved. Two nematodes were found attacking M. calvescens: Ditylenchus drepanocercus Goodey, causing
angular leaf spots and a new species of Ditylenchus sp.,
which is being presently described and causes severe
galling on foliage. The former nematode was studied
in detail (Seixas et al., 2004a, 2004b), but priority is
being given to the latter nematode as it is easier to manipulate and causes a more severe disease. Its evaluation has provided promising results, and it is being
tested in quarantine in Hawaii. Bacteriologists were
also involved after a bacterial disease was found attacking Tradescantia fluminensis. The etiological agent
was identified as Burkholderia andropogonis; pathogenicity was demonstrated but host-range tests appear
to discourage further evaluations of its potential for a
classical introduction.
Latin American weed biological
control science at the crossroads
The challenges of re-inaugurating
classical weed biological control in
Latin American countries
Latin America still holds a plethora of natural enemies of important native and exotic weeds that may
be used in classical or inundative weed biological control worldwide. Sadly, the potential of the discipline
for tackling weed infestations in agricultural lands and
for mitigating biological invasions in Latin America
remains virtually untapped. To change this, there are
significant challenges to be overcome to raise the disciplines status and to maintain the structures developed
by past and present researchers. Some of these issues
will be discussed below.
In a recent assessment of weed biological control,
for classical biological control only about 5% of nearly
1000 programmes worldwide were implemented in
Latin America (Ellison and Barreto, 2004). The majority of the programmes were in the USA, Australia,
South Africa, Canada and New Zealand. The paucity of
programmes in Latin America was attributed to a series
of factors, among which are:
115
XII International Symposium on Biological Control of Weeds
1. The lack of long-term funding and tendency to withdraw funding as soon as one promising agent fails to
perform well regardless of other promising agents.
2. A lack of recognition from the public, government
officials and local scientists of the importance of exotic invasive weeds. Among weed scientists in Latin
America, there is a persistent myth that tropical Latin America is immune to invasions by exotic plants.
Ellison and Barreto (2004) refute this assumption
with examples of important exotic invasions into
natural, semi-natural ecosystems and in agricultural
systems. Only since the early 2000s has the threat
to agriculture, forestry, cattle ranching and the natural environment by introduced species (including
weeds) started to be recognized in Latin America.
There is a virtual absence of examples of practical
use of the inundative approach in weed biological control in Latin America. This mirrors a lack of commercial success for bioherbicides on a world scale as discussed by Evans et al. (2001), including reasons such
as: poor target selection, poor strain selection, strain
instability, mass production difficulties, low shelf-life,
problems with time of application and poor formulations. Nevertheless, this has not discouraged Latin
American scientists from attempting to develop such
products (Table 1).
The risk of depending on local heroes
and the need for a strategy for expanding
and perpetuating the discipline
discipline of weed biological control, allowing funding
of research and the establishment of new labs.
To consolidate the discipline in Latin America,
highly successful classical weed biological control
programmes should be implemented as quickly as possible. Such successful programmes must receive wide
publicity. Piggy-backing on other successful projects
is the only way to ensure such success. Pre- and postrelease ecological and economical evaluations would
allow for a clear demonstration of the benefits of such
projects and provide for the support of future proposals. Publicity is needed to educate the public, other scientists and the authorities, to encourage further funding
and promote new scientific vocations that will guarantee a future for the activity. Even within scientific
forums, there is little effort by Latin American weed
biological control scientists to publicize their activities
and their outstanding past record. Few Latin American
weed scientists are aware of the successful history of
weed biological control, the highly advantageous cost/
benefit ratios demonstrated for some important programmes, or even of the fact that the majority of weed
species in any country are aliens that could be targeted
by classical weed biological control. A more active role
should be played by the Latin American weed biological control scientists within the various discipline societies and at relevant meetings.
New rules for collecting in Latin America:
field scientists bureaucracy
Of the 17 research groups listed in Table 1, five
of the labs have either suspended or terminated their
work in this field as a result of changed political and
administrative priorities or retirement or death of the
lead scientist. Further, in most of cases (except USDAARS SABCL and CSIRO labs), activity relies on the
enthusiasm of one leading scientist. Several of these
scientists are either about to retire or already retired but
continuing their activities at a slowing pace. The sole
example of an ongoing programme of classical weed
biological control in Latin America, aimed at the weed
U. europaeus, relies almost completely on H. Norambuenas work in INIA-CRI Carillanca, Chile. This disciplines continuity cannot rely on isolate individuals.
For some labs, all the activity depends on a single or
few projects, and once funding becomes scarce or the
project ends, activity is likely to cease. Unfortunately,
in a limited period, a drastic reduction in the number
of weed biological control labs in Latin America may
take place. Latin America needs urgently to have more
examples such as that of J.H. Pedrosa-Macedo, a forest entomologist and weed biological control scientist
that prepared a second generation of scientists that are
active in the field. This depends very much on the governments and institutional recognition at international,
national and regional levels of the importance of the
In the past, insects or fungi attacking weeds were
generally regarded as irrelevant to everyone but the
weed biological control scientists. Field entomologists
and plant pathologists could explore distant places and
collect natural enemies. This was admissible as there
were no laws governing such procedures, and this remains the case for many countries. In the last two or
three decades, the public and government authorities
worldwide became aware of the value and the need to
preserve the biodiversity of ecosystems: international
agreements, such as the Convention on Biodiversity,
were developed and supported by national legislations.
An unfortunate consequence is that exploration for classical biological control agents is sometimes not treated
separately from profit-oriented bioprospection for new
drugs or other compounds. Some countries have novel
anti-biopiracy legislation with highly conservative
safeguards that make it difficult to conduct exploration.
To collect in Latin American, indeed anywhere in the
world, it is necessary to obtain updated information
about the legislation concerning collecting activities
for the countries to be visited. In Brazil, for example,
it is mandatory to work with local collaborators and
to leave duplicates of specimens in a Brazilian collection. Such cooperative links invariably prove beneficial
to the programme by allowing for systematic surveys
by in-country scientists and may contribute to raising a
116
Latin American weed biological control science at the crossroads
more permanent interest for the discipline. Legal issues
involved in collecting in Latin America are evolving
quickly. For instance, recently in Brazil, after lobbying
by the scientific community, legislation was revised,
releasing all scientific collection that does not involve
genes, organic molecules or extracts from native species in Brazil for commercial use from the previous
bureaucratic burdens established in 2001. Fortunately,
this has placed collecting biological control agents in
Brazil back at the situation it was in the 1990s.
In search of collaboration
for mutual benefit
Weed biological control science in Latin America
owes considerably to the weed biological control scientists of developed countries who have been actively engaged in training scientists from those countries in this
field and providing encouragement, partnership and
funding opportunities that allowed for several among
the existing labs to start and maintain their activities.
To maintain and develop these relationships so that
Latin America weed biocontrol science can prosper, it
is important to share resources and information and to
develop training for new biological control scientists. It
is important to establish cooperation on a target weed
by target weed basis, as it is not fair to undertake collections of biological control agents for a wide range
for weeds under a single agreement.
To increase cooperation and collaboration, there is
the possibility for mutual exchange of classical biological control agents. For instance, some of the worst weeds
in the Indian subcontinent are from Latin America (E.
crassipes, C. odorata, I. carnea, L. camara, M. micrantha, P. hysterophorus, among others); meanwhile,
among the worst weeds in Latin America are plants that
are natives of India, Pakistan and neighboring countries
(C. benghalensis, C. rotundus, Dichrostachys cinerea
(L.) Wight and Arn., H. coronarium, Rottboellia cochinchinensis, Saccharum spontaneum L.). Brazil and Argentina are the centre of origin for some noxious weeds
in South Africa (Campuloclinium macrocephalum, M.
unguis-cati and P. aculeata), while the African grass,
Eragrostis plana Nees, is problematic in Brazil causing severe losses to cattle ranchers (Kissmann, 1997).
We should work towards developing collaborative approaches that would provide mutual benefit rather than
the current mainly one-way movement of agents from
Latin America.
Target selection: a critical issue for the
discipline in Latin America
The target weeds which were chosen for bioherbicide development in Latin America (Table 1) are all
highly damaging in many important crops and are often
intractable by chemical means thus justifying a market
for a one-weed-product. In the case of classical biologi-
cal control, choosing the right weed can be more difficult. An obviously target for the weed scientists may
not be a priority for government or environmentalists.
In Brazil, where there has been no previous history of
a classical introduction against any weed, the choice of
the target is a delicate issue.
There are a number of very important weeds that
are also cultivated providing conflict of interest around
control. Examples are Pinus species and fodder grasses.
Clearly these are not target weeds suitable for Latin
America regarding the challenge of trying to raise
awareness and gain acceptance for biological control.
The focus should be on one or few selected exotic weed
species that will raise no conflicts and that cause significant environmental or agricultural problems so that
control brings uncontroversial benefit that could be
used for advertising the success of the discipline. Several weed species fit into this frame. Some were already
mentioned, such as E. plana and H. coronarium, but
others might be contemplated, such as Tecoma stans
(L.) Juss. ex Kunth (Bredow et al., 2004). Another option is to piggy-back on a successful programme developed elsewhere in the world, for example, Cryptostegia, which invades extensive areas of the Brazilian
northeast (Herrera and Major, 2006). A highly successful programme against this weed involving the introduction of two natural enemies from Madagascar was
carried out in Australia (Tomley and Evans, 2004).
Re-opening the incomplete project against itch grass,
R. cochinchinensis (Lour.) W.D. Clayton, may also be
helpful. This project was interrupted in 1990 before the
host-specific head smut fungus, Sporisorium ophiuri
(P. Henn.) Vanky could be released (Ellison and Evans,
1995; Reeder et al., 1996; Reeder and Ellison, 1999;
Snchez-Garita, 1999). A renewed effort from CAB
International and Centro Agronmico Tropical de Investigacin y Enseanzas (CATIE) might resolve the
pending issues allow for a pioneering introduction of a
weed biological control agent in Central America with
potential benefits for the whole of Latin America.
The situation in Latin America is currently favorable for actions that may consolidate weed biological
control and help it gain the respect as a valuable discipline that offers unique solutions to major weed problems. The moment requires firm action from the weed
biological control scientists in Latin America and their
cooperators.
Acknowledgements
The author wishes to acknowledge the following colleagues for providing relevant information and ideas
that were critical for the preparation of this manuscript:
C. Ellison, C. Wikler, H. Evans, J. Briano, J. Medal,
J.H. Pedrosa-Macedo, M. Vitorino, R. Pitelli, S. Mello
and T. Heard. The author also thanks the Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico
(CNPq) for financial support.
117
XII International Symposium on Biological Control of Weeds
References
Anonymous (2006) Grass agents slow to reveal their secrets.
Whats New in Biological Control of Weeds 36, 45.
vila, Z.R., Mello, S.C.M., Ribeiro, Z.M.A. and Fontes,
E.M.G. (2000) Produo de inculo de Alternaria cassiae. Pesquisa Agropecuria Brasileira 35, 533541.
vila, Z.R., Mello, S.C.M. and Borges Neto, C.R. (2005)
Influncia de meios de cultura e do regime de luz no cresci
mento e esporulao de Cercospora caricis. Summa Phytopathologica 31, 194200.
vila, Z.R. and Pitelli, R.A. (2005) Avaliao de adjuvantes
no crescimento e virulncia de Cercospora piaropi, agente
de controle biolgico do aguap. Summa phytopathologica 31, 201203.
Barreto, R.W. and Evans, H.C. (1998) Fungal pathogens of
Euphorbia heterphylla and E. hirta in Brazil and their potential as weed biocontrol agents. Mycopathologia 141,
2136.
Barreto, R.W., Evans, H.C. and Ellison, C.A. (1995) The
mycobiota of the weed Lantana camara in Brazil, with
particular reference to biological control. Mycological Research 99, 769782.
Barreto, R.W., Seixas, C.D.S. and Killgore, E. (2001) Colletotrichum gloeosporioides f.sp. miconiae: o primeiro fungo
fitopatognico brasileiro a ser introduzido no exterior para
o controle biolgico clssico de uma planta invasoras (Miconia calvescens). In: VII Simpsio de Controle Biolgico,
Livro de Resumos. Poos de Caldas, Brazil, p. 109.
Borges Neto, C.R., Mello, S.C.M., Ribeiro, Z.M.A., Malty, J.
and Fontes, E.M.G. (2000) Influncia da idade da planta,
perodo de umidificao e concentrao de inculo no
desenvolvimento de sintomas provocados por Cercospora caricis em tiririca. Fipopatologia Brasileira 25,
138142.
Borges Neto, C.R. and Pitelli, R.A. (2004) Adjuvantes e
herbicidas e a infectividade de Fusarium graminearum,
agente potencial de biocontrole de Egeria densa e Egeria
najas. Planta Daninha 22, 7783.
Bredow, E.A., Pedrosa-Macedo, J.H. and Vitorino, M.D.
(2004) Amarelinho Tecoma stans (Bignoniaceae)uma
ornamental multiuso ou plstica invasora. In: Bredow, E.A.
and Pedrosa-Macedo, J.H. (eds) Princpios e Rudimentos
do Controle Biolgico de Plantas. Imprensa UniversitriaUFPR, Curitiba, Brazil, pp. 3197.
Broughton, S. (2000) Review and evaluation of Lantana biocontrol programs. Biological Control 17, 272286.
Burckhardt D., Hanson P. and Madrigal L. (2005) Diclidophlebia lucens, n. sp (Hemiptera: Psyllidae) from Costa
Rica, a potential control agent of Miconia calvescens
(Melastomataceae) in Hawaii. Proceedings of the Entomological Society of Washington 107, 741749.
Center, T.D. (1982) The waterhyacinth weevils Neochetina
eichhorniae and N. bruchi. Aquatics 4, 1619.
Cronk, Q.C.B. and Fuller, J.L. (1995) Plant Invaders. Chapman and Hall, London, UK.
Cuda, J., Medal, J.C., Vitorino, M.D. and Habeck, D.H. (2005)
Supplementary host specificity testing of the sawfly Heteroperreyia hubrichi, a candidate for classical biological
control of Brazilian peppertree, Schinus terebinthifolius in
the USA. BioControl 50, 195201.
Cullen, J.M. (1974) Seasonal and regional variation in the
success of organisms imported to combat skeleton weed
Chondrilla juncea L. in Australia. In: Wapshere, A.J.
(ed.) Proceedings of the III International Symposium on
Biological Control of Weeds. Commonwealth Institute of
Biological Control, Ascot, UK, pp. 111117.
Davis, C.J., Yoshioka, E. and Kafler, D. (1992) Biological
control of lantana, prickly pear, and hamakua pamakani. In: Stone, C.P., Smith, C.W. and Tunison, J.T. (eds)
Alien Plant Invasions in Native Ecosystems of Hawaii:
Management and Research. Research Corporation of the
University of Hawaii, Natural Park Resources Study Unit,
Honolulu, HI, pp. 411431.
Elango, D.E., Holden, A.N.G. and Prior, C. (1993) The potential of plant pathogens collected in Trinidad for biological
control of Chromolaena odorata (L.) King and Robinson.
International Journal of Pest Management 39, 393396.
Ellison, C.A. and Barreto, R.W. (2004) Prospects for the
management of invasive alien weeds using co-evolved
fungal pathogens: a Latin American perspective. Biological Invasions 6, 2345.
Ellison C.A. and Evans H.C. (1995) Present status of the
biological control programme for the graminaceous weed
Rottboellia cochinchinensis. In: Delfosse, E.S. and Scott,
R.R. (eds.) Proceedings of the VIII International Symposium on Biological Control of Weeds. DSIR/CSIRO, Melbourne, Victoria, Australia, pp. 493500.
Ellison, C.A., Pereira, J.M., Thomas, S.E., Barreto, R.W. and
Evans, H.C. (2006) Studies on the rust Prospodium tuberculatum, a new classical biological control agent released against the invasive alien weed Lantana camara in
Australia. 1. Life-cycle and infection parameters. Australasian Plant Pathology 35, 309319.
Evans, H.C., Greaves, M.P. and Watson, A.K. (2001) Fungal
biocontrol agents of weeds. In: Butt, T.M., Jackson, C.
and Magan, N. (eds) Fungi as Biocontrol Agents. CAB
International, Wallingford, UK, pp. 169192.
Faria, A.B.V., Barreto, R.W. and Cuda J.P. (2008) Fungal
pathogens of Schinus terebinthifolius from Brazil as potential biocontrol agents. In: Julien, M.H., Sforza, R.,
Bon, M.C., Evans, H.C., Hatcher, P.E., Hinz, H.L. and
Rector, B.G. (eds) Proceedings of the XII International
Symposium on Biological Control of Weeds. CAB International, Wallingford, UK.
Forno I.W., Kassulke R.C. and Harley K.L.S. (1992) Host
specificity and aspects of the biology of Calligrapha pantherina (Col, Chrysomelidae), a biological-control agent
of Sida acuta [Malvaceae] and S. rhombifolia in Australia.
Entomophaga 37, 409417.
Gandolfo, D., McKay, F., Medal, J.C. and Cuda, J.P. (2007)
Open-field host specificity test of Gratiana boliviana
(Coleoptera: Chrysomelidae), a biological control agent
of tropical soda apple (Solanaceae) in the United States.
Florida Entomologist 90, 223228.
Goeden, R.D. (1978) Part II: Biological control of weeds. In:
Clausen, C.P. (ed.) Introduced Parasites and Predators
of Arthropod Pests and Weeds: A World Review. USDAAgriculture Handbook No. 480, USDA, Washington, DC,
pp. 357414.
Harley, K.L.S., Kassulke, R.C., Sands, D.P.A. and Day, M.D.
(1990) Biological control of water lettuce, Pistia stratiotes (Araceae) by Neohydronomus affinis (Coleoptera:
Curculionidae). Entomophaga 35, 363374.
Herrera, O. and Major, I. (2006) Visitantes perigosos no
Nordeste-Brasil Tropical. Cincia Hoje 38, 4244.
118
Latin American weed biological control science at the crossroads
Hight, S., Horiuchi, I., Vitorino, M.D., Wikler, C. and Ma
cedo, J.H.P. (2003) Biology, host specificity tests, and risk
assessment of the sawfly Heteroperreyia hubrichi, a potential biological control agent of Schinus terebinthifolius
in Hawaii. Biocontrol 48, 461476.
Hill, M.P. (1999) Biological control of red water fern, Azolla
filiculoides Lamarck (Pteridophyta: Azollaceae), in South
Africa. In: Olckers, T. and Hill, M.P. (eds) Biological
Control of Weeds in South Africa (19901998). African
Entomology Memoir 1, 119124.
Hill, M.P. and Cilliers, C.J. (1999) Azolla filiculoides Lamarck (Pteridophyta: Azollaceae), its status in South Africa and control. Hydrobiologia 415, 203206.
Jimenez, M.M. and Charudattan, R. (1998) Survey and evaluation of Mexican native fungi for potential biocontrol of
waterhyacinth. Journal of Aquatic Plant Management 36,
145148.
Jimenez, M.M. and Lopez, E.G. (2001) Host range of Cercospora piaropi and Acremonium zonatum, potential fungal
biocontrol agents for waterhyacinth in Mexico. Phytoparasitica 29, 175177.
Julien, M. and Griffiths, M.W. (1998) Biological Control of
Weeds: A World Catalogue of Agents and Their Target
Weeds. CAB International, Wallingford, UK.
Kissmann, K.G. (1997) Plantas Infestantes e Nocivas. Vo. 1.
2nd Ed. BASF, So Paulo, Brazil.
Lonsdale, W.M., Farrell, G. and Wilson, C.G. (1995) Biological control of a tropical weeda population-model and
experiment for Sida acuta. Journal of Applied Ecology
32, 391399.
Macedo, D.M., Barreto, R.W. and Pomella, A.W.V. (2008) Factors affecting mass production of Duosporium yamadanum
in rice grains. In: Julien, M.H., Sforza, R., Bon, M.C., Evans, H.C., Hatcher, P.E., Hinz, H.L. and Rector, B.G. (eds)
Proceedings of the XII International Symposium on Biological Control of Weeds. CAB International, Wallingford,
UK.
Marchiori, R., Nachtigal, G.F., Coelho, L., Yorinori, J.T. and
Pitelli, R.A. (2001) Comparison of culture media for the
mass production of Bipolaris euphorbiae and its impact
on Euphorbia heterophylla dry matter accumulation.
Summa Phytopathologica 27, 428432.
McFadyen, R.E.C. Vitelli M. and Setter, C. (2002) Host specificity of the rubber vine moth, Euclasta whalleyi PopescuGorj and Constantinescu (Lepidoptera: Crambidae: Py
raustinae): field host-range compared to that predicted by
laboratory tests. Australian Journal of Entomology 41,
321323.
McFadyen, R.E.C. and Willson, B. (1997) A history of biocontrol of weeds. In: Julien, M. and White, G. (eds) Biological Control of Weeds: theory and practical applications. ACIAR, Canberra, pp. 1722.
Medal, J. 2003. Historia del control biolgico de malezas.
In: Medal, J.H., Norambuena, H. and Gandolfo, D. (eds)
Memorias del Primer Curso Latinoamericano de Control Biolgico de Malezas. University of Florida-IFAS,
Gainesville, FL, pp. 112.
Medal, J.C., Pedrosa-Macedo, J.H., Vitorino, M.D., Habeck,
D.H., Gillmore, J.L. and Sousa, L.P. (1999) Host specificity of Heteroperreyia hubrichi Malaise (Hymenoptera,
Pergidae), a potential biological control agent of Brazilian
peppertree (Schinus terebinthifolius Raddi). Biological
Control 14, 6065.
Medal, J.C., Sudbrink, D., Gandolfo, D., Ohashi, D. and Cuda,
J.P. (2002) Gratiana boliviana, a potential biocontrol agent
of Solanum viarum: Quarantine host-specificity testing in
Florida and field surveys in South Amrica. Biocontrol 47,
445461.
Nechet, K.L., Barreto, R.W. and Mizobuti, E.S. (2006) Bipolaris euphorbiae as a biological control agent for wild
poinsettia (Euphorbia heterophylla): host-specificity and
variability in pathogen and host populations. BioControl
51, 259275.
Nechet, K.L., Vieira, B.S., Barreto, R.W. and Silva, A.A.
(2008) Combination of a mycoherbicide with selected
chemical herbicides for control of Euphorbia heterophylla.
In: Julien, M.H., Sforza, R., Bon, M.C., Evans, H.C.,
Hatcher, P.E., Hinz, H.L. and Rector, B.G. (eds.) Proceedings of the XII International Symposium on Biological
Control of Weeds. CAB International, Wallingford, UK.
Norambuena, H., Carrillo, R. and Neira, M. (1986) Introduccion, establecimiento y potencial de Apion ulicis como
antagonista de Ulex europaeus en el sur de Chile. Entomophaga 31, 310.
Norambuena, H. and Ormeo, J. (1991) Control biolgico de
malezas: fundamentos y perspectivas en Chile. Agricultura Technica 51, 210219.
Norambuena, H. and Piper, G.L. (2000) Impact of Apion ulicis Forster L. on seed dispersal. Biological Control 17,
267271.
Norambuena, H., Martnez, G., Carrillo, R. and Neira, M.
(2007) Host specificity and establishment of Tetranychus
lintearius (Acari: Tetranychidae) for biological control
of gorse, Ulex europaeus (Fabaceae) in Chile. Biological
Control 40, 204212.
Oehrens, E.B. (1963) Posibilidades de introduccin de hongos uredinales como factores de control biolgico para
malezas dicotiledneas de Chile. Simiente 23, 17.
Oehrens, E.B. (1977) Biological control of blackberry
through the introduction of rust, Phragmidium violaceum,
in Chile. FAO Plant Protection Bulletin 25, 2628.
Oehrens, E.B. and Gonzales, S.M. (1974) Introduccin de
Phragmidium violaceum (Shultz) Winter como factor de
control biologico de zarzamora (Rubus constrictus Ldr.
and M. y R. ulmifolius Schott.). Agro Sur 2, 3033.
Oehrens, E.B. and Gonzales, S.M. (1975) Introduccion de
Uromyces galegae (Opiz) Saccardo como factor de control biologico de alega (Galega officinalis L.). Agro Sur
3, 8791.
Oehrens, E.B. and Gonzales, S.M. (1977) Dispersin, ciclo
biologico y danos causados por Phragmidium violaceum
(Schultz) Winter en zarzamora (Rubus constrictus Ldr.
and M. y R. ulmifolius Schott.) en las zonas centro-sur y
sur de Chile. Agro Sur 5, 7385.
Olckers,T. and Hill, M.P. (1999) Biological control of weeds
in South Africa (19901998). African Entomology Memoir 1, 1182.
Ostermeyer, N. and Grace, B.S. (2007) Establishment, distribution and abundance of Mimosa pigra biological control
agents in northern Australia: implications for biological
control. Biocontrol 52, 703720.
Pedrosa-Macedo, J.H. (2000) Biology and behavior of the
strawberry guava sawfly, Haplostegus epimelas Konow,
1901 (Hym.: Pergidae) in southern Brazil. Proceedings
of the Entomological Society of Washington 102, 129
134.
119
XII International Symposium on Biological Control of Weeds
Pedrosa-Macedo, J.H., Dalmolin, A. and Smith, C.W. (2007a)
O Araazeiro: Ecologia e Controle Biolgico. FUPEF do
Paran, Curitiba, Brazil.
Pedrosa-Macedo, J.H., Weigert, J.K., Scapini, L.A., Niederhartamann, D., Bebiano, D.R. Fowler, S.V and Waipara,
N. (2007b) Estudos bioecolgicos sobre Tradescantia fluminensis (Commelinaceae) e seus inimigos naturais associados, no Paran. Floresta 37, 3141.
Pereira, J.M. and Barreto, R.W. (2000) Additions to the mycobiota of the weed Lantana camara (Verbenaceae) in
Southeastern Brazil. Mycopathologia 151, 7180.
Pereira, O.L. and Barreto, R.W. (2005) The mycobiota of the
weed Mitracarpus hirtus in Minas Gerais (Brazil), with
particular reference to fungal pathogens for biological
control. Australasian Plant Pathology 34, 4150.
Pereira, O.L. and Barreto, R.W. (2007) Fungos para o controle
biolgico de Psidium cattleianum. In: Pedrosa-Macedo,
J.H., Molin, A., and Smith, C.W. (eds.) O Araazeiro.
Ecologia e Controle Biolgico. FUPEF do Paran, Curitiba, Brazil, pp. 183190.
Pereira, O.L., Barreto, R.W., Cavalazzi, J.R.P. and Braun, U.
(2007) The mycobiota of the cactus weed Pereskia aculeata in Brazil, with comments on the life-cycle of Uromyces pereskiae. Fungal Diversity 25, 167180.
Pereira, O.L., Barreto, R.W. and Waipara, N. (2008) Pathogens from Brazil for classical biocontrol of Tradescantia fluminensis. In: Julien, M.H., Sforza, R., Bon, M.C.,
Evans, H.C., Hatcher, P.E., Hinz, H.L. and Rector, B.G.
(eds.) Proceedings of the XII International Symposium on
Biological Control of Weeds. CAB International, Wallingford, UK.
Perkins, R.C.L. and Swezey, O.H. (1924) The introduction
into Hawaii of insects that attack lantana. Experimental
Station of the Hawaii Sugar Planters AssociationBulletin
16, Honolulu.
Pitelli, R.L.C.M., Penariol, M.C., Pitelli, A.M.C.M. and Pitelli,
R.A. (2007) Host specificity of a Brazilian isolate of Alternaria cassiae (Cenargen CG593) under greenhouse
conditions. Phytoparasitica 35, 123128.
Reeder, R.H. and Ellison, C.A. (1999) Estado actual de la investigacin en control biologico clasico de la Rottboellia
cochinchinensis con el carbon Sporisorium ophiuri: potencial y riesgos (Current status of research into the classical biological control of Rottboellia cochinchinensis with
the smut Sporisorium ophiuri: potencial and risks). In:
Snchez-Garita, V. (ed.) Control Biologico de Rottboellia
cochinchinensis. CATIE, Costa Rica, pp. 101135.
Reeder, R.H., Ellison, C.A., and Thomas, M.A. (1996) Population dynamic aspects of the interaction between the
weed Rottboellia cochinchinensis (itch grass) and the potential biological control agent Sporisorium ophiuri (head
smut). In: Moran, V.C. and Hoffmann, J.H. (eds.) Proceedings of the IX International Symposium on Biological
Control of Weeds. University of Cape Town, South Africa,
pp. 205211.
Room, P.M., Harley, K.L.S., Forno, I.W. and Sands, D.P.A.
(1981) Successful biological control of the floating weed
Salvinia. Nature 294, 7880.
Sankaran, K.V., Puzari, K.C., Ellison, C.A., Sreerama Kumar,
P. and Dev, U. (2008) Field release of the rust fungus Puccinia spegazzinii to control Mikania micrantha in India:
protocols and awareness raising. In: Julien, M.H., Sforza,
R., Bon, M.C., Evans, H.C., Hatcher, P.E., Hinz, H.L. and
Rector, B.G. (eds.) Proceedings of the XII International
Symposium on Biological Control of Weeds. CAB International, Wallingford, UK.
Snchez-Garita, V. (1999) Control Biolgico de Rottboellia
cochinchinensis. CATIE, Costa Rica.
Segura, R. and Heard, T.A. (2004) The CSIRO Mexican Field
Station history and current activities. In: Cullen JM, Briese DT, Kriticos DJ, Lonsdale WM, Morin L, Scott JK.
(eds.) Proceedings of the XI International Symposium on
Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 145148.
Seixas, C.D.S., Barreto, R.W. and Killgore, E. (2007) Fungal pathogens of Miconia calvescens (Melastomataceae)
from Brazil, with reference to classical biological control.
Mycologia 99, 99111.
Seixas, C.D.S., Barreto, R.W., Freitas, L.G., Maffia, L.A.
and Monteiro, F.T. (2004a) Ditylenchus drepanocercus
(Nematoda), a potential biological control agent for Miconia calvescens (Melastomataceae): host-specificity and
epidemiology. Biological Control 31, 2937.
Seixas, C.D.S., Barreto, R.W., Freitas, L.G., Monteiro, F.T.
and Oliveira, R.D.L. (2004b) Ditylenchus drepanocercus rediscovered in the Neotropics causing angular leaf
spots on Miconia calvescens. Journal of Nematology 36,
481486.
Soares, D.J. and Barreto, R.W. (2008) Fungal survey for biocontrol agents of Ipomoea carnea from Brazil. In: Julien,
M.H., Sforza, R., Bon, M.C., Evans, H.C., Hatcher, P.E.,
Hinz, H.L. and Rector, B.G. (eds.) Proceedings of the XII
International Symposium on Biological Control of Weeds.
CAB International, Wallingford, UK.
Sosa, A.J., Traversa, M.G., Delhey, R., Kiehr, M., Cardo,
M.V. and Julien, M.H. (2008) Biological control of lippia
(Phyla canescens): surveys for the plant and its natural
enemies in Argentina. In: Julien, M.H., Sforza, R., Bon,
M.C., Evans, H.C., Hatcher, P.E., Hinz, H.L. and Rector,
B.G. (eds.) Proceedings of the XII International Symposium on Biological Control of Weeds. CAB International,
Wallingford, UK.
Spencer, N.R. and Coulson, J.R. (1976) The biological control of alligatorweed, Alternanthera philoxeroides, in the
United States of America. Aquatic Botany 2, 177190.
Tomley, A.J. and Evans, H.C. (2004) Establishment of, and
preliminary impact studies on, the rust, Maravalia cryptostegiae, of the invasive alien weed, Cryptostegia grandiflora in Queensland, Australia. Plant Pathology 53,
475484.
Trujillo, E.E. (2005) History and success of plant pathogens
for biological control of introduced weeds in Hawaii. Biological Control 33, 113122.
Vieira, B.S. and Barreto, R.W. (2005) Lewia chlamidosporiformans sp. nov. from Euphorbia heterophylla. Mycotaxon 94, 245248.
Vieira, B.S., Nechet, K.L. and Barreto, R.W. (2008) Lewia
chlamidosporiformans, a mycoherbicide for control of
Euphorbia heterophylla: isolate selection and mass production. In: Julien, M.H., Sforza, R., Bon, M.C., Evans,
H.C., Hatcher, P.E., Hinz, H.L. and Rector, B.G. (eds.)
Proceedings of the XII International Symposium on Bio-
120
Latin American weed biological control science at the crossroads
logical Control of Weeds. CAB International, Wallingford, UK.
Vitorino, M.D., Pedrosa-Macedo, J.H. and Smith, C.W.
(2000) The biology of Tectococcus ovatus Hempel (Heteroptera: Eriococcidae) and its potential as a biocontrol
agent of Psidium cattleianum (Myrtaceae). Proceedings
of the X International Symposium on Biological Control
of Weeds, Montana, pp. 651657.
Waterhouse, D.F.W. and Norris, K.R. (1987) Biological Control Pacific Prospects. Inkata Press, Melbourne, Victoria,
Austalia.
Wikler, C. and Vitorino, M.D. (2007) Entomofauna associada
ao araazeiro no Estado do Paran. In: Pedrosa-Macedo,
J.H., Dalmolin, A. and Smith, C.W. (eds.) O Araazeiro:
Ecologia e Controle Biolgico. FUPEF do Paran, Curitiba, Brazil, pp. 9398.
Wikler, C., Smith, C.W and Pedrosa-Macedo, J.H. (1996)
The stem-gall wasp Eurytoma sp. (Hymenoptera: Eurytomidae), a potential biological control agent against
Psidium cattleianum. In: IX International Symposium
on Biological Control of Weeds. Stellenbosch, p. 219
221.
Wilson, C.L. (1969) Use of plant pathogens in weed control.
Annual Review of Plant Pathology 7, 411433.
Yorinori, J.T. (1985) Biological control of wild poinsettia
(Euphorbia heterophylla) with pathogenic fungi. In: Delfosse, E.S. (ed.) Proceedings of the VI International Symposium on Biological Control of Weeds, 1985. Vancouver
Agriculture Canada, pp. 677681.
Yorinori, J.T. (1987) Controle biolgico de ervas daninhas
com microrganismos. In: Fundao Cargill (ed.) Anais
da II Reunio sobre Controle Biolgico de Doenas de
Plantas, 1987. Escola Superior de Agricultura Luis de
Queiroz, Piracicaba, So Paulo, pp. 2030.
Yorinori, J.T. and Gazziero, D.L.P. (1989) Control of wild
poinsettia (Euphorbia heterophylla) with Helminthosporium sp. In: Delfosse, E.S. (ed.) Proceedings of the VII
International Symposium on Biological Control of Weeds.
Istituto Sperimentale per la Patologia Vegetale, Rome,
Italy, pp. 571576.
121
Galling guilds associated with Acacia
dealbata and factors guiding selection
of potential biological control agents
R.J. Adair1
Summary
The Australian tree Acacia dealbata Link (Mimosaceae) invades natural ecosystems in both the
Northern and Southern Hemispheres, including areas beyond its natural range in Australia. Biological
control is under development in South Africa using the seed-feeding curculionid Melanterius maculatus Lea. A diverse range of galling insects occur on A. dealbata in Australia, most exhibiting high
levels of host specificity and niche partitioning within their host. Galling insects have successfully
contributed to biocontrol of other Acacia species in South Africa. Australian galling insects from
A. dealbata have considerable potential for adoption as novel or complementary biocontrol agents.
Factors governing the selection of potential agents are considered in the context of impact on the host,
efficacy and compatibility with the utilization of the host for timber, pulp, floriculture and fire-wood
harvesting, particularly in resource-poor regions of the world. The potential for biological control of
A. dealbata in invaded habitats in Australia is also discussed.
Keywords: wattle, conflict of interest, agent selection.
Introduction
Silver wattle, Acacia dealbata Link (Mimosaceae:
Botrycephalae), is a widespread and conspicuous tree
indigenous to forests and woodlands of southeastern Australia (Costermans, 1983). The species has a
broad habitat range with two sub-specific taxa (subsp.
dealbata, subsp. subalpina) that are delineated by altitude (Kodela and Tindale, 2001). A. dealbata is often abundant in early post-fire vegetation succession,
where mass germination of soil-stored seed is triggered
by burning. In its native habitat, A. dealbata provides
ecosystem functions such as food and habitat for fauna
(Broadhurst and Young, 2006) and fixing atmospheric
nitrogen. The species silvery bipinnate foliage and
abundant production of bright yellow flowers in winter
to early spring contributes to the popularity of A. dealbata in horticulture. Large, naturalized populations
of A. dealbata now occur in many countries and can
require management to protect natural and social assets
(Sheppard et al., 2006; Adair, 2008). Biological control
Department of Primary Industries, PO Box 48, Frankston, Victoria,
Australia 3199 <[email protected]>.
CAB International 2008
of A. dealbata occurs in South Africa where large-scale
invasions make other forms of suppression difficult to
implement. This paper examines the role of classical
biological control of A. dealbata using gall-forming
agents and how such agents may affect commercial and
utilitarian values of the host tree.
A. dealbatathe invader
While A. dealbata is native to eastern Australia, extensive and expanding naturalized populations occur in
south-west Western Australia, where the species was
introduced for horticultural purposes. Although southern Western Australia has an astoundingly rich native
Acacia flora (Hnatiuk and Maslin, 1988), there are no
native Botrycephalae, and very few Western Australian
acacias are large woody trees. Consequently, invasion
of A. dealbata into the native vegetation in Western
Australia may have undesirable ecological impacts, although quantitative impact data both in Australia and
elsewhere are lacking. In South Africa, A. dealbata has
been problematic as early as 1915 (Henkel, 1915) and
is now a weed of national importance due to negative
impacts on water management and biodiversity conservation (Le Maitre et al., 2002; Nel et al., 2004). More
recently, in Europe, A. dealbata was listed as one of the
122
Galling guilds associated with Acacia dealbata and factors guiding selection of potential biological control agents
top 20 invasive plants suggested as targets for biological control (Sheppard et al., 2006). Invasions in southern France post-1910 have progressively replaced local
vegetation including cork oaks, l`arbousier (Arbutus
unedo L.) and heather (http://www.worldwidewattle.
com/). A. dealbata is also naturalized in New Zealand,
western North America, Madagascar, Japan and Chile
(Randall, 2002).
Utilitarian values of A. dealbata
In Australia, A. dealbata is utilized in habitat restoration programs and urban landscaping projects. The
species is not utilized commercially, although pollen
used by honey bees contributes to the apiary industry.
In New Zealand and North America, A. dealbata is utilized in horticulture but with limited economic value.
In contrast, the exploitation of A. dealbata is well developed in southern Europe and South Africa where
the species services quite different industries in each
of these regions.
In Europe, A. dealbata was introduced around 1816
(Cavanagh, 2006) where acacias (mimosa) are grown
for horticultural and floricultural purposes. The mimosa cut flower industry in France occupies around
200 ha with an estimated value of 34 million/year
(Roland, 2006). Hybrids of A. dealbata and selected
cultivars form the basis of the industry and produce
flower crops between December and March. Whether
these selections and hybrids have naturalized in Europe is uncertain, but the creation and invasion of de
novo genotypes by hybridization can complicate classical biological control programs. Essential oils from
the flowers of A. dealbata are used as a fixative and
blending agent in the manufacture of high-grade perfumes and soaps, and the industry consumes around 1
million of refined mimosa absolute per year (Roland,
2006). More recently, the French tourism industry has
promoted the virtues of the Route de Mimosa during
the main flowering season with numerous festive activities linked to this period, undoubtedly contributing to
local economies in the Bormes-les-Mimosas to Grasse
region.
In South Africa, silvicultural operations use Acacia
mearnsii De Wild. And, to a limited extent, Acacia decurrens Willd. A. dealbata is not commercially cultivated, but extensive areas of naturalized and invasive
populations of A. dealbata in eastern South Africa are
the legacy of early experimental and development programs. Resource-poor communities utilize A. dealbata
for fuel wood, charcoal and construction timber where
harvesting is carried out ad hoc and driven by localized domestic needs (de Neergaard et al., 2005). The
contribution of A. dealbata to the regional and national
economies of South Africa has not been calculated, although limited and careful extrapolation from the cost
benefit analysis undertaken for A. mearnsii (de Wit et
al., 2001) could possibly be made. In the A. mearnsii
case, biological suppression programs were strongly
beneficial to the national interest.
Where A. dealbata threatens important assets, bona
fide utilitarian values need to be taken into account
when designing biological control strategies to reduce
levels of conflict of interest. Historically, potential conflicts of interests are avoided by: (1) not initiating biological control programs, (2) undertaking a costbenefit
analysis and proceeding with biological control where
it is in the public interest, or (3) by targeting specific
organs on the host and avoiding negative impacts on
utilitarian interests.
Biological control of Australian acacias
Classical biological control of Australian acacias
was pioneered in South Africa, where eight species
are currently subject to active research, development
or agent redistribution programs. All of these programs have succeeded in the establishment of one
or more agents, and several targets are now subject
to satisfactory levels of suppression (Dennill et al.,
1999; Hoffmann et al., 2002). Two general approaches to biological control of acacias have been adopted
in South Africa, each largely governed by the level
of conflict of interest with commercial or utilitarian interests. Economically important species (A.
mearnsii, Acacia melanoxylon R. Br., A. dealbata,
A. decurrens Willd.) are targeted solely for biological control of reproductive organs with seed-feeding
curculionids (Melanterius spp.) that have no negative impacts on vegetative growth of the host plant.
In contrast, acacia species of little or no economic
value (Acacia cyclops A. Cunn. ex G. Don., Acacia
longifolia (Andrews) Willd., Acacia saligna (Labill.)
H.L.Wendl., Acacia pycnantha Benth.) are subject to
biological control of a range of plant organs where
Melanterius spp. or flower-galling Cecidomyiidae
are used to target reproductive organs along with
Trichilogaster (Hymenoptera: Pteromlaidae) or Uromycladium (Fungi: Uredinales), which gall vegetative organs.
In Australia, biological control of invasive acacia
species that have transgressed substantial geographical
barriers (trans-continental invaders) is advocated. A.
dealbata invasions in Western Australia are suggested
as targets for biological control (Adair, 2008). Biological control of A. longifolia in Portugal has commenced
following successful control in South Africa (Sheppard
et al., 2006).
Galling agents and biological control
strategies for Australian acacias
Galling organisms vary in their impact on the host
plant depending largely on the mode of physiological
interaction with the host (innate impact), gall densi-
123
XII International Symposium on Biological Control of Weeds
ties, phenological synchronization, location on the host
and capacity to divert and accumulate resource allocation (Hartnet and Abrahamson, 1979; Dorchin et al.,
2006).
Dennill (1988) outlines ecological hypotheses underpinning the successful suppression of A. longifolia
in South Africa. In this system of forced commitment,
diversion of host resources to gall development occurs
at the expense of normal growth functions. In comparison, the flower-galling cecidomyiids Dasineura dielsi
Rbsaamen and Dasineura rubiformis Kolesik proposed for biological control of Australian acacias induce
gall structures with biomass and calorific allocations
the same as or less than normal fruit production.
Therefore, disruption to vegetative growth beyond that
created by normal fruit formation is unlikely (Adair,
2005). In such cases where host trees are prevented
from producing heavy fruiting loads, vegetative growth
was found to be either unaffected or accelerated. This
process is termed commitment release (Adair, 2005)
and may be applicable to situations where conflicts of
interest are associated with the targeting of vegetative
organs.
In the case of A. dealbata, a diverse assemblage
of galling agents is known with a range of innate impacts.
Galling biota of A. dealbata
in Australia
In an extensive survey of Acacia in southern Australia
(Adair, 2005), records and accessions of gall-forming
insects collected on A. dealbata were extracted and
combined with published data records. Thirteen gallinducing species were recorded on A. dealbata: seven restricted to reproductive organs; two restricted to
leaves; two restricted to stems of various size classes;
one restricted to vegetative and reproductive buds;
and two that attacked a range of host organs (Table 1).
More than half of the taxa (61%) belonged to the Cecidomyiidae (Diptera), a family that is well-known from
Australian Mimosaceae (Adair, 2005).
All recorded gall-forming biota from A. dealbata
have restricted host ranges, at least within the Botrycephalae, with the exception of the fungus U. notabile Mc
Alpine (Uredinales), which is recorded from numerous
bipinnate species of Australian Acacia (Marks et al.,
1982), although host-specific biotypes are known to occur within this genus (Morris, 1999). Perilampella sp.
(Pteromalidae) appears to be confined to A. dealbata,
and ?Cecidomyia sp. is restricted to a small group of
closely related Botrycephalae, where A. dealbata is its
principal host. Densities of gall-forming organisms associated with A. dealbata are generally low, but most
species are widespread within the natural distribution
of this species. Dasineura sp. 2 appears to be restricted
to south-west New South Wales.
The average dry weight of galled tissue compared
to the average weight of the same un-galled organs
was used in this study to indicate the general level or
direction of resource partitioning to gall structures
(Table 1). Galls that are heavier than normal un-galled
organs (positive gall biomass ratio) indicate a possible
resource sink. Conversely, galled tissues with lower
biomass than the same organs without galls may indicate the absence of such a resource sink. The were more
cecidogenic organisms associated with A. dealbata
with negative to neutral gall biomass ratios (61%)
than those with positive gall biomass ratios (38%;
Table 1).
Selection of potential galling agents
High costs and safety concerns in the development
and release of biological control agents necessitate
careful selection of organisms destined for detailed
evaluation. Five selection filters are proposed here for
pre-screening potential galling agents for A. dealbata,
(1) impact efficacy, (2) host specificity, (3) conflict of
interest, (4) climatic compatibility and (5) risk of parasitism.
Impact efficacy: Efficacy filters preceding host specificity evaluation can effectively narrow the range of
organisms for further consideration and may improve
the prospects for success (Raghu, et al., 2006). Ecological modelling designed to identify weak points
in the hosts life history (Briese, 2006), together with
pre-release impact assessment (McClay and Balciunas, 2006), are useful tools for quantitative efficacy
evaluation. However, manipulative techniques to
gauge density-based impacts of endophagous organisms, particularly on large woody plants, are somewhat problematic. While density-related impacts will
be important, they remain largely untestable for large
trees, except perhaps in situations where outbreak
populations occur in the agents natural range, e.g.
D. rubiformis in Western Australia (Adair, 2005). In
the case of A. dealbata, the innate impacts of galling
organisms form the initial efficacy filter, and organisms with impact-class scores of 1 and 2 (Table 1)
are dropped from further consideration. Modelling
insect reproductive capacity and survivorship predictions combined with estimation of population densities likely to achieve effective damage to the host may
be the only realistic way of further filtering for efficacy
of A. dealbata agents.
Host specificity: Galling organisms are generally host
specific (monophagous) or have a host range restricted
to closely related plant species (stenophagous); (Ananthakrishnan, 1984). Nearly all galling agents on A. dealbata are stenophagous or polyphagous within Acacia.
Host-specificity filtering needs to consider commercial
and utilitarian interests of potentially susceptible nontarget taxa and should be performed in a regional context
124
Gall-forming organisms from Acacia dealbata in Australia.
Family
Genus species
Common name
Host organ
Host rangea
Biologyb
Cecidomyiidae
Asphondylia sp. 1
Asphondylia sp. 2
Seed galler
Pubescent bud
galler
Eastern budseed galler
Inflated floret
galler
Fleshy floret
galler
Hollow galler
Red plush
galler
Pinnule galler
Bud-shoot
galler
Rachis galler
Lop-sided stem
galler
Small-stem
galler
Galling rust
fungus
Seed
Flower bud
S
S
U, C
M, C
Flower
bud, seed
Ovary
Asphondylia sp. 3
Dasineura pilifera
Dasineura sp. 1
Dasineura sp. 2
?Cecidomyia sp.
Tetrastichinae
Undescribed genus
Perilampella
?hecteaeush
Perilampella sp.i
Undetermined
Lepidoptera
Undetermined
Uredinales
Uromycladium
notabile
125
Pteromalidae
Impact
classc
3
3
Geographical
distributiond
2
2
Conflict of
intereste
1
1
Gall: organ
biomass ratiof
N
Region excluded
for biocontrolg
A, SA,E
M,A
C
U, S
Ovary
U, S
Ovary
Ovary
S
S
?U, S
U, C
3
3
1
2
1
1
Pinnule
Bud
S
S
?M, S
U, C
2
6
2
2
2
2
+
+
A, SA, E
E
M (Ad)
S
U, C
?M, C
5
5
3
3
2
3
+
+
E
A, SA, E
Stem
?M, C
A, SA, E
Stem, fruit,
leaf
M,C
Leaf rachis
Stem
A, SA,E
Ad, Acacia dealbata; M, monophagous; S, stenophagousrestricted to the Botrycephalae; P, polyphagousfound on species within a number of subgenera.
U, Univoltine; M, multivoltine; A, alternates between host organs; S, pupation occurs in soil; C, life cycle completed within gall.
c
2, Minor disruption to normal growth processes but impact unlikely to affect host fitness even in high densities; 3, impact restricted to reproductive organs and gall biomass equal to or less than fruit biomass; 5, moderate disruption to normal growth processes; 6, significant disruption to normal growth processes.
d
1, Restricted within natural range of A. dealbata; 2, widespread across natural range of A. dealbata; 3, distribution uncertain.
e
1, Lowwill not conflict with commercial interests associated with Australian Acacias; 2, moderatepotential to conflict with some commercial interests; 3. highpotential to conflict with most commercial interests.
f
, Gall biomass is lower than host organ; N, gall biomass is approximately equal to host organ; +, gall biomass is greater than host organ.
g
A, Australia; SA, South Africa,; E, Europe.
h
Records from A. mearnsii by van den Berg (1980, unpublished records) are likely to be the result of misidentification of the host plant.
i
Taxonomic position requires verification using molecular diagnostics, and feeding range assessed using no-choice tests.
a
Galling guilds associated with Acacia dealbata and factors guiding selection of potential biological control agents
Table 1.
XII International Symposium on Biological Control of Weeds
by addressing local industry issues. Using natural host
records, three regionally based host-specificity filters
are established for Australia, South Africa and Europe.
In Western Australia, several Botrycephalae acacias are
cultivated commercially, but none are used by the silvicultural industries. Based on host specificity, all galling organisms restricted to the Botrycephalae should
be considered as potential biocontrol agents for A. dealbata. Only Asphondylia sp. 3 and the polyphagous
biotypes of Uromycladium should be excluded on this
basis. In South Africa, A. mearnsii and A. decurrens are
commercially exploited, and galling organisms known
from these hosts that have positive gall biomass ratios
and are capable of affecting vegetative growth need to
be excluded. Therefore, Asphondylia sp. 3, a new cecidomyiid genus (pinnule galler), Tetrastichinae sp., the
stem-galling lepidopteran, and the polyphagous biotypes of Uromycladium should be excluded in South
Africa. In Europe, organisms that attack commercially
important Botrycephalae acacias are excluded where
host structures of importance are affected (Table 1).
Conflicts of interest: The conflict of interest filter relates to direct impacts on the targeted host, A. dealbata.
In Europe and South Africa, organisms that directly or
indirectly (e.g. due to positive gall biomass ratios) affect structures of importance are excluded. Therefore,
organisms affecting vegetative organs and pre-flowering structures are excluded as potential agents for Europe. In South Africa, only organisms with potential to
affect vegetative growth are excluded.
Climatic compatibility: Close climate matching between natural and intended areas of introduction may
influence the success of biological control outcomes
(Dhileepan et al., 2006). The galling organisms associated with A. dealbata occur over a broad geographic
area and climate range, including considerable variation in altitude and rainfall patterns. The only clear
exclusion based on climatic considerations was Dasineura sp. 2, where known occurrences have a low
match (using Climex and Climate) with introduced
occurrences of A. dealbata in Western Australia. High
match levels for this species occur in Europe and South
Africa (unpublished data).
Parasitism: Endophagous organisms tend to be susceptible to parasitism (Askew, 1980), and failure of some
biocontrol programs using galling agents are attributed
to high parasitism levels (Muniappan and McFadyen,
2005). Methods for predicting parasitism impacts remain elusive (Adair and Neser, 2006). However, gallforming agents that experience low parasitism levels
(<30%) have been successful in suppressing their hosts
(Muniappan and McFadyen, 2005). Larger gall size and
chamber number can reduce parasitism levels in some
gall-forming agents (Manongi and Hoffmann, 1995),
however, this association is not consistent (Waring and
Price, 1989). High endemic parasitism of Australian
and South African analogues of Asphondylia sp. 2 and
Asphondylia sp. 3 that induce single-chambered, thinwalled galls, warrant exclusion due to the high probability of attack by parasitoids. In contrast, the galls
of ?Cecidomyia sp. consist of long compacted hairs
that surround a hard woody kernel. In Australia, ?Cecidomyia sp. is parasitized by a specialized dipteran
(Chloropidae: Gaurax sp.), but remarkably few hymenopterans, which typically dominate cecidomyiid
galls. The parasitism risks of this insect in a biological control context may be lower than the more simply
structured Asphondylia spp. galls.
Conclusions
The successful suppression of invasive Australian
acacias using classical biological control has been
achieved through the use of galling agents that induce
a debilitating resource allocation commitment in their
host (Dennill, 1988) and seed-feeding agents, either in
combination or as a single-agent introduction. A. dealbata is utilized for commercial and domestic purposes
in both the Southern and Northern Hemispheres. The
biological control strategy adopted for invasive noncommercial acacias in South Africa has limited application for A. dealbata, except in Western Australia
where the level of conflicts of interest is low. In other
regions, host- and organ-specific gall-inducing organisms known from A. dealbata may contribute to the
biological suppression of this plant. Control programs
that focus on suppression of seed-producing organs to
avoid conflicts of interest need to be guided by the potential of the agents to achieve ecologically meaningful
levels of control. While host impacts induced by endophagous organisms creating resource sinks on vegetative growth are difficult to test or predict a priori,
control targets for solely seed-reducing organisms are
more achievable through modelling of the life history
attributes and population dynamics of the host. A. dealbata may be a density-independent species, and therefore, suppression by seed-reducing organisms, such as
Melanterius maculatus Lea and Bruchophagus acaciae
(Cameron) (Hymenoptera) would need to achieve very
high levels of control before population-level impacts
can be obtained.
Galling organisms for A. dealbata are available to
contribute to the reduction of reproductive output of
A. dealbata, even in situations where the host is commercially utilized. However, a compatible combination of agents is more likely to achieve high levels of
seed reduction than a single agent alone, based on the
enormous resource allocation of A. dealbata to flower
and fruit production. Seed-feeding organisms that can
find food at low density levels, such as Melanterius
ventralis Lea (Donnelly and Hoffmann, 2004), but respond rapidly to sudden increases in food availability
may work well in combination with organisms that
attack pre-fruiting stages of the reproductive cycle. The
126
Galling guilds associated with Acacia dealbata and factors guiding selection of potential biological control agents
sequence of introduction of combinations of biological
control agents remains debatable (Impson et al., 2008),
but introductions following a reverse phenological sequence (seed-feeders before flower-feeders) may favour the establishment of organisms that target the end
of the reproductive process, which could otherwise be
disadvantaged (Briese, 2006).
A series of five selection filters presented here identifies gall-inducing organisms potentially suitable for
suppression of A. dealbata at three levels of conflict
of interest: low (Australia), moderate (South Africa)
and high (Europe). Efficacy of impact should precede
other selection filters (Raghu et al., 2006), and while
difficult to quantify for organisms restricted to reproductive organs on large perennial trees, manipulative
techniques are technically possible (Balciunas and
Burrows, 1993).
Acknowledgements
I thank S. Neser, N. Dorchin and S. Raghu for valuable contributions that assisted in improving an earlier draft of this paper. John Weiss performed climate
matches for Dasineura sp. 2. The Department of Water
Affairs and Forestry, South Africa (Working for Water
Program) contributed to the survey of gall-forming organisms on Australian acacias. David Yeates (ANIC)
kindly identified the Chloropid parasitoid.
References
Adair, R.J. (2005) Seed-reducing Cecidomyiidae as potential
biological control agents for invasive Australian wattles in
South Africa, particularly Acacia mearnsii and A. cyclops.
PhD Thesis, University of Cape Town, South Africa.
Adair, R.J. (2008) Biological control of Australian native
plants, in Australia, with an emphasis on acacias. Muelleria 26, 6778.
Adair, R.J. and Neser, O. (2006) Gall-forming Cecidomyiidae
from acacias: can new parasitoid assemblages be predicted? In: Ozaki, K., Yukawa, J., Ohgushi, T. and Price, P.W.
(eds) Galling Arthropods and their Associated Ecology
and evolution. Springer-Verlag, Tokyo, pp. 133147.
Ananthakrishnan, T.N. (1984) Adaptive strategies in cecidogenous insects. In: Ananthakrishnan, T.N. (ed.) Biology
of Gall Insects. Oxford and IBH Publishing, New Delhi,
India, pp. 19.
Askew, R.R. (1980) The diversity of insect communities in
leaf mines and plant galls. Journal of Animal Ecology 49,
817829.
Balciunas, J.K. and Burrows, D.W. (1993) The rapid suppression of the growth of Melaleuca quinquenervia saplings
in Australia by insects. Journal of Aquatic Plant Management 31, 265270.
Briese, D.T. (2006) Can an a priori strategy be developed for
biological control? The case of Onopordum spp. thistles in
Australia. Australian Journal of Entomology 45, 317323.
Broadhurst, L.M. and Young, A.G. (2006) Reproductive
constraints for the long-term persistence of fragmented
Acacia dealbata (Mimosaceae) populations in southeast
Australia. Biological Conservation 133, 512526.
Cavanagh, T. (2006) Historical aspects of wattles: the cultivation of Australian acacias in Great Britain and Europe
during the 18th and 19th centuries. In: Proceedings Acacia2006. Knowing and growing Australian wattles. SGAP
Victoria, Australia, pp. 6982.
Costermans, L. (1983) Native trees and shrubs of south-eastern Australia. Rigby, Adelaide, Australia, 422 pp.
Dennill, G.B. (1988) Why a gall former can be a good biocontrol agent: the gall wasp Trichilogaster acaciaelongifoliae
and the weed Acacia longifolia. Ecological Entomology
13, 19.
Dennill, G.B., Donnelly, D., Stewart, K. and Impson, F.
(1999) Insects used for the biological control of Australian Acacia species and Paraserianthes lophantha (Willd.)
Nielsen (Fabaceae) in South Africa. African Entomological Memoir 1, 4554.
de Neergaard, A., Saarnak, C., Hill, T., Khanyile, M., Berosa,
A.M., Birch-Thomsen, T. (2005) Australian wattle species in the Drakensberg region of South Africaan invasive alien or a natural resource? Agricultural Systems
85, 216233.
de Wit, M.P., Crookes, D.J. and van Wilgen, B.W. (2001)
Conflicts of interest in environmental management: estimating the costs and benefits of a tree invasion. Biological
Invasions 3, 167178.
Dhileepan, K., Wilmont Senaratne, K.A.D. and Raghu, S.
(2006) A systematic approach to biological control agent
exploration and prioritisation for prickly acacia (Acacia
nilotica ssp. indica). Australian Journal of Entomology
45, 303307.
Donnelly, D. and Hoffmann, J.H. (2004) Utilisation of an unpredictable food source by Melanterius ventralis (Coleoptera: Curculionidae), a seed-feeding biological control
agent of Acacia longifolia (Mimosaceae) in South Africa.
Biocontrol 49, 225235.
Dorchin, N., Cramer, M. and Hoffmann, J.H. (2006) Photosynthesis and sink activity of wasp-induced galls in Acacia pycnantha. Ecology 87, 17811791.
Hartnet, D.C. and Abrahamson, W.G. (1979) The effect of
stem gall insects on life history patterns in Solidago canadensis. Ecology 60, 910917.
Henkel, J.S. (1915) Report of the Conservator of Forests.
Department of Water Affairs and Forestry, South Africa,
pp. 1617.
Hnatiuk, R.J. and Maslin, B.R. (1988) Phytogeography of
Acacia in Australia in relation to climate and species richness. Australian Journal of Botany 36, 361383.
Hoffmann, J.H., Impson, F.A.C., Moran, V.C. and Donnelly,
D. (2002) Trichilogaster gall wasps (Pteromalidae) and
biological control of invasive golden wattle trees (Acacia
pycnantha) in South Africa. Biological Control 25, 6473.
Impson, F., Moran, V.C., Kleinjan, C., Hoffmann, J.H. and
Moore, J.A. (2008) Multiple species introductions of
biological control agents against weeds: look before you
leap. In: Proceedings of the 12th International Symposium
on Biological Control of Weeds (this volume).
Kodela, P.G. and Tindale, M.D. (2001) Acacia dealbata
subsp. subalpina (Fabaceae: Mimosoideae), a new subspecies from south-eastern Australia. Telopea 9, 319322.
Le Maitre, D.C., van Wilgen, B.W., Gelderblom, C.M., Bailey, C., Chapman, R.A. and Nel, J.A. (2002) Invasive alien
127
XII International Symposium on Biological Control of Weeds
trees and water resources in South Africa: case studies of
the cost and benefits of management. Forest Ecology and
Management 160, 143159.
Manongi, F.S. and Hoffmann, J.H. (1995) The incidence of
parasitism in Trichilogaster acaciaelongifoliae, a gallforming biological control agent of Acacia longifolia in
South Africa. African Entomology 3, 147151.
Marks, G.C., Futher, B.A. and Walters, N.E.M. (1982) Tree
diseases in Victoria. Forests Commission Victoria, Melbourne, Australia, 149 pp.
McClay, A. and Balciunas, J.K. (2006) The role of pre-release
efficacy assessment in selecting classical biocontrol agents
for weedsapplying the Anna Karenina principle. Biological Control 35, 197207.
Morris, M.J. (1999) The contribution of the gall-forming rust
fungus Uromycladium tepperianum (Sacc.) McAlp. to the
biological control of Acacia saligna (Labill.) Wendl. (Fabaceae) in South Africa. African Entomology Memoir 1,
125128.
Muniappan, R. and McFadyen, R.E. (2005) Gall-inducing
arthropods used in the biological control of weeds In: Raman, A. Schaefer, C.W. and Withers, T.M. (eds) Biology,
ecology, and evolution of gall-inducing arthropods, Vol 2.
Science Publishers, Enfield, NH, pp. 709730.
Nel, J.A., Richardson, D.M., Rouget, M., Mgidi, T.N.,
Mdzeke, N., Le Maitre, D., van Wilgen, B.W., Schonegevel, L., Henderson, L. and Neser, S. (2004) A proposed classification of invasive plant species in South
Africa: towards prioritising species and areas for management action. South African Journal of Science 100,
5364.
Raghu, S., Wilson, J.R. and Dhileepan, K. (2006) Refining
the process of agent selection through understanding plant
demography and plant response to herbivory. Australian
Journal of Entomology 45, 308316.
Randall, R. (2002) A Global Compendium of Weeds. R.G. and
F.J. Richardson, Melbourne, Australia, 905 pp.
Roland, W.A. (2006) Australian acacias in Europe. In: Proceedings Acacia2006. Knowing and growing Australian
wattles. SGAP Victoria, Australia, pp. 8386.
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006) Top
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 93117.
Waring, G.L. and Price, P.W. (1989) Parasitoid pressure and
the radiation of a gall forming group (Cecidomyiidae:
Asphondylia spp.) on creosote bush (Larrea tridentata).
Oecologia 79, 293299.
128
Biological control of Miconia calvescens
with a suite of insect herbivores from Costa
Rica and Brazil
F.R. Badenes-Perez,1,2 M.A. Alfaro-Alpizar,3
A. Castillo-Castillo3 and M.T. Johnson4
Summary
Miconia calvescens DC. (Melastomataceae) is an invasive tree considered the most serious threat
to the natural ecosystems of Hawaii and other Pacific islands. We evaluated nine species of natural
enemies that feed on inflorescences or leaves of M. calvescens for their potential as biological control
agents, comparing their impact on the target plant, host specificity, and vulnerability to biotic interference. Among herbivores attacking reproductive structures of M. calvescens, a fruit-galling wasp
from Brazil, Allorhogas sp. (Hymenoptera: Braconidae), and a flower- and fruit-feeding moth from
Costa Rica, Mompha sp. (Lepidoptera: Momphidae), were the most promising agents studied. The
sawfly Atomacera petroa Smith (Hymenoptera: Argidae) from Brazil was thought to have the highest
potential among the defoliators evaluated.
Keywords: herbivory, host specificity, biotic interference.
Introduction
Miconia calvescens DC. (Melastomataceae) is a small
tree native to Central and South America that is considered a serious threat to natural ecosystems in Hawaii
and other Pacific islands because of its ability to invade
native forests (Medeiros et al., 1997). Its devastating
effects are most evident in Tahiti, where it has displaced
over 65% of the native forest and threatens many endemic species (Meyer and Florence, 1996). Herbicidal
and mechanical removal are the main methods used to
contain the spread of M. calvescens, but control is difficult and costly, especially in remote areas (Medeiros
et al., 1997; Kaiser, 2006). Since M. calvescens is a
tree of significant size (up to 15 m high) and prolific
reproduction (Medeiros et al., 1997; Meyer, 1998), a
combination of agents attacking vegetative and repro-
Pacific Cooperative Studies Unit, University of Hawaii at Manoa,
Honolulu, HI 96822, USA.
2
Current address: Department of Entomology, Max Planck Institute for
Chemical Ecology, Hans-Knoell-Str. 8, D-07745 Jena, Germany.
3
Escuela de Biologia, Universidad de Costa Rica, San Jose, Costa Rica.
4
Institute of Pacific Islands Forestry, USDA Forest Service, Pacific
Southwest Research Station, Volcano, HI 96785, USA.
Corresponding author: F.R. Badenes-Perez <fbadenes-perez@ice.
mpg.de>.
CAB International 2008
ductive structures may be necessary to achieve effective biological control.
Several insects and pathogens have been identified
as potential agents in surveys conducted in the native
range of M. calvescens in Brazil and Costa Rica by
Johnson (unpublished data) and others (Burkhart, 1995;
Barreto et al., 2005; Picano et al., 2005). In the interest of avoiding unnecessary introductions and making
efficient use of limited resources to evaluate potential
agents, we wish to prioritize future work and focus on
highly host-specific agents that hold the greatest promise
for impacting M. calvescens in Hawaii. Because biotic
interference can be a significant obstacle to successful
weed biocontrol in Hawaii, we have been attempting to
predict which potential agents are most susceptible to
parasitoids, predators and pathogens (Johnson, in this
Proceedings). In this paper, we evaluate nine insect species that feed on inflorescences or leaves of M. calvescens based on impact on M. calvescens, host specificity
and vulnerability to natural enemies.
Methods and materials
In Brazil, we studied four species attacking M. calvescens: a leaf-rasping sawfly, Atomacera petroa Smith
(Hymenoptera: Argidae); a defoliating caterpillar, Antiblemma leucocyma Hampson (Lepidoptera: Noctui
dae);a fruit-galling wasp, Allorhogas sp. (Hymenoptera:
129
XII International Symposium on Biological Control of Weeds
Braconidae); and a fruit-boring weevil, Apion sp. (Coleoptera: Curculionoidea). In Costa Rica, we evaluated
five species of Lepidoptera: an undescribed Mompha
sp. (Momphidae), Erora opisena (Druce) (Lycaenidae), Temecla paron (Godman and Salvin) (Lycaenidae) and Parrhasius polibetes (Cramer) (Lycaenidae),
all of which fed on inflorescences, and the leaf-roller
Ategumia lotanalis (Crambidae).
To quantify the impact of the leaf feeders on M. calvescens, we observed the percentage of leaves attacked
in the field, and we measured area damaged per leaf
using WinFOLIA leaf area analysis software (Regent
Instruments Inc., 2003). To quantify the impact of the
inflorescence feeders, the percentage of flower and fruit
attacked by each species in the field were determined.
Additionally, in the case of internal fruit feeding,
we compared the number of seeds in infested versus
healthy fruits. Host specificity of each insect was assessed through a combination of observations of plants
growing in association with M. calvescens at field sites,
laboratory feeding tests and knowledge of host ranges
of related insects. Our evaluation of the potential for
biotic interference was based upon evidence of attack
by natural enemies in the native range and comparison
with enemies known for related herbivores in Hawaii.
For each insect, each criterion (impact on plant, host
specificity, potential for biotic interference) was assigned a score from 1 (low) to 3 (high). An overall assessment of each potential biocontrol agent was made
by summing scores across criteria.
Results
Host specificity
Gall-forming Allorhogas spp. represent a highly diverse group with each species being quite host specific
(Hanson, personal comment), like most gall-formers
tend to be (Julien and Griffiths, 1998; Dennill et al.,
1999; Hoffmann et al., 2002). Specificity of Apion sp.
has not been tested, but adults were found only on M.
calvescens in the field. Apion sp. tend to be host specific and have been used with some success in weed
biocontrol (Julien and Griffiths, 1998; McClay and De
Clerck-Floate, 1999; Norambuena and Piper, 2000).
Mompha sp. has been reared only from fruits of M.
calvescens in Costa Rica, where fruits of several Miconia spp. and other Melastomataceae have been repeatedly sampled (Chacn, 2007). E. opisena, P. polibetes
and T. paron were only seen on M. calvescens in our
field surveys (Badenes-Perez, unpublished data), but
no focussed host specificity studies have been conducted. Larvae of P. polibetes have also been found
feeding on Euphorbiaceae (Zikan, 1956), Leguminoseae (DAraujo e Silva et al., 1968), Malpighiaceae
(http://janzen.sas.upenn.edu/) and Vochysiaceae (Diniz
and Morais, 2002). Larval food plants of neotropical
lycaenids are poorly known, but most are thought to
be polyphagous (Downey, 1962; Robbins and Aiello,
1982). Studies in the laboratory and the native habitat
of A. leucocyma and A. petroa indicated that they only
attacked M. calvescens, while A. lotanalis attacked
other Melastomataceae besides M. calvescens in
the laboratory (Badenes-Perez and Johnson, 2007a;
Badenes-Perez and Johnson, 2008).
Impacts
Biotic interference
Each of the fruit-feeding species caused variable levels of damage in the field, usually not attacking a high
percentage of fruit but damaging moderate to high levels of seeds within the fruit they did attack, e.g. 80% and
60% reduction of seeds in attacked fruits for Allorhogas
and Apion sp., respectively (Badenes-Perez and Johnson, 2007b). In addition, evidence of Apion sp. causing
premature abscission of fruits suggests that this species
could indirectly reduce viability and germination of
seeds. Impacts by the three species of Lycaenidae can be
very high because each larva completely consumed large
portions of an inflorescence before flowering or many
individual immature fruits after flowering (Badenes-
Perez, unpublished data). Among defoliators, A. leucocyma and A. lotanalis were considered to have very
high impact based on levels of damage seen in the field
as well as leaf area consumed per insect (Badenes-
Perez and Johnson, 2008). Both of these lepidopterans
attacked young foliage in addition to older leaves, with
potentially high costs to plant fitness. In contrast, the
sawfly A. petroa was found to attack primarily older
leaves, and each larva removed relatively modest areas
of leaf tissue (Badenes-Perez and Johnson, 2007a).
Allorhogas sp. was sometimes attacked by a eulophid parasitoid (Hymenopetera: Eulophidae: Tetrastichinae) in its natural habitat in Brazil, but there are no
Allorhogas sp. present in Hawaii, and it appears likely
that specialized enemies of this gall wasp are absent
(Badenes-Perez and Johnson, 2007b). No natural enemies of Apion sp. were observed in the field in Brazil, but
opportunities to assess biotic interference were limited
by low densities of this insect (Badenes-Perez and Johnson, 2007b). A fungal pathogen is thought to limit the effective range of biocontrol by the gorse herbivore Apion
ulicis (Forsters) (Coleoptera: Curculionoidea) in Hawaii
(Julien and Griffiths, 1998). It is therefore possible that
our species from M. calvescens might be similarly affected. Despite the presence of several parasitoids of
Mompha sp. in Costa Rica, rates of parasitism were relatively low (Alfaro-Alpizar, unpublished data). A relative occupying a very similar niche, Mompha trithalama
Meyrick (Lepidoptera: Momphidae), introduced from
Trinidad to Hawaii, is well established and attacks a
high percentage of Clidemia hirta (L.) D. Don. (Melastomataceae) fruits in Hawaii (Conant, 2002), although its
population dynamics and overall efficacy have not been
130
Biological control of Miconia calvescens with a suite of insect herbivores from Costa Rica and Brazil
assessed in detail. In general, insects feeding internally
in M. calvescens fruits are expected to escape impacts
from generalist enemies in Hawaii. The probability of
biotic interference in Hawaii was considered moderate
to high for E. opisena, P. polibetes and T. paron because,
feeding externally, they would be exposed to a variety of
generalist enemies of Lepidoptera. Although they have
not been well studied, there are a few species of native
and introduced lycaenids in Hawaii that may already
have specialized enemies. The mimetic appearance of
our three species on M. calvescens may, however, help
them avoid some predators and parasitoids. In fact, parasitism of these species in their native Costa Rican range
was low (Badenes-Perez and Johnson, 2007a).
Biotic interference was considered highly probable for A. leucocyma because of the high levels of
parasitism found in Brazil (Badenes-Perez and Johnson, 2008) and because of high parasitism of Antiblemma acclinalis Hbner (Lepidoptera: Noctuidae),
established but apparently ineffective as a biological
control agent of C. hirta in Hawaii (Conant, 2002).
Biotic interference also seems likely for A. lotanalis
because several species of parasitoids and a hemipteran predator were observed attacking it in Costa Rica
(Castillo-Castillo, unpublished data) and because another biocontrol agent of C. hirta, Ategumia matutinalis (Guenee) (Lepidoptera: Crambidae), also appears
to be highly parasitized in Hawaii (Conant, 2002). In
contrast to these lepidopterans, the sawfly A. petroa
was considered less likely to be attacked in Hawaii because no parasitoids and predators were observed in
the natural habitat of A. petroa in Brazil and because
there are no native species of Argidae and only two
other introduced sawflies in Hawaii (Badenes-Perez
and Johnson, 2007a).
When the criteria of impact, specificity and biotic
interference were considered together, Allorhogas sp.
Table 1.
and Mompha sp. emerged as the strongest candidates
among herbivores attacking reproductive structures of
M. calvescens (Table 1). Other insects showing relatively high overall potential for biological control were
the defoliator A. petroa and the inflorescence feeder
Apion sp. Less likely to become effective biological
control agents of M. calvescens were A. leucocyma, A.
lotanalis, E. opisena, T. paron and P. polibetes because
of their high probability of experiencing biotic interference in Hawaii and/or the possibility of low host specificity.
Discussion
Our assessment of the potential of the insects studied
as biological control agents of M. calvescens may be
preliminary but is still helpful as an initial evaluation.
Other insects being evaluated as biocontrol agents of
M. calvescens that have not been included here are: the
fruit-feeding weevil Anthonomus monostigma Champion (Coleoptera: Curculionidae) (Chacn, 2007), the
stem/leaf-feeding weevil Cryptorhynchus melastomae
Champion (Coleoptera: Curculionidae) (Reichert, 2007),
the sap-sucking Diclidophlebia spp. (Hemiptera: Psyllidae) (Morais, 2007; Burckhardt et al., 2005), and
the defoliating caterpillar Euselasia chrysippe Bates
(Lepidoptera: Riodinidae) (Allen, 2007). As additional
information becomes available, insects will need to be
re-evaluated. Other practical factors that affect prioritization are the viability of insect rearing and the difficulty
to obtain permits to export insects from native areas.
Acknowledgements
We thank Drs Robert Barreto and Marcelo Picano as
well as people working in their laboratory groups at
the Universidade Federal de Viosa. We also thank Dr
ine insect species selected as potential biological control agents of Miconia calvescens based on impact, host speciN
ficity and probability of biotic interference in Hawaii. For each insect, each criterion was assigned a score from 1 to 3
indicated in parenthesis: 1 = low, 2 = moderate or 3 = high. Risks of biotic interference were assigned negative scores.
Finally, an overall assessment of each potential biocontrol agent was made by summing scores across criteria.
Insect species
Plant part attacked
Impact on plant Host specificity
Probability of biotic
Overall potential for
interference in Hawaii biological control
Allorhogas sp.
Mompha sp.
Apion sp.
Erora opisena
Temecla paron
Parrhasius
polibetes
Atomacera petroa
Antiblemma
leucocyma
Ategumia lotanalis
Inflorescences
Inflorescences
Inflorescences
Inflorescences
Inflorescences
Inflorescences
High (3)
High (3)
High (3)
High (3)
High (3)
High (3)
High (3)
High (3)
Unknowna (3)
Unknowna (2)
Unknowna (2)
Low (1)
Lowb (1)
Lowb,c (1)
Moderateb,c (2)
Moderateb (2)
Moderateb (2)
Moderateb (2)
High (5)
High (5)
Moderate-High (4)
Low-Moderate (3)
Low-Moderate (3)
Low (2)
Leaves
Leaves
Moderate (2)
High (3)
High (3)
High (3)
Lowa (1)
Highb,c (3)
Moderate-High (4)
Low-Moderate (3)
Leaves
High (3)
Moderatea (2)
Highb,c (3)
Low (2)
Only observed on M. calvescens and based on host specificity of related insects.
b
Based on field and/or laboratory observations.
c
Based on published studies with same and/or related species.
a
131
XII International Symposium on Biological Control of Weeds
Paul Hanson and Kenji Nishida at the Universidad de
Costa Rica, Dr Robert Robbins at the Smithsonian Institution, Dr Ichiro Nakamura at the University of Buffalo and Dr Isidro Chacn at the Instituto Nacional de
Biodiversidad of Costa Rica for their help with insect
identification. Funding was provided by the Hawaii
Invasive Species Council and Forest Service International Programs.
References
Allen, P. (2007) Demografa, patrn de supervivencia y efectos de tamao de grupo en larvas gregarias de Euselasia
chrysippe (Lepidoptera: Riodinidae), un potencial agente
de control biolgico de Miconia calvescens (Melastomataceae) en Hawai. MSc Thesis. Escuela de Biologia, Universidad de Costa Rica, San Jose, Costa Rica, 51 pp.
Badenes-Perez, F.R. and Johnson, M.T. (2007a) Ecology,
host specificity and impact of Atomacera petroa Smith
(Hymenoptera: Argidae) on Miconia calvescens DC (Melastomataceae). Biological Control 43, 95101.
Badenes-Perez, F.R. and Johnson, M.T. (2007b) Ecology and
impact of Allorhogas sp. (Hymenoptera: Braconidae) and
Apion sp. (Coleoptera: Curculionoidea) on fruits of Miconia calvescens DC (Melastomataceae) in Brazil. Biological Control 43, 317322.
Badenes-Perez, F.R. and Johnson, M.T. (2008) Biology, herbivory, and host specificity of Antiblemma leucocyma
(Lepidoptera: Noctuidae) on Miconia calvescens (Melastomatacaea) in Brazil. Biocontrol Science and Technology
18, 183192.
Barreto, R.W., Seixas, C.D.S., Killgore, E. and Gardner,
D.E. (2005) Mycobiota of Miconia calvescens and related
species from the neotropics, with particular reference to
potential biocontrol agents. Technical Report 132 pp. 24.
Pacific Cooperative Studies Unit, University of Hawaii at
Manoa, Honolulu, HI.
Burckhardt, D., Hanson, P. and Madrigal, L. (2005) Diclidophlebia lucens, n. sp. (Hemiptera: Psyllidae) from Costa Rica, a potential control agent of Miconia calvescens
(Melastomataceae) in Hawaii. Proceedings of the Entomological Society of Washington 107, 741749.
Burkhart, R.M. (1995) Natural enemies of Miconia calvescens. Hawaii Department of Agriculture, Honolulu, HI.
Chacn, M.E.J. (2007) Historia natural de Anthonomus monostigma (Coleoptera: Curcuilionidae) y su potencial
como agente de control biolgico de Miconia calvescens
(Melastomataceae). MSc Thesis. Escuela de Biologia,
Universidad de Costa Rica, San Jose, Costa Rica, 85 pp.
Conant, P. (2002) Classical biological control of Clidemia
hirta (Melastomatacea) in Hawaiiusing multiple strategies. In: Smith, C.W., Denslow, J.S. and Hight, S. (eds)
Workshop on biological control of invasive plants in native Hawaiian ecosystems. Technical Report 129. Pacific
Cooperative Studies Unit, University of Hawaii at Manoa,
Honolulu, HI, pp. 1320.
DAraujo e Silva, A.G., Gonalves, C.R., Galvao, D.M.,
Gonalves, J.L., Gomes, J., do Nascimento Silva, M. and
de Simoni, L. (1968) Quarto Catalogo dos Insetos que
Vivem nas Plantas do Brasil, seus Parasitos e Predadores
(ed. M. d. Agricultura), Rio de Janeiro, Brasil, 622 pp.
Dennill, G.B., Donnelly, D., Stewart, K. and Impson, F.A.C.
(1999) Insect agents used for the biological control of the
Australian Acacia species and Paraserianthes lophantha
(Willd.) Nielsen (Fabaceae) in South Africa. African Entomological Memoir 1, 4554.
Diniz, I.R. and Morais, H.C. (2002) Local pattern of host
plant utilization by lepidopteran larvae in the cerrado veg
etation. Entomotropica 17, 115119.
Downey, J.C. (1962) Hostplant relations as data for butterfly
classification. Systematic Zoology 11, 150159.
Hoffmann, J.H., Impson, F.A. C., Moran, V.C. and Donnelly,
D. (2002) Biological control of invasive golden wattle trees (Acacia pycnantha) by a gall wasp, Trichilogaster sp.
(Hymenoptera: Pteromalidae), in South Africa. Biological
Control 25, 6473.
Julien, M.H. and Griffiths, M.W. (1998) Biological control of
weeds: a world catalogue of agents and their target weeds.
CAB International, Wallingford, UK, 223 pp.
Kaiser, B.A. (2006) Economic impacts of non-indigenous
species: Miconia and the Hawaiian economy. Euphytica
148, 135150.
McClay, A. and De Clerck-Floate, R. (1999) Establishment
and early effects of Omphalapion hookeri (Kirby) (Coleoptera: Curculionidae) as a biological control agent for
scentless chamomile, Matricaria perforata Merat (Asteraceae). Biological Control 14, 8595.
Medeiros, A.C., Loope, L.L., Conant, P. and McElvaney, S.
(1997) Status, ecology and management of the invasive
plant Miconia calvescens DC (Melastomataceae) in the
Hawaiian Islands. Bishop Museum Occasional Papers 48,
2336.
Meyer, J.-Y. (1998) Observations on the reproductive biology of Miconia calvescens DC (Melastomataceae), an alien
invasive tree on the island of Tahiti (South Pacific Ocean).
Biotropica 30, 609624.
Meyer, J.-Y. and Florence, J. (1996) Tahitis native flora endangered by the invasion of Miconia calvescens DC (Melastomataceae). Journal of Biogeography 23, 775781.
Morais, E.G.F. (2007) Diclidophlebia smithi (Hemiptera:
Psyllidae) como agente de controle biolgico da planta
invasora Miconia calvescens. MSc Thesis. Universidade
Federal de Viosa, Viosa, Brazil, 68 pp.
Norambuena, H. and Piper, G.L. (2000) Impact of Apion ulicis Forster on Ulex europaeus L. seed dispersal. Biological Control 17, 267271.
Picano, M.C., Barreto, R.W., Fidelis, E.G., Semeao, A.A.,
Rosado, J.F., Moreno, S.C., de Barros, E.C., Silva, G.A.
and Johnson, M.T. (2005) Biological control of Miconia
calvescens by phytophagous arthropods. Technical Report
134 pp. 31. Pacific Cooperative Studies Unit, University
of Hawaii at Manoa, Honolulu, HI.
Regent Instruments Inc. (2003) WinFOLIA 2003d, Quebec,
Canada, 70 pp.
Reichert, E. (2007) Cryptorhynchus melastomae (Coleoptera: Curculionidae) as a potential biocontrol agent for
Miconia calvescens (Melastomataceae) en Hawaii. MSc
Thesis. Department of Natural Resource Sciences, McGill
University, Montreal, Canada, 66 pp.
Robbins, R.K. and Aiello, A. (1982) Foodplant and oviposition records for Panamenian Lycaenidae and Riodinidae.
Journal of the Lepidopterists Society 36, 6575.
Zikan, J.F. (1956) Beitrag zur biologie von 12 Theclinen-arten.
Dusenia 7, 139148.
132
Giving dyers woad the blues: encouraging
first results for biological control
G. Cortat,1 H.L. Hinz,1 E. Gerber,1 M. Cristofaro,2,3
C. Tronci,3 B.A. Korotyaev4 and L. Gltekin5
Summary
Dyers woad, Isatis tinctoria L. (Brassicaceae), has been cultivated since Roman times throughout
Europe for the blue indigo dye extracted from its leaves and was introduced by early colonists into
North America. Today, it is a declared noxious weed in ten western US states. A literature survey for
insects, mites and pathogens associated with dyers woad revealed several biological control candidates. Three were found in 2004 during preliminary field surveys in Switzerland and Germany:
Ceutorhynchus rusticus Gyllenhal and Aulacobaris fallax H. Brisout, both root-mining weevils, and
Psylliodes isatidis Heikertinger, a shoot-mining flea beetle. Results of host-specificity tests conducted
at CABI Europe-Switzerland are particularly promising for C. rusticus, a very damaging species able
to kill overwintering rosettes. Results of additional host-specificity tests with P. isatidis are necessary
to decide whether it is worth continuing with this species, while A. fallax is not specialized enough to
be further considered. In 2006, additional field surveys were conducted in Turkey, Bulgaria, Romania
and Kazakhstan to find new candidates. Based on the material identified thus far, two species are of interest, a flea beetle preliminarily identified as Psylliodes sophiae var. tricolor Weise and a root-mining
weevil preliminarily identified as Aulacobaris near fallax. For both species, rearing colonies were established in Switzerland, and methods for host-specificity tests were developed. A literature survey revealed 62 species to be associated with dyers woad in Europe. Of the ten species only described from
dyers woad (I. tinctoria) or closely related Isatis species, four are of particular interest, viz. the rootmining weevil Aulacobaris licens Reitter, an as-yet undescribed Lixus sp. and the two seed-feeding
weevils Bruchela exigua Motschulsky and Ceutorhynchus peyerimhoffi Hustache. Surveys will be
conducted in 2007 to find at least two of these four species, and investigations on already available
agents will continue.
Keywords: Isatis tinctoria, weed biological control, host range testing.
Introduction
Dyers woad, Isatis tinctoria L. (Brassicaceae), is a
winter annual, biennial or short-lived perennial mustard of Eurasian origin. The original distribution of
dyers woad in Europe and Eurasia is difficult to reconstruct because it has been widely distributed and
grown by humans as a dye crop since Roman times.
It was introduced to North America by early colonists
CABI Europe-Switzerland, Rue des Grillons 1, 2800 Delmont,
Switzerland.
2
ENEA-Casaccia, Rome, Italy.
3
BBCA, Via del Bosco 10, 00060 Sacrofano (Rome), Italy.
4
Russian Academy of Sciences, Zoological Institute, 199034 St. Petersburg, Russia.
5
Atatrk University, Department of Plant Protection, 25240 Erzurum,
Turkey.
Corresponding author: G. Cortat <[email protected]>.
CAB International 2008
1
as a textile dye crop and then accidentally spread as a
contaminant of crop seed (Hegi, 1986; McConnell et
al., 1999). Today, dyers woad is a declared noxious
weed in ten western US states (USDA-NRCS Plants
National Database, http://plants.usda.gov; Invaders
Database, http://invader.dbs.umt.edu). It was estimated
that dyers woad reduced crop and rangeland production in Utah by $2 million in 1981, and its infestation
doubled there within 10 years (Evans and Chase, 1981,
in McConnell et al., 1999). Unlike many other mustard
weeds, dyers woad does not depend on disturbance to
establish and can readily invade and dominate wellvegetated rangeland sites.
In spring 2004, an initiative was started by a consortium of southeastern Idaho and Utah counties, administered through the Hoary Cress Consortium (Dr Mark
Schwarzlaender, University of Idaho) to investigate
the potential for classical biological control of dyers
woad. During preliminary surveys conducted by CABI
133
XII International Symposium on Biological Control of Weeds
Europe-Switzerland in 2004, three species were prioritized as biological control candidates, and investigations started.
Here we present data on the biology and host-specificity testing of the three species under study, as well as
a list of additional candidates resulting from extensive
literature and field surveys conducted in 2006.
Biological control candidates
studied since 2004
One flea beetle and two weevil species were found
during preliminary surveys in southern Germany and
southern Switzerland in 2004, and host-specificity tests
were started in 2005. The current test plant list for dyers woad comprises about 100 plant species and varieties, nearly two thirds of which are native to North
America (Wilson and Littlefield, 2006).
Aulacobaris fallax H. Brisout
(Coleoptera: Curculionidae)
Females of this univoltine weevil lay eggs from the
end of March until early summer into the petioles of
dyers woad or directly into the root. Larvae mine in
the petioles, root crown and root of dyers woad, in
which they also pupate. Heavily infested plants are visibly stunted and may die (Hinz et al., 2004).
In no-choice oviposition and development tests conducted with 39 test plant species, mining and/or larvae
were detected in 16 plant species, 12 of which are indigenous to North America. Subsequently, these plants
were exposed in multiple-choice field cage tests. Apart
from dyers woad, mines or larvae were found in eight
test plant species, all native to North America (Hinz et
al., 2005, 2006). Although in five cases only, one or
two of the plants offered were attacked, results indicate that the range of plants A. fallax females accept
for oviposition and on which larvae are able to develop
is broader than expected and includes several North
American species. In conclusion, A. fallax is not regarded as specific enough for consideration as a biological
control agent for dyers woad.
Ceutorhynchus rusticus Gyllenhal
(Coleoptera: Curculionidae)
Females of this weevil lay eggs from September
throughout winter until early spring. Eggs are mostly
laid directly into the leaf surface under the epidermis.
Larvae feed in the petioles and later in the root crown.
Mature larvae leave the plants to pupate in the soil, and
adults of the next generation start to emerge from the
end of May onwards. Newly emerged weevils aestivate during July and August and resume feeding at the
beginning of September (Hinz et al., 2004). The main
shoot of heavily attacked plants often dies, which easily distinguishes attacked plants in the field.
During no-choice oviposition and development
tests, 45 test species were exposed, including 24 native
to North America. Of the 43 species for which results
(adult emergence) are available, adults thus far have
only emerged from three test species, viz. Arabis holboellii Hornemann, Schoenocrambe linifolia Nutt., and
Stanleya viridiflora Nutt. However, while an average
of 8.1 1.5 adults emerged from dyers woad (control)
plants, only an average of 0.5 0.5 to 1.5 0.6 adults
emerged from these non-target test plants. In subsequent choice tests with two of the species using five
replicates each, no attack occurred on A. holboellii, and
S. linifolia sustained attack at a lesser degree than dyers woad (Hinz et al., 2005, 2006). Hoffmann (1954)
and Freude et al. (1983) each listed I. tinctoria as the
only host plant for C. rusticus. Although C. rusticus appears to have been collected previously from Rorippa
palustris (L.) Besser (B. Korotyaev, unpublished data),
we did not find attack on three other Rorippa species (Rorippa amphibia L., Rorippa sinuata Nutt. and
Rorippa sylvestris (L.) Bess.) that we tested (Hinz et
al., 2006).
In conclusion, C. rusticus is regarded as a very promising biological control candidate for dyers woad. We
will continue host-specificity tests in 2007 and are also
planning to conduct an experiment to quantify the impact of C. rusticus on dyers woad.
Psylliodes isatidis Heikertinger
(Coleoptera: Chrysomelidae)
In autumn, females of this flea beetle lay eggs in the
soil close to the root crown of the host plant. Larvae
hatch in early spring and feed in the developing shoots
of dyers woad. Mature larvae leave the plants to pupate in the soil, and adults of the subsequent generation
emerge during June. After a brief feeding period, the
beetles aestivate. This species shows good potential for
rapid population increase and at high densities can kill
shoots or whole plants (Hinz et al., 2004).
During no-choice larval transfer and development
tests with nine test species, adult flea beetles emerged
from four. However, a lower proportion of larvae successfully developed into adults on test plants (312%)
compared to control (dyers woad) plants (28%). In a
subsequent multiple-choice field-cage test exposing
14 test species together with dyers woad, adult beetles emerged from nine test species. However, while
an average of 46 adults (maximum 130) emerged from
individual dyers woad plants, only 0.3 to 3.6 emerged
from test species (Fig. 1). We assume that the density
of adults released into the field cage was artificially
high (considering the high number of adults that emerged from controls), and plants may have been placed
too close to each other (Hinz et al., 2005). A second
multiple-choice field cage test will be established
in autumn 2006 releasing fewer beetles and exposing fewer test plants (Hinz et al., 2006). Depending
134
Giving dyers woad the blues: encouraging first results for biological control
Figure 1.
50
40
30
20
10
0
Isatis tinctoria
Armoracia rusticana
Brassica oleracea sabauda
Erysimum asperum
Erysimum inconspiccum
Lepidium campestre
Lepidium latifolium
Lesquerella ludoviciana
Peltaria alliacea
Physaria chambersii
Reseda lutea
Rorippa amphibia
Rorippa sinuata
Rorippa sylvestris
Schoenocrambe linifolia
Mean number of adults
emerged
60
Mean number of Psylliodes isatidis adults emerged per plant
during a multiple-choice field-cage test established at CABI
Europe-Switzerland in September 2005.
on results obtained in summer 2007, we will make a
decision whether it is worth continuing work with P.
isatidis.
New biological control candidates
Emphasis in 2006 was placed on field surveys in parts
of the likely area of origin of dyers woad with the aim
to find additional candidates as well as to collect plant
samples to investigate the genetic variability of the genus Isatis in Eurasia. Five field trips were conducted,
three to central and northeastern Turkey, one to Romania and Bulgaria and one to Kazakhstan. Twenty-two
field sites of dyers woad were found. At each site,
adult insects were sampled, plants were dissected, and
plant parts (leaves, shoots, seeds, and roots) with immature insect stages or signs of mite or pathogen attack were collected and taken back to the lab. Immature stages were reared to adulthood and sent, together
with field-collected adults, to taxonomists for identification. Combining data of the last 3 years, we have
reared or sampled 52 insect species, 33 of which are
known to feed and/or develop on plants of the family
Brassicaceae. Half of the species (n = 17) are weevils
(Curculionoidea), and most of these belong to the genus Ceutorhynchus Germar. Based on identifications
available thus far, two species could be of interest,
viz. a stem-mining flea beetle, preliminarily identified
as Psylliodes sophiae var. tricolor Weise, and a root-
mining weevil, Aulacobaris sp. near fallax (H. Brisout),
which is similar to A. fallax but most likely a distinct
subspecies or a different species.
Apart from field surveys, we also continued to collect information on insects, mites and fungi associated
with dyers woad in the literature as well as by contacting taxonomists and collaborators. Combined with
previous data, 56 insect species were found plus two
nematodes and four fungi associated with dyers woad.
About one third were weevils (Curculionidae and Urodontidae), followed by flea beetles (Chrysomelidae,
Alticini; n = 10). Ten species are only described from
I. tinctoria or closely related Isatis species. Of these,
four are of particular interest, viz. a root-mining weevil,
Aulacobaris licens Reitter, an as-yet undescribed Lixus
sp. and two seed-feeding weevils, Bruchela exigua
Motschulsky (=Urodon exiguus), and Ceutorhynchus
peyerimhoffi Hustache.
In the following, short descriptions of these new potential biological control candidates are given.
P. sophiae var. tricolor Weise
(Coleoptera: Chrysomelidae)
This flea beetle was found in Bulgaria, central
Turkey and Kazakhstan and preliminarily identified as
P. sophiae var. tricolor. Larvae mined in the root crown
of dyers woad. At several sites, large numbers of this
species were observed, and larval mining can be quite damaging, causing stunting of shoots. Therefore, although
P. sophiae var. tricolor is described as oligophagous
135
XII International Symposium on Biological Control of Weeds
and should also occur on another brassica species (e.g.
Sisymbrium sophiae L.) (Freude et al., 1966), we decided to prioritize it as a candidate. Beetles collected
in the field were overwintered in an incubator at 3C
between 8 November 2006 and 26 February 2007.
Females started to lay eggs in March, a colony was
established,and methods for host-specificity tests were
developed. The main emphasis will be to investigate
whether a host race exists that is specific to dyers woad.
Aulacobaris sp. pr. fallax (H. Brisout)
(Coleoptera: Curculionidae)
This species was found in central Turkey. Similar to
A. fallax that was collected in southern Germany (see
above), the larvae mine in root crowns and roots of
dyers woad, in which they also pupate. However, this
populations oviposition behaviour is slightly different, and slight morphological differences were found
(B. Korotyaev, unpublished data). In addition, A. fallax
has never been recorded from Turkey (B.A. Korotyaev and L. Gltekin, unpublished data). Specimens
from Germany and Turkey might therefore belong to
two different species or subspecies. We are currently
establishing a colony and plan to conduct preliminary
host-specificity tests with plant species that supported
development of A. fallax in earlier tests. In addition,
the morphology of specimens sampled in Turkey and
Germany will be studied in detail.
B. exigua (Motschulsky)
(Coleoptera: Anthribidae)
B. exigua (= U. exiguus) is a seed-feeding weevil. It
occurs in the southern Ukraine and the northern Caucasus region and may be monophagous on dyers woad
(Korotyaev, 1988). We are planning a field trip to these
regions in the second half of May to collect adults, start
a colony in quarantine at CABI and begin preliminary
host-specificity tests.
C. peyerimhoffi Hustache
(Coleoptera: Curculionidae)
C. peyerimhoffi is known from Algeria, Italy and
Greece. Specimens found in southeastern Turkey are
most probably Ceutorhynchus isatidis Colonnelli (E.
Colonnelli, B. Korotyaev, L. Gltekin, personal communication). The larvae of this species develop in
the seeds of dyers woad and thus far have only been
collected from I. tinctoria in Europe (E. Colonnelli,
personal communication). However, the species was
originally described based on specimens collected in
Algeria on the closely related Isatis djurdjurae Coss.
(Hustache, 1916; Peyerimhoff, 1919). A field trip is
planned in 2007 to collect this species.
We are not currently working with a seed feeder as
candidate, and dyers woad spreads exclusively by seed;
therefore, B. exigua and C. peyerimhoffi are promising
choices as new candidates.
A. licens (Reitter)
(Coleoptera: Curculionidae)
Conclusions and outlook
The root-mining weevil A. licens occurs principally
on Isatis glauca, species from northeastern Turkey
closely related to I. tinctoria. During a field trip at the
beginning of April, some specimens were collected and
brought back to the quarantine facility. Weevils readily fed and laid eggs on cut dyers woad (I. tinctoria)
shoots offered in cylinders, indicating that this plant is
also part of the host range of A. licens. Provided that A.
licens will also successfully complete larval development on dyers woad, a colony will be established, and
host-specificity tests will start in 2008.
Lixus (Compsolixus) sp.
(Coleoptera: Curculionidae)
Several specimens of this probably undescribed
species have been collected only on I. glauca in northeastern Turkey, while a few more have been found in
European collections, mainly from Turkey. Although
most species of Lixus Fabricius are not monophagous,
a seemingly narrow distribution of this species and its
rarity in the collections indicate the possibility of a restricted host range, and it warrants a closer examination
as a candidate for dyers woad. The larvae are believed
to be stem and/or root miners.
Since the start of the project in 2004, three candidates
have been investigated, two weevils and one flea beetle. The root-miner C. rusticus shows very good potential as a biological control agent for dyers woad,
and host-specificity tests are well advanced. While the
second weevil, A. fallax, turned out to be not specific
enough, additional multiple-choice tests with the flea
beetle P. isatidis will reveal whether it can be further
considered as candidate.
As a result of additional field and literature surveys
conducted in 2006, six additional species, five weevils
and one flea beetle, were prioritized as new candidates.
Three of the agents have already been collected, while
the other three species will be collected during planned
field trips later in 2007 or in 2008.
In addition to surveys for natural enemies, plant
material was collected for morphological and molecular analyses. In Eurasia, the genus Isatis L. is represented by roughly 50 species (Hegi, 1986), and various
subspecies of I. tinctoria are recognized. Observations
during field surveys confirm this information, with
considerable variability in leaf shape and colour, seed
shape and phenology of dyers woad seen. A thorough
revision of the genus would therefore urgently be
needed, especially since this will directly influence sur-
136
Giving dyers woad the blues: encouraging first results for biological control
vey areas and agent selection. Plant material collected
was sent to Dr John Gaskin (USDA, ARS, NPARL,
Sidney, MT), who has kindly offered to process the
material.
In conclusion, prospects for the biological control of
dyers woad look very promising. After only 3 years,
host-specificity tests with three insect species are well
advanced, and we have identified at least six additional
species as candidates. Geographic areas (Turkey and
Caucasus) with a wide variety of apparently specialized feeders on Isatis spp. were identified and will be
further surveyed for biological control candidates in
the future.
Acknowledgements
We thank Bethany Muffley and Cristobal Tostado for
additional technical assistance and Florence Willemin
and Christian Leschenne (all CABI EuropeSwitzerland) for plant propagation. We would also like to thank
the following taxonomists for species identifications:
Prof Paolo Audisio (Universit degli Studi di Roma
La Sapienza, Roma, Italy; Nitidulidae), Dr Alexander J. Konstantinov (Systematic Entomology Laboratory, USDA, Washington, DC; Alticinae), Dr Serguei
Yu. Sinev (Russian Academy of Sciences, Zoological
Institute, St. Petersburg, Russia; Pyraustidae) and Dr
Michael von Tschirnhaus (University of Bielefeld, Bielefeld, Germany; Agromyzidae). We also like to thank
Dr Alecu Diaconu (Institute of Biological Research,
Iasi, Romania) and Ion Schiopu for facilitating our field
trip in Romania and helping us to find Isatis sites. We
specifically thank Dr Roman V. Jashenko (Institute of
Zoology, Almaty, Kazakhstan) and Dr Anna Ivashenko
(Institute of Botany, Almaty, Kazakhstan) for facilitating our field trip to Kazakhstan and providing access
to the herbarium collection in Almaty. Financial support for this project was kindly provided by the Idaho
State Department of Agriculture, the US Forest Service
State and Private Forest, as well as various counties in
Idaho, Utah and Wyoming through the Panhandle Lakes
RC&D Hoary Cress Consortium. This program would
not be possible without the continuing dedication and
initiative of Jim Hull (Weed Superintendent, Franklin
County, ID) and Mark Schwarzlaender (University of
Idaho). The study by L. Gltekin and B. Korotyaev
was supported by the Collaborative Linkage Grant
no. 981318 of the NATO Life Science and Technology
Programme. The work of B. Korotyaev was supported
also by the Russian Foundation for Basic Research,
Grant nos. 04-04-81026-Bel2004a and 07-04-00482a,
and was performed with the use of the collection of
the Zoological Institute, Russian Academy of Sciences
(UFC ZIN no. 2-2.20), contract no. 02.452.11.7031
with Rosnauka (2006-RI-26.0/001/070).
References
Colonnelli, E. (2004) Catalogue of Ceutorhynchinae of the
world with a key to genera. Argania, Barcelona, 124 p.
Freude, H., Harde, K.W. and Lohse, G.A. (1966) Die Kfer
Mitteleuropas. Band 9: Cerambydidae, Chrysomelidae.
Goecke & Evers Verlag, Krefeld, Germany, 299 p.
Freude, H., Harde, K.W. and Lohse, G.A. (1983) Die Kfer
Mitteleuropas. Band 11: Ryncophora. Goecke & Evers
Verlag, Krefeld, Germany, 342 p.
Hegi, G. (1986) Illustrierte Flora von Mitteleuropa. Spermatophyta, Band IV Teil 1. Angiospermae, Dicotyledones 2.
Paul Parey, Berlin, Hamburg, Germany pp. 126131.
Hinz, H.L., Cortat, G. and Gerber, E. (2004) Biological control of dyers woad, Isatis tinctoria. Annual Report 2006.
Unpublished Report, CABI Bioscience Switzerland Centre, Delmont, Switzerland.
Hinz, H.L., Cortat, G. and Gerber, E. (2006) Biological control of dyers woad, Isatis tinctoria. Annual Report 2005.
Unpublished Report, CABI Bioscience Switzerland Centre, Delmont, Switzerland.
Hinz, H.L., Gerber, E., Krebs, C. and Cortat, G. (2005) Biological control of dyers woad, Isatis tinctoria. Annual
Report 2004. Unpublished Report, CABI Bioscience
Switzerland Centre, Delmont, Switzerland.
Hoffmann, A. (1954) Faune de France 59: Coleoptres, Curculionides. Editions Paul Lechevalier, Paris, France, pp. 487
1207.
Hustache, A. (1916) Deux nouveaux Ceuthorrhynchini (Col.
Curculionidae). Bulletin de la Socit entomologique de
France 21, 6970.
Korotyaev, B.A. (1988) Contribution to the knowledge of
the superfamily Curculionoidea (Coleoptera) of the fauna
of the USSR and adjacent countries. In Proceedings of
the Zoological Institute of the Academy of Sciences of the
USSR 170, 122163. (in Russian).
Korotyaev, B.A. (1992) On the trophic specialization of Palaearctic weevils of the subfamily Ceutorhynchinae (Coleoptera, Curculionidae). In Proceedings of the 4th ECE/
13th SIEEC Gdll, Budapest 1991, 510512.
McConnell, E.G., Evans, J.O. and Dewey, S.A. (1999) Dyers woad. In: Sheley, L. and Petroff, K. (eds) Biology and
Management of Noxious Rangeland Weeds. OSU Press,
Corvalis, OR, pp. 231237.
Peyerimhoff, P. (1919) Note sur la biologie de quelques Coloptres Phytophages du Nord-Africain (3e srie). Annales
de la Socit entomologique de France 88, 169258.
Wilson, L. and Littlefield, J. (2006) Proposed host specificity
test plant list for potential biological control agents of
hoary cresses, Lepidium draba L. appelianum, perennial
pepperweed, L. latifolium, and dyers woad, Isatis tinctoria (Brassicaceae). Submitted by J. Littlefield to USDA
APHIS Technical Advisory Group (TAG) in June 2006.
TAG No. 06-04.
137
Herbivores associated with
Arundo donax in California
T.L. Dudley,1,* A.M. Lambert,1, 2,* A. Kirk3 and Y. Tamagawa1
Summary
The Old World grass, Arundo donax L. (giant reed), is a serious invader of California riparian areas,
and its purported ecosystem impacts led to its consideration as a target for biological control development. However, the herbivore complex in the Arundo adventive range has not been characterized, so
there is little information regarding insects that may hinder biological control efforts by interfering
with the release of new agents or that could be promoted as augmentative biological control agents if
they have a substantial impact on the target weed. Here we report the results of surveys in California,
with emphasis on three presumably non-indigenous insects that inflict significant damage to the host.
One is a shoot-boring wasp, Tetramesa romana (Walker) (Hymenoptera: Eurytomidae), with a range
limited to southern California and that damages shoots generally less than 1 cm in diameter. A shoot
fly, Cryptonevra sp., is also associated with shoot damage and often mortality of secondary stems. A
third herbivore, the aphid Melanaphis donacis (Passerini), is widespread in the southern and central
parts of the state but has less apparent impact to the host. T. romana and Cryptonevra sp. are currently candidates for biological control development and introduction from overseas locations. Their
established presence in California suggests that efforts could be revised to focus on documentation of
host ranges and impacts under field rather than in quarantine conditions, in anticipation of future re
distribution in North America.
Keywords: biological control, Cryptonevra, Melanaphis donacis, Tetramesa romana,
giant reed; herbivore.
Introduction
Arundo donax L. (giant reed) may be the most destructive invader of California riparian areas, displacing
native vegetation, transpiring excessive groundwater,
posing erosion and wildfire risks and providing poor
wildlife habitat (Bell, 1994; Dudley, 2000; Kisner,
2004). Few arthropods appear to be associated with
Arundo in California (Herrera and Dudley, 2003; Kirk
et al., 2003), and most are using it as opportunistic
structural habitat rather than as a food source. A variety
of herbivores is found in the region of origin (from the
Mediterranean Basin and across southern Asia), and
the level of herbivore impacts is considered to be much
Marine Science Institute, University of California, Santa Barbara, CA
93106, USA.
2
Department of Biology, Eastern Connecticut State University, Willimantic, CT, USA.
3
USDA-ARS European Biological Control Laboratory, Montpellier,
France.
Corresponding author: T.L. Dudley <[email protected]>.
* These authors contributed equally to this work.
CAB International 2008
1
greater than in California or other areas where Arundo
is invasive (Kirk et al., 2003). Classical biological control is being developed using several candidate agents
from Eurasia (Kirk and Widmer, 2004), but implementation is not anticipated for at least several years.
A standard element of a biological control programme should involve documentation of herbivores
attacking the target weed in its adventive range to determine if new agents are needed or if effects of existing herbivores can be enhanced, as well as to evaluate
the potential for interference with introduction of new
agents (Harris, 1975; Olckers and Hulley, 1995). As
part of the programme to build the ecological framework for justifying Arundo biological control, we are
characterizing and comparing Arundo herbivores and
associated plant condition in California and Southern
Europe. Early investigations indicated that several nonindigenous insects, including a shoot-boring wasp of
the family Eurytomidae, are present in California. The
wasp was subsequently identified as Tetramesa romana
(Walker), a widespread Mediterranean species and
a primary candidate in the Arundo biological control
programme. Members of this genus are highly host
138
Herbivores associated with Arundo donax in California
specific (Al-Barrak et al., 2004), which makes them
particularly suitable for biological control.
Our initial objective is to determine the geographic
extent of this and other current Arundo herbivores in
California and to quantify their efficacy against this
host weed. The larger objective is to increase the impact
of existing herbivores through augmentative measures
and/or to distribute them more widely to provide more
extensive host suppression. We are currently evaluating their host specificity and appropriateness for mass
rearing and redistribution.
Materials and methods
To examine Arundo herbivore distributions in California,
we conducted monthly insect surveys of Arundo stands
on the Santa Clara River and less frequent (once or
twice during study period) extensive surveys of Arundo-
infested areas throughout the southwestern US (Table
1). Line transects 100 m long were established within
Arundo vegetation, and samples were collected from
0.5 m2 quadrats placed at 10-m intervals along transect
lines. As of April 2007, a total of 994 Arundo shoots
within all plots surveyed were cut at the soil surface
and bundled together for transport back to the laboratory, where they were stored at 8C until processed.
Primary or main shoots and side shoots or secondary
shoots were examined separately. Shoot lengths and
diameters were recorded, and shoots were visually examined for herbivores and evidence of herbivore damage. These were then split lengthwise and examined for
internal feeders. After dissection, shoots were dried (2
days at 55C) to determine dry weight biomass. All recovered insects were sent to the USDA European Biological Control Laboratory in Montpellier, France for
identification, and voucher specimens were deposited
in the Santa Barbara Museum of Natural History, Santa
Table 1.
Barbara, CA. Plots were also used to determine shoot
density and biomass per unit area for analysis of plant
growth differences between native and introduced
ranges and evaluation of impacts of biological control
agents after release and establishment.
Plant use by T. romana (oviposition, feeding and
pupation sites) and potential impacts of infestation on
Arundo were evaluated. We focused on this species because it is one of the primary agents being tested as a
potential biological control (Kirk and Widmer, 2004).
Shoot length, basal diameter and biomass of shoots and
side shoots infested with T. romana, were compared with
those of uninfested plants. Main shoot and side shoot
data were analyzed separately using Students t test.
Results and discussion
No native insect herbivores were found using Arundo
as a significant food source, in contrast to prior surveys that show numerous arthropods using this plant
for non-consumptive purposes (Herrera and Dudley,
2003). Two non-native insects and another unidentified
(but potentially non-native) insect were recovered from
Arundo shoots during sampling. An aphid, Melanaphis
donacis (Passerini), was found throughout the sampling
range (except very northern California) with greatest
abundance in coastal Arundo populations. Aphids feed
primarily on the apical shoots and less mature, distal
leaves, and although reaching high population densities
in early spring in some locations, only minor damage
was observed on plants in one location. Coccinellids
were abundant on Arundo shoots several weeks after
peak aphid densities and may have been responsible for
the decrease in aphids by late May - aphids were often
not present during sampling in July through December.
We did not quantify aphid densities during surveys for
several reasons; there was a lack of any visible aphid
Sampling locations (river systems) by county, sampling intensity, and insects recorded in California.
County
Alameda
Humboldt
Imperial
Inyo
Kern
Los Angeles
Mendocino
Monterey
Orange
Riverside
San Bernardino
San Diego
San Luis Obispo
Santa Barbara
Santa Clara
Ventura
Yolo
# Sampling sites
1
2
2
2
1
4
2
5
2
3
3
5
2
7
1
8
1
Insects
Melanaphis donacis (Passerini)
None
Melanaphis donacis
None
None
Melanaphis donacis, Tetramesa romana (Walker)
None
Melanaphis donacis
Melanaphis donacis, Tetramesa romana
Melanaphis donacis, Tetramesa romana
Tetramesa romana (one site)
Melanaphis donacis, Tetramesa romana
Melanaphis donacis
Melanaphis donacis, Cryptonevra sp., Tetramesa romana
Melanaphis donacis
Melanaphis donacis, Cryptonevra sp., Tetramesa romana
Melanaphis donacis
139
XII International Symposium on Biological Control of Weeds
feeding effect, aphids were often dislodged from stems
during stem sampling and collection, aphid movement
reduced accurate measurements, and the time and labor-
intensive nature of counting aphids would have substantially reduced our ability to adequately sample Arundo
populations throughout the state.
Dipteran larvae were recovered from over 80% of
Arundo shoots in one area on the floodplain terrace in
the Santa Clara River, Ventura County, CA. We infrequently found similar damage throughout this river system and the nearby Ventura River. Several larvae feed
together in the upper nodes on main shoots of plants,
and feeding damage resulted in atypical witches broom
shoot growth with 25% stem mortality in infested shoots
(Fig. 1). The witches broom shoot growth was evident
in most shoots covering a 2-ha radius. We did not examine these damaged shoots for bacterial disease, which can
cause this type of deformity; however, we assume that
feeding damage on the primary shoot promoted secondary shoot production, as similarly occurs in stems with
wind-damaged primary shoots. The gregarious larvae
were associated with hour glass shoot damage (Fig. 2)
and may be an inquiline species that feeds on microbes
that colonize damage from another fly species (A. Kirk,
personal observations). Chloropid flies (Cryptonevra
Figure 1.
sp.) in the Mediterranean region primarily attack developing canes up to about 75 cm in length and cause similar hour glass damage to what we observed. In addition,
it or another species of Chloropidae attacks side shoots
and/or leading shoots on taller canes and are often followed by several inquiline species - up to 18 inquiline
spp. have been recorded from southern France (A. Kirk,
unpublished data). Damage prevents shoots from elongating during growth and results in stunted shoots and a
witches broom appearance from increased side shoot
production (Fig. 1). We are focusing our sampling efforts during the spring growing period when herbivores
producing the initial shoot damage may be present.
Cryptonevra sp. is a candidate for potential biological control introduction (Tracy and Deloach, 1998), so
further documentation of its distribution and impacts in
North America and verification that the species present here is the same as the Mediterranean taxon being
tested by US Department of Agriculture-Agricultural
Research Service (USDA-ARS) are essential for future
agent selection.
The shoot-boring wasp, T. romana, was collected
from Arundo populations throughout southern California (Fig. 3). Its exit holes and gall-like formations
produced during larval feeding are evident on primary
Arundo donax L. shoot infested with Cryptonevra sp. in the
Santa Clara River, CA (arrow). Note the atypical witches
broom appearance of the infested shoot relative to the other,
uninfested shoots.
140
Herbivores associated with Arundo donax in California
Figure 2.
Arundo donax L. shoot damage caused by Cryptonevra sp.
feeding. The arrow points to the characteristic hourglass
damage.
and secondary shoots. We have observed oviposition
on shoots both in greenhouse cultivation and at field
sites (Fig. 4). Shoot infestation was variable (range: 0%
to 80%) with a mean of 23.1 4.4% (SE) shoots infested across all sites sampled (n = 994). However, in
March 2007, we observed about 2.5 ha of an Arundo
stand near the Santa Clara River (Ventura Co.) with ca.
Figure 3.
99% of side shoots infested and killed. A recent survey
at this same location suggests that infestation levels of
current-year shoots are still high.
Wasp densities are highest on smaller diameter main
shoots (primarily new shoots) and side shoots of larger,
mature shoots (Fig. 5). Wasp densities can exceed 35
individuals on a main shoot and six individuals on a
Distribution of Tetramesa romana (Walker) (Hymenoptera:
Eurytomidae) in California as of April 2007.
141
XII International Symposium on Biological Control of Weeds
Figure 4.
Tetramesa romana (Walker) (Hymenoptera: Eurytomidae)
ovipositing in an Arundo side shoot. Bar represents 2mm.
side shoot, but average 4.9 3.6 individuals per main
shoot and 3.4 1.9 per side shoot. Infested main shoots
were shorter and smaller in diameter than uninfested
shoots (shoot height, t = 3.02, df = 315, p = 0.003; shoot
diameter, t = 3.93, df = 315, p < 0.001), but biomass
was not significantly different between infested and uninfested shoots. Infested side shoots were thinner than
herbivore free shoots (t = 2.0, df = 197, p = 0.05). Side
Figure 5.
shoot height and biomass were not significantly affected by wasp infestation. In France, T. romana infests
shoots over a broader diameter range (A. Kirk, personal observation); Arundo shoots are also, on average,
thinner and shorter than in North American populations
(A. Lambert, unpublished data).
Biomass reduction in infested shoots could not be
inferred from these data for two reasons: (1) extreme
Distribution of emergence holes of Tetramesa romana (Walker) over the stem density range (mm) of Arundo donax L. main and side shoots.
142
Herbivores associated with Arundo donax in California
variability in measures of morphological character
(shoot length, diameter and biomass) even with substantial replication and (2) wasps colonize primarily
new shoots and secondary shoots, which have lower
biomass than mature and main shoots, respectively
(Spencer et al., 2006). Therefore, wasp impact may
be less noticeable on young shoots relative to that of
an herbivore attacking the larger mature shoots. The
frequency of damage but lack of substantial impact of
T. romana to Arundo populations in our surveys does not
necessarily reflect an absence of impact for this species.
It is anticipated that this herbivore, in association with
other agents such as the rhizome-feeding scale insect,
Rhizaspidiotus donacis (Leonardi), may inflict greater
damage to host plants (A. Kirk, personal observations).
Further surveys are necessary to accurately determine the full extent and limits of the T. romana distribution and its potential to infest a greater proportion of
shoots. For example, as our sampling size increases,
the relationship between wasp density and main and
side shoot mortality becomes stronger. We are currently
evaluating above- and belowground biomass relationships of infested and uninfested plants in a common
garden experiment to elucidate the effect of herbivory
on resource allocation. Experiments are also being conducted by the USDA-ARS to determine the impact of
T. romana on water use by Arundo (J. Goolsby, USDA,
Weslaco, TX, personal communication).
Genetic comparisons are underway in collaboration
with D. Kazmer and W. Jones, USDA-ARS, Sidney, MO,
and European Biological Control Laboratory (EBCL),
using newly eclosed females collected throughout California and matching with wasps collected in the Mediterranean region to determine the origin of California
T. romana. It is unclear at this time whether this wasp
is a recent introduction in California and is still in its
establishment phase, in which case its potential (future)
range may be much greater. Alternatively, if T. romana
has been established for a long period, then dispersal
may have achieved the full extent of its potential range,
and the likelihood for redistribution is poorer. Wasps
have been found throughout the watersheds of Ventura
County, whereas stands in other counties, particularly
at the northern and southern extents of its apparent
range, are often colonized more heterogeneously with
uninfected streams in moderately close (e.g. <10 km)
proximity to colonized stands. This may indicate that
the colonization process is, as yet, incomplete. Patchy
distributions could also represent reduced physiological compatibility with climate regimes at the limits of
the T. romana distribution. Such a distribution in the
warm- temperate coastal region of California is not
dissimilar to other organisms introduced for biological
control of other California pests, e.g. parasitoid Aphytis
spp. against the citrus red scale (Rosen and DeBach,
1978, 1979). Furthermore, it is not yet clear whether
there may have been the co-introduction of T. romana
parasitoids (A. Kirk, unpublished data) and subsequent
population regulation, which could also explain the
variation in attack rate observed over T. romanas distribution in California.
The unintentional introduction of T. romana and
other apparent specialist herbivores to North America
provides the opportunity to conduct host-specificity
trials and other experimental studies in the field, which
will provide instructive ecological information less
easily obtained under quarantine conditions. Plant
growth and insect behavior are often constrained or
atypical in highly controlled environments (Blossey et
al., 1994a,b), so results in open-field settings are expected to be more representative of natural responses.
We suggest that an Arundo biological control programme based on development of California T. romana and Cryptonevra sp. as augmentative biological
control agents for local population enhancement and/or
re-distribution in North America should be carried out
in parallel with foreign importation and further quarantine testing. Host-range testing should be continued to
validate host specificity prior to agent movement, but
conducting this work with insects already established
in North America would enhance the ecological validity of the tests and conserve financial resources. This
survey of California Arundo herbivores also validates
the advice of Harris (1975) and others that biological
control be developed as a logical progression, with
evaluation of weed ecology and associated herbivores
in the invasive range before importing new organisms.
Acknowledgements
We thank J. ten Brinke, K. Kennedy, V. Frankel and
T. Lemein for field assistance and R. Sforza and two
anonymous reviewers for editorial assistance. We appreciate the information and study site access provided
by The Nature Conservancy Santa Clara River Project,
particularly E.J. Remson and C. Cory. Financial assistance was provided, in part, by the Santa Clara River
Trustee Council/US-FWS grant no. 81440-5-G021 and
the University of California Integrated Pest Management Program grant no. 2006-34439-17024 (USDACSREES no. 2006-34439-17024).
References
Al-Barrak M., Loxdale, H.D., Brooks, C.P., Dawah, H.A.,
Biron, D.G. and Alsagair, O. (2004) Molecular evidence
using enzyme and RAPD markers for sympatric evolution in British species of Tetramesa (Hymenoptera: Eurytomidae). Biological Journal of the Linnaean Society 83,
509525.
Bell, G. (1994) Biology and growth habits of giant reed
(Arundo donax). In: Jackson, N.E., Frandsen, P., Douthit,
S. (eds) Arundo donax Workshop Proceedings, Ontario,
CA. pp. 16.
Blossey, B., Schroeder, D., Hight, S.D. and Malecki, R.A.
(1994a) Host specificity and environmental impact of the
weevil Hylobius transversovittatus, a biological control
143
XII International Symposium on Biological Control of Weeds
agent of purple loosestrife (Lythrum salicaria). Weed Science 42, 128133.
Blossey, B., Schroeder, D., Hight, S.D. and Malecki, R.A.
(1994b) Host specificity and environmental impact of two
leaf beetles (Galerucella calmariensis and G. pusilla)
for biological control of purple loosestrife (Lythrum salicaria). Weed Science 42, 134140.
Dudley, T.L. (2000) Arundo donax L., In: Bossard, C.C.,
Randall, J.M., Hoshovsky, M.C. (eds) Invasive Plants of
Californias Wildlands. University of California Press,
Berkeley, CA, pp.5358.
Harris, P. (1975) General approach to biological control of
weeds in Canada. Phytoprotection 56, 135141.
Herrera, A. and. Dudley, T.L. (2003) Invertebrate community
reduction in response to Arundo donax invasion at Sonoma
Creek. Biological Invasions 5, 167177.
Kisner, D.A. (2004) The effect of giant reed (Arundo donax)
in riparian areas of Camp Pendleton Marine Corps Base,
California. MSc Thesis, San Diego State University, CA.
Kirk, A., Widmer, T., Campobasso, G., Carruthers, R. and
Dudley, T. (2003) The potential contribution of natural
enemies from Mediterranean Europe to the management
of the invasive weed Arundo donax (Graminae; Arundinae). Proceedings of the California Invasive Plant Council Symposium 7, 6268.
Kirk, A.A. and Widmer, T.L. (2004) Biological control of Giant Reed (Arundo donax) an invasive plant species in the
USA. USDA-ARS Petition to APHIS for Technical Advisory Group. European Biological Control Laboratory,
Montpellier, France.
Olckers T., and Hulley, P.E. (1995) Importance of preintroduction surveys in the biological control of Solanum
weeds in South Africa. Agriculture, Ecosystems and Environment 52, 179185.
Rosen, D. and DeBach, P. (1978) Diaspididae. In: Clausen,
C.P. (ed.) Introduced Parasites and Predators of Arthropod Pests and Weeds: a World Review. United States
Department of Agriculture, Agricultural Handbook 480,
pp. 78129.
Rosen, D. and DeBach P. (1979) Species of Aphytis of the
World. Dr W. Junk, The Hague, The Netherlands. 801 p.
Spencer, D.F., Liow, P., Chan, W., Ksander, G.G. and Getsinger, K.D. (2006) Estimating Arundo donax shoot biomass.
Aquatic Botany 84, 272276.
Tracy J.L. and Deloach. J. (1998) Suitability of classical
biological control of giant reed (Arundo donax) in the
United States. In: Bell, C.E. (ed.) Proceedings of the
Arundo and Saltcedar Workshop. University of California Cooperative Extension Service, Holtville, CA,
pp. 73109.
144
Which species of the thistle biocontrol
agent Trichosirocalus are present in
New Zealand?
R. Groenteman,1,2 D. Kelly,1 S.V. Fowler2 and G.W.Bourdt3
Summary
Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae), nodding thistle crown weevil, was
introduced to New Zealand in 1989 for classical biological control of Carduus nutans L. (nodding
thistle). Later, it was introduced from New Zealand to Australia. In 2002, a revision of the species
concluded that T. horridus was in fact a complex of three species, with distinct host plant genus
preferences: T. horridus, Trichosirocalus mortadelo Alonso-Zarazaga and Snchez-Ruiz, and Trichosirocalus briesei Alonso-Zarazaga and Snchez-Ruiz with preferences for Cirsium, Carduus and
Onopordum thistles, respectively. In the revision, crown weevils on Carduus thistles in Australia were
identified as all T. mortadelo, sourced from New Zealand. This suggests that the original introductions
into New Zealand were wholly or partly T. mortadelo. A survey conducted to confirm which species
of Trichosirocalus are present in New Zealand shows that all adults collected here are T. horridus
regardless of whether the source host was Cirsium vulgare (Savi) Tenore or C. nutans. This presents
two paradoxes: firstly, that T. mortadelo was collected from New Zealand and shipped to Australia but
has not been found in our surveys, and secondly, that T. horridus does not show distinct preference for
Cirsium thistles in New Zealand as reported elsewhere.
Keywords: Trichosirocalus horridus, T. mortadelo, Carduus nutans, Cirsium vulgare,
Onopordum acanthium.
Introduction
The thistle biological control agent Trichosirocalus
horridus (Panzer) was first introduced to New Zealand
in 1984 and was released in the Canterbury plains of
the South Island (Jessep, 1989b). By 1989, established
populations were available for distribution to further
sites in both the North and South Islands, and the wee
vil has been considered established in New Zealand
ever since. The weevils introduced to New Zealand are
originally from Carduus nutans L. from Neuenburg,
Germany, and were established on C. nutans in Canada
School of Biological Sciences, University of Canterbury, Private Bag
4800, Christchurch, New Zealand.
2
Landcare Research, Gerald St, PO Box 40, Lincoln, 7640, New Zealand.
3
AgResearch Ltd., Lincoln, PO Box 4749, Christchurch 8140, New Zealand.
Corresponding author: R. Groenteman <[email protected]>.
CAB International 2008
1
(Agriculture Canada, Regina Station, SK) prior to their
introduction (Jessep, 1989a, 1989b; Julien and Griffiths,
1998; P. Harris, personal communication).
Preliminary trials in New Zealand soon after the
weevils establishment suggested it readily attacked
the three Carduus spp. present in the South Island,
C. nutans, Carduus pycnocephalus L. and Carduus
tenuiflorus Curtis (the fourth Carduus sp., present only
in New Zealands North Island, Carduus acanthoides
L., was not tested), and two Cirsium spp., Cirsium vulgare (Savi) Tenore and Cirsium palustre (L.) Scopoli.
It did not attack Cirsium arvense (L.) Scopoli, Silybum
marianum (L.) Gaertner and Onopordum acanthium
L. (Jessep, 1989b). The damage caused by weevil attack to both Carduus and Cirsium spp. in these trials
was substantial.
In 1992, T. horridus from New Zealand was introduced to Australia for C. nutans control, and releases
were initiated in 1993 (Woodburn, 1997). At the same
time, scientists from Commonwealth Scientific and
Industrial Research Organisation (CSIRO) Australia
145
XII International Symposium on Biological Control of Weeds
were engaged in research in Europe towards a biocontrol programme for Onopordum spp. thistles (Briese
et al., 1994). One of the highly prioritized biocontrol
agents was identified as T. horridus, partly because it
had already been used successfully in Virginia for controlling the related C. nutans thistle (Kok, 1986, cited
by Briese et al., 1994). The weevils were introduced
from Europe into quarantine in Australia, where it
was noticed that they exhibited some consistent nonsimilarities to the T. horridus introduced from New
Zealand for C. nutans control (T. Woodburn, personal
communication). Specimens were sent to a taxonomist
(M. Alonso-Zarazaga), who re-described the species
and revealed that, what was until then regarded as one
species, T. horridus, was in fact a complex of three species (Alonso-Zarazaga and Snchez-Ruiz, 2002). The
samples sent from Australia included specimens of the
Onopordum specialist from quarantine and specimens
from the population established on C. nutans near Canberra. The latter are considered progeny of the weevils
introduced from New Zealand (Alonso-Zarazaga and
Snchez-Ruiz, 2002). The European samples from Onopordum in quarantine in Australia were identified as a
new species, Trichosirocalus briesei Alonso-Zarazaga
and Snchez-Ruiz, with trophic linkage to the thistle
genus Onopordum (larvae and adults did not survive
on Carduus spp. and Cirsium spp.). Specimens from
C. nutans from the Canberra region were, too, identified
as a new species, Trichosirocalus mortadelo AlonsoZarazaga and Snchez-Ruiz, associated with thistles of
the genus Carduus, mainly C. nutans (Alonso-Zarazaga
and Snchez-Ruiz, 2002). It was assumed that the weevils introduced from New Zealand to Australia were
either T. mortadelo or a mixture of T. mortadelo and
T. horridus. The species T. horridus was now associated
with thistles of the genus Cirsium (Alonso-Zarazaga
and Snchez-Ruiz, 2002).
In Australia, efforts to establish weevils originating
from New Zealand on Cirsium spp. were unsuccessful
(A. Swirepick, personal communication), which settles
with their identification as T. mortadelo, the Carduus
specialist. In New Zealand, however, the weevils occur
on both Carduus spp. and C. vulgare, although their
effect on C. vulgare populations is, for the most part,
somewhat poor. Trichosirocalus spp. are extremely attractive biocontrol agents for weeds that exhibit a longlived seed bank, such as C. nutans and Onopordum spp.,
as they reduce the plants vigour such that they either
produce less seeds or even die prior to any seed production; and they also reduce plant biomass at the time
that lapses until seed reserves in the soil are depleted
(Briese, 2006). It was therefore desirable to find out
whether the non-satisfactory control of C. vulgare in
New Zealand was attributed to T. horridus, the Cirsium
specialist, not being present here. In 2006, a survey was
conducted in New Zealand to record which thistle species are attacked by which Trichosirocalus spp.
Methods and materials
A field survey was initiated in autumn 2006 and lasted
to the end of summer 2007. Regional Councils staff
visited release sites in different parts of the country,
collected weevil specimens and, where possible, estimated whether thistle populations decreased since
the weevils introduction into their region. All thistle
species present at each site were noted, as well as the
relative abundance of damaged rosettes (which can be
readily distinguished by the black frass secreted by the
feeding larvae). Where weevils were found on more
than one thistle species at one site, they were collected
separately for each thistle species. Sample size varied
between sites, depending on the difficulty to obtain
specimens. All weevils were preserved in 70% ethanol
prior to identification.
For identification, all weevils were dissected under
a stereo microscope. Their gender was recorded, and
males aedeagi were examined against the key provided in Alonso-Zarazaga and Snchez-Ruiz (2002).
All specimens were then mounted and deposited at the
University of Canterbury arthropod collection for future referencing.
Results and discussion
Different responses from Regional Councils in different parts of the country resulted in large variation in
the number of samples collected per region. Eight regions (three in the North Island and five in the South
Island, Figure 1) and 51 sites were sampled, and 744
adults were dissected, 337 of which were males. In one
region (Auckland), no weevils could be collected. All
males in all samples were identified as true T. horridus
(Table 1). Where estimated, C. nutans populations in
New Zealands South Island appear to have decreased
since T. horridus introduction (Table 2).
Three main points can be highlighted: (1) there is
no evidence for any species other than T. horridus being present in New Zealand (Table 1); (2) T. horridus
readily attacks Carduus, Cirsium and Onopordum thistles (Tables 1 and 3); and (3) C. vulgare appears to be,
for the most part, affected not as strongly as C. nutans
and O. acanthium by the weevil (Tables 2 and 3). The
latter two points are inconsistent with the recent re-
description of the species (Alonso-Zarazaga and
Snchez-Ruiz, 2002). According to this description,
T. horridus should be able to feed on Carduus spp. but
should show preference for Cirsium spp., in particular to C. vulgare. In addition, it has been indicated that
T. horridus might be seen on Onopordum spp. in mixed
thistle stands but that it would not feed on it (M. Alonso-
Zarazaga, personal communication).
The presence and feeding of T. horridus on O. acanthium in New Zealand is of special interest: in the early
1990s, not long after the weevil had been introduced to
146
Which species of the thistle biocontrol agent Trichosirocalus are present in New Zealand?
Figure 1.
Regions of New Zealand visited in the survey with points marking: (a) thistle species present in each region and,
(b) species from which weevils were collected.
Table 1.
Geographic origin of Trichosirocalus spp. samples and species ID.
Region
North Island
Central NI
Number of sites sampled
Host thistle species
Number of males examined
Carduus nutans
20
Carduus nutans
Carduus tenuiflorus
South Island
Tasman
Carduus nutans
20
North Canterbury
Carduus nutans
39
Mid Canterbury
2
2
29
5
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
19
13
158
12
Carduus nutans
15
Cirsium vulgare
Onopordum acanthium
22
Greater Wellington
South Canterbury
Central Otago
Total
44b
337
Based on male aedeagus; identified using the key provided in Alonso-Zarazaga and Snchez-Ruiz (2002).
In some sites, weevils were collected from more than one thistle species.
147
Species IDa
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
Trichosirocalus
horridus
XII International Symposium on Biological Control of Weeds
Table 2.
Estimates of changes to Carduus nutans and Cirsium vulgare populations since the introduction of T. horridus.
Number of sites where populations were
Estimated to have declined
Carduus nutans
North Island
South Island
Cirsium vulgare
North Island
South Island
Estimated to have not declined
3
16
3
6
0
17
1
2
1
2
3
18
New Zealand and had established, T. Jessep (who was
responsible for its introduction and distribution), realizing the weedy status of Onopordum spp. in Australia, decided to establish the weevil on O. acanthium in
New Zealand as a measure to prevent this thistle from
becoming weedy here (M. Turner, personal communication). O. acanthium (the only Onopordum spp. pre
sent in New Zealand) is not common and can mainly
be found in Central Otago. A site was located in Central
Otago, hosting an isolated dense O. acanthium stand,
with no C. nutans to be found at a radius of 10 km. In
1993, 1300 weevils, collected from C. nutans, were released at the site (M. Turner, personal communication).
The weevils have established and have dramatically
reduced O. acanthium density at the site over several
years. Today, only few thistles are germinating, fewer
reach the flowering stage, and ones that do flower show
Table 3.
reduced height and vigour (R. Groenteman, personal
observation). Moreover, with the loss of prickliness attributed to the weevils larval feeding, the capitula appear palatable to the deer grazing on the farm, which
further helps depleting the soil seed bank. The landowner assured us that he does not manage thistles on
the farm by any other control means or land management practices. The weevils were never redistributed
from that original site; yet we were able to collect them
from another O. acanthium stand, about 50 km (aerial
distance) away from the original site. At this distant site,
part of the plants were high and vigorous and exhibited
no weevil feeding signs, whereas others appeared much
less vigorous, exhibited feeding signs, and hosted the
weevil (R. Groenteman, personal observation). It seems
that the weevils dispersed by themselves to the site, and
their impact is beginning to be noticeable. Adults that
Relative abundance of different thistle species in different regions of New Zealand and fraction of sites in which
Trichosirocalus horridus adults and/or damaged rosettes were abundant.
Region
North Island
Auckland
Central NI
Greater Wellington
South Island
Tasman
North Canterbury
Mid Canterbury
South Canterbury
Central Otago
NZ total
Not estimated (unknown site history)
Host species
Presencea
Relative abundance
of thistle rosettes
Frequency of sites in which adult
T. horridus and/or damaged
rosettes were abundant
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
1/1
1/1
3/3
2/3
2/2
2/2
moderately common
moderately common
very common
moderately to very common
rare
rare
0/1
0/1
1/3
0/2
1/2
1/2
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
Carduus nutans
Cirsium vulgare
Onopordum
acanthium
Cirsium nutans
Carduus vulgare
1/1
1/1
3/3
2/3
3/3
3/3
30/30
14/30
2/3
2/3
2/3
moderately common
rare
moderately to very common
rare
varies
moderately common
moderately common mostly
rare to moderately common
moderately common
rare to moderately common
moderately to very common
0/1
0/1
3/3
0/2
1/3
0/3
21/30
5/14
2/2
1/2
2/2
45/46
27/46
29/45
7/27
Presence is expressed as the fraction of sites in which the thistle species was recorded, out of the total number of sites visited in each region.
148
Which species of the thistle biocontrol agent Trichosirocalus are present in New Zealand?
were collected from O. acanthium at both sites were
identified, again, as T. horridus. At the self-introduced
site, adult weevils were also collected from C. vulgare,
and these, too, were identified as T. horridus (Table 1).
The results from the survey thus far present two par
adoxes: firstly, that T. mortadelo was collected from
New Zealand and shipped to Australia but has not been
found in our survey, and secondly, that T. horridus does
not show a distinct preference for Cirsium thistles in
New Zealand as reported elsewhere.
What, then, might the origin of T. mortadelo in Australia be? It appears that although CSIRO scientists
have been engaged in searches for biocontrol agents in
several European countries in the mid 1990s, the only
Trichosirocalus sp. introduced into Australia as a result of this were the Onopordum specialist from Spain,
later named T. briesei; and the only weevils released
on Carduus thistles a few years earlier were of New
Zealand origin (A. Sheppard, D. Briese, T. Woodburn,
A. Swirepick, personal communication). If any Trichosirocalus spp. had been collected from Carduus thistles
in Europe as part of these surveys, they would have
been examined in quarantine only and not released
(A. Sheppard, personal communication), whereas the
Carduus weevils used in the revision were collected
from established populations in Australia (D. Briese,
personal communication).
It has been suggested that T. mortadelo may have
disappeared from New Zealand or was out-competed
by T. horridus. An examination of 30 specimens that
were collected in South Canterbury in 1990, as part of
a Lincoln University Master thesis, does not support
this possibility. These specimens were collected at the
original release site in New Zealand, where the weevils
have been established the longest. This, most probably,
was the site that sourced the weevils for the introduction to Australia. Sixteen of these specimens were
males, and they were identified as T. horridus. Thus, it
is hard to believe T. mortadelo was there at the time but
was not represented in the sample.
A third paradox arises from the revision itself
(Alonso-Zarazaga and Snchez-Ruiz, 2002) and is supported by the personal communications surrounding
this survey: that there is no evidence for T. horridus
presence in Australia.
Conclusions and outlook
To conclude, it appears that the Trichosirocalus spp.
fauna in New Zealand consists of only one species,
T. horridus, feeding on thistle hosts belonging to the
genera Carduus, Cirsium and Onopordum. It is unclear
why the weevils in New Zealand exhibit host utilization that is inconsistent with the recent revision. It also
remains unclear where T. mortadelo in Australia originates from and why T. horridus is absent there. A similar survey in Australia would shed some light on these
matters.
It is probably not desirable to re-introduce T. horridus from C. vulgare to achieve better control of this
thistle, since the specialized populations may blend
with the existing populations already well established
on Carduus. However, it might be useful to harvest
weevils from the few sites in New Zealand where they
do well on C. vulgare and redistribute those. It is certainly unjustified to introduce the Carduus specialist,
T. mortadelo, since T. horridus is well established on
C. nutans and appears to have a significant negative
effect on this weed.
Acknowledgements
We thank Lynley Hayes for coordinating the survey;
Bridget Keenan, Greg Hoskins, Harvey Phillips, Lindsay Grueber, Neil Gallagher and Malinda Matthewson
for collecting weevils, with special thanks to Murray
Turner; Pauline Syrett and Rowan Emberson for additional material; Andy Sheppard, David Briese, Anthony
Swirepick, Tim Woodburn and Miguel Alonso-Zarazaga
for revealing the Australian side of the story; Stephen
Thorpe for ID guidance; and Clair Galilee for technical
assistance. The project was funded by the New Zealand
Foundation for Research Science and Technology.
References
Alonso-Zarazaga, M.A. and Snchez-Ruiz, M. (2002) Revision of the Trichosirocalus horridus (Panzer) species
complex, with description of two new species infesting
thistles (Coleoptera : Curculionidae, Ceutorhynchinae).
Australian Journal of Entomology 41, 199208.
Briese, D.T. (2006) Can an a priori strategy be developed for
biological control? The case of Onopordum spp. thistles in
Australia. Australian Journal of Entomology 45, 317323.
Briese, D.T., Sheppard, A.W., Zwlfer, H. and Boldt, P.E.
(1994) Structure of the phytophagous insect fauna of
Onopordum thistles in the Northern Mediterranean Basin.
Biological Journal of the Linnean Society 53, 231253.
Jessep, C.T. (1989a) Carduus nutans L., nodding thistle
(Asteraceae). In: Cameron, P.J., Hill, R.L., Bain, J. and
Thomas, W.P. (eds) A Review of Biological Control of Invertebrate Pests and Weeds in New Zealand 1874 to 1987.
pp. 339342.
Jessep, C.T. (1989b) Introduction of the crown weevil (Trichosirocalus horridus) as an additional biocontrol agent
against nodding thistle. Proceedings of New Zealand
Weed and Pest Control Conference 42, 5254.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds, Fourth Edition: A World Catalogue of Agents
and their Target Weeds. CABI, Wallingford, UK 223 p.
Kok, L.T. (1986) Impact of Trichosirocalus horridus (Coleoptera, Curculionidae) on Carduus thistles in pastures.
Crop Protection 5, 214217.
Woodburn, T.L. (1997) Establishment in Australia of Trichosirocalus horridus a biological control agent for Carduus
nutans, and preliminary assessment of its impact on plant
growth and reproductive potential. Biocontrol Science
and Technology 7, 645656.
149
Bionomics and seasonal occurrence of
Larinus filiformis Petri, 1907 (Coleoptera:
Curculionidae) in eastern Turkey, a
potential biological control agent for
Centaurea solstitialis L.
L. Gltekin,1 M. Cristofaro,2,3 C. Tronci3 and L. Smith4
Summary
We conducted studies on the life history of Larinus filiformis Petri, 1907 (Coleoptera: Curculionidae:
Lixinae) to determine if it is worthy of further evaluation as a classical biological control agent of
Centaurea solstitialis L. (Asteraceae: Cardueae), yellow starthistle. The species occurs in Armenia,
Azerbaijan, Turkey and Bulgaria. Adults have been reared only from C. solstitialis. In eastern Turkey,
adults were active from mid-May to late July and oviposited in capitula (flower heads) of C. solstitialis from mid-June to mid-July. In the spring, before females begin ovipositing, adults feed on the
immature flower buds of C. solstitialis, preventing them from developing. Larvae develop in about 6
weeks and destroy all the seeds in a capitulum. The insect is univoltine in eastern Turkey, and adults
hibernate from mid-September to mid-May.
Keywords: Larinus filiformis, Centaurea solstitialis, bionomics.
Introduction
Centaurea solstitialis L. (Asteraceae: Cardueae), yellow starthistle, is an important invasive alien weed in
rangelands of the western USA (Maddox and Mayfield,
1985; Sheley et al., 1999; DiTomaso et al., 2006). Although six species of insects have been introduced to
the USA for biological control of this weed, there is still
interest to find additional prospective agents (Balciu
nas, 1998; Smith, 2004a; Pitcairn et al., 2006). Greece
and Turkey are considered to be the geographic centre
of diversity of C. solstitialis (Wagenitz, 1975; Dostl,
Atatrk University, Faculty of Agriculture, Plant Protection Department, 25240 TR Erzurum, Turkey.
2
ENEA C.R. Casaccia, s.p. 25, Via Anguillarese 301, 00060 S. Maria di
Galeria, Rome, Italy.
3
Biotechnology and Biological Control Agency, Via del Bosco 10, 00060
Sacrofano, Rome, Italy.
4
USDA-ARS, Western Regional Research Center, Albany, CA 94710,
USA.
Corresponding author: L. Gltekin <[email protected]>.
CAB International 2008
1
1976). Recent explorations carried out in Eastern Turkey revealed the presence of Larinus filiformis Petri,
1907 (Coleoptera: Curculionidae), a weevil strictly
associated with C. solstitialis (Cristofaro et al., 2002,
2006; Gltekin et al., 2006). L. filiformis was originally
described from Arax Valley (Armenia) and is included
in the Lixinae subfamily (Petri, 1907; Ter-Minassian,
1967). However, nothing is known about its biology.
The goal of this work was to clarify this insects life
history, including host range, seasonal occurrence and
geographic distribution in Turkey.
Methods and materials
Field surveys and observations were conducted from
early spring through summer during 2003. The goals
were to collect live adults for biological experiments
and to observe hibernation places, initiation of adult
activity in the spring, adult feeding, mating, oviposition, larval feeding and development, host plants, oviposition preference, season of occurrence of different
developmental stages and to collect associated parasitoids. The principal study sites were located in Bingl
150
Bionomics and seasonal occurrence of Larinus filiformis Petri, 1907 (Coleoptera: Curculionidae)
Province (from 35 km north-east of Bingl to 15 km
west of Bingl) and in Idr Province, (from 6 km east
of Tuzluca to 7 km east of Idr). Both provinces are
temperate regions. The Bingl region is characterized
by Quercus forest with open areas, including abandoned fields where C. solstitialis commonly occurs.
Idr Province contains the Aras river valley and Ar
mountain lowland area (Idr Plain). The Aras valley
is quite desertified and eroded and dominated mainly
by semi-desert vegetation. When we found a site with
at least 100 C. solstitialis plants, we searched yellow
starthistle plants for signs of L. filiformis (e.g. feeding
and oviposition damage and presence of adults). During 1997 to 2006, while conducting a survey of Larinus
biodiversity, the lead author recorded host plant associations in eastern Turkey by examining many plants in
the tribe Cardueae (Asteraceae). Mature flower heads
were collected and held to rear adult insects from
known host plants.
Results and discussion
Insect morphology
Adult body length is 4.5 to 6.5 mm. The body is
black, the scape, funicle and tarsi are chestnut-brown,
and the tibia and apex of femur are brownish-black
(Figure 1). Golden colour hair-like scales are scattered
sparsely on pronotal disc and body. Elytra are clothed
with bifurcate short whitish-grey scales, which are more
dense on second to fourth intervals (where the first interval is the longitudinal stripe closest to the center of
the insect when wings are folded) and on the lateral
margins. The rostrum is cylindrical, weakly curved, 1.0
to1.5 times as long as pronotum and is distinctly longer
and more shiny in females than males (Figure 1). Elytra
are parallel sided at the basal half and then gradually
curved toward the apex.
Figure 1.
Life history
First adult activity in the spring was recorded during the third week of May. By the end of May, most of
the insects had emerged from hibernation. Early season field collections suggest that males become active
earlier than females. Mating was not observed in the
field until the end of May, but it continued throughout
the adult life span. Adults feed on young buds, on the
central growing tip of the plant and on leaves of C. solstitialis. Later in the season, they feed on the internal
receptacle tissue and flower parts of immature and mature flower heads. Adults were active in the field from
19 May to 3 August 2003.
Based on our observations in the field and cage
studies, three conditions were necessary to begin oviposition: temperature above 25C, feeding for about
20 days and availability of mature C. solstitialis flower
heads. In the field, oviposition was recorded from midJune to mid-July. Females chewed an ovate hole with
their rostrum in the lower part of mature, rarely preflowering flower heads and laid a single egg into the
receptacle tissue. The oviposition hole is then closed
with a secretion. Eggs hatched in about 13 days under
laboratory conditions.
In the field survey, the first larvae were found at the
beginning of July in the Bingl region (1,000 m asl).
The larvae fed on receptacle tissue enlarging a niche,
eventually consuming all the flower parts and receptacle. Larvae fastened together remaining flower parts
and frass to form a hard cell inside the flower head.
Pupation occurs inside the flower head, about 3 days
after mature larvae ceased feeding. Larvae were seen
from the beginning of July to mid-September. Attacked
flower heads can easily be distinguished because they
dried prematurely without flowering.
First adult emergence was recorded in late July in
Idr region. The adults waited inside flower heads about
Adults of Larinus filiformis Petri.
151
XII International Symposium on Biological Control of Weeds
1 week before exiting from the top of the flower. New
adults were recorded in the field until mid-September
at the Idr site. Hibernation started in mid- to late Sep
tember: Adults generally hide under rocks, dry plant
parts or debris in groups of two to three individuals.
Geographical distribution and
host plant range
L. filiformis occurs from eastern Armenia, Nakhi
chevan autonomous region of Azerbaijan and in Turkey, from the east border through central Anatolia and
near the southern Mediterranean coast (Figure 2). One
specimen was determined from Bulgaria, which was a
new record for Europe (Gltekin et al., 2006).
Three years of extensive field observations, collections and laboratory rearing, carried out by our group
in Turkey, indicate that L. filiformis is monophagous on
C. solstitialis.
Natural enemies
Six hymenopteran parasitoid species were reared
from larval and pupal stages of L. filiformis: Bracon
urinator (F.), Bracon tshitsherini Kok., Exeristes roborator F., Aprostocetus sp. and two unidentified wasp
species belonging to the families Eurytomidae and
Ormyridae.
Conclusions
Our results indicate that L. filiformis is closely associated with C. solstitialis in the field in eastern Turkey.
Figure 2.
The insect was very common on C. solstitialis and was
never found feeding on other nearby plants. The adults
destroy many immature flower heads, and larvae destroy all the seeds within flower heads that they infest.
This damage is very similar to that caused by Larinus
minutus Gyll. on spotted and diffuse knapweeds, Centaurea stoebe L. and Centaurea diffusa L.; L. minutus
Gyll. is a biological control agent that appears to be reducing populations of these two weeds in North America (Smith, 2004b; Seastedt et al., 2007). Therefore,
L. filiformis warrants further evaluation as a prospective biological control agent for this weed, especially if
it does not interfere with the other seed head insects that
are already established (Pitcairn et al., 2004; 2005).
Acknowledgements
We are sincerely grateful to Professor Vladimir I. Dorofeyev (Komarov Botanical Institute, Russian Academy
of Sciences, St. Petersburg, Russia) for identification
of plants; Dr S.V. Belokobylskii, Dr D.R. Kasparyan,
Dr K. Dzhanokmen (Zoological Institute, Russian
Academy of Sciences, St. Petersburg, Russia) for identification of parasitoid wasps; Dr Boris A. Korotyaev
(Zoological Institute, Russian Academy of Sciences,
St. Petersburg, Russia) for years of supervision, support and training of the senior author in weevil taxonomy. The senior author was supported by grants from
Collaborative Linkage Grants no. 978845 and NRCLG-981318 of the NATO Life Science and Technology Programme; BBCA research grant and TUBITAKTOVAG-105O038.
Geographic distribution of Larinus filiformis.
152
Bionomics and seasonal occurrence of Larinus filiformis Petri, 1907 (Coleoptera: Curculionidae)
References
Balciunas, J. (1998) The future of biological control for yellow starthistle. In: Hoddle, M.S. (ed.) Proceedings, California Conference on Biological Control: Innovations in
Biological Control Research. University of California,
Berkeley, CA, pp. 9395.
Cristofaro, M., Hayat, R., Gultekin, L., Tozlu, G., Zengin,
H., Tronci, C., Lecce, F., Sahin, F. and Smith, L. (2002)
Preliminary screening of new natural enemies of yellow
starthistle, Centaurea solstitialis L. (Asteraceae) in Eastern Anatolia. In: zbek, H., Gl, . and Hayat, R. (eds)
Proceeding of the Fifth Turkish National Congress of Biological Control. 47 September, 2002, Erzurum, Turkey,
pp. 287295.
Cristofaro, M., Hayat, R., Gltekin, L., Tozlu, G., Tronci,
C., Lecce, F., Paolini, A. and Smith, L. (2006) Arthropod
communities associated with Centaurea solstitialis L. in
Central and Eastern Anatolia. In: VIIIth European Congress of Entomology, Abstract Book. September 1722,
Izmir, Turkey, p.148.
DiTomaso, J., Kyser, G.B. and Pitcairn, M.J. (2006) Yellow
starthistle Management Guide. Cal-IPC Publication 200603. California Invasive Plant Council, Berkeley, CA.
Dostl, J. (1976) Centaurea L. In: Tutin, T.G., Heywood,
V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M. and Webb, D.A. (eds) Flora Europaea. Cambridge University Press, Cambridge, vol. 4, 254301.
Gltekin, L., Cristofaro, M., Tronci, C. and Smith, L. (2006)
Life history of Larinus filiformis Petri (Coleoptera: Curculionidae), potential biological control agent for Centaurea
solstitialis L., and geographical distribution in Turkey. In:
VIIIth European Congress of Entomology, Abstract Book.
September 1722, 2006. Izmir, Turkey, p.151.
Maddox, D.M. and Mayfield, A. (1985) Yellow starthistle infestations are on the increase. California Agriculture 39,
1012.
Petri, K. (1907) Bestimmungs-Tabelle der Gattungen Larinus
Germar (inclus. Stolatus Muls.), Microlarinus Hochhuth,
Rhinocyllus Germar und Bangasternus Gozis aus dem
europischen, mediterranen, west- und nordasiatischen
Faunengebiete. Verhandlungen des naturforschendenVerei
nes in Brnn 45, 51146.
Pitcairn, M.J., Piper, G.L. and Coombs, E.M. (2004) Yellow
starthistle. In: Coombs, E.M., Clark, J.K., Piper, G.L. and
Cofrancesco, A.F., Jr. (eds) Biological Control of Invasive
Plants in the United States. Oregon State Univ. Press,
Corvallis, OR, pp. 421435.
Pitcairn, M.J.,Woods, D.M. and Popescu, V. (2005) Update
on the long-term monitoring of the combined impact of
biological control insects on yellow starthistle. In: Woods,
D.M. (ed.) Biological Control Program Annual Summary,
2004. California Department of Food and Agriculture,
Plant Health and Pest Prevention Services, Sacramento,
CA, pp. 2730.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, J.
(2006) Yellow starthistle continues its spread in California. California Agriculture 60, 8390.
Seastedt, T.R., Knochel, D.G., Garmoe, M. and Shosky, S.A.
(2007) Interactions and effects of multiple biological control insects on diffuse and spotted knapweed in the Front
Range of Colorado. Biological Control 42, 345354.
Sheley, R.L., Larson, L.L., and Jacobs, J.J. (1999) Yellow
starthistle. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State Univ. Press., Corvallis, OR, pp. 408416.
Smith, L. (2004a) Prospective new agents for biological control of yellow starthistle. In: Proceedings 56th Annual
California Weed Science Society, 1214 January, 2004,
Sacramento, CA, pp. 136138.
Smith, L. (2004b) Impact of biological control agents on
diffuse knapweed in central Montana. In: Cullen, J.M.,
Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L.
and Scott, J.K. (eds) Proceedings of the XI International
Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 589593.
Ter-Minassian, M.E. (1967) Zhuki-dolgonosiki podsemejstva
Cleoninae fauny SSSR. Tsvetozhily i stebleedy (triba Lixini). Nauka, Leningrad. (English translation: Weevils of
the Subfamily Cleoninae in the Fauna of the USSR. Tribe
Lixini. USDA Agricultural Research Service, Washington,
D. C. by Amerind Publishing Co. Pyt. Ltd., New Delhi,
1978.)
Wagenitz, G. (1975) 79. Centaurea L. In: Davis, P.H. (ed.)
Flora of Turkey. Edinburgh University Press, Edinburgh,
vol. 5, pp. 465585.
153
All against one: first results of a newly formed
foreign exploration consortium for the
biological control of perennial pepperweed
H.L. Hinz,1 E. Gerber,1 M. Cristofaro,2 C. Tronci,3 M. Seier,4
B.A. Korotyaev,5 L. Gltekin,6 L. Williams7 and M. Schwarzlnder8
Summary
Perennial pepperweed (PPW), Lepidium latifolium L., is a mustard of central Asian origin that is
invading natural and cultivated habitats in North America and is difficult to control with conventional
means. Biological control of PPW is hampered by the fact that it is relatively uncommon in its native
range and that it has more than 30 native North American congeners. In addition, detailed information
on phytophagous organisms and diseases associated with PPW in its native range is sparse. In 2005,
a foreign exploration consortium was formed, and in 2006, five field trips were conducted: three to
Turkey, one to Kazakhstan and one to Romania and Bulgaria. A total of 28 field sites of PPW were
sampled. Based on identifications available thus far and combined with data of previous opportunistic
surveys, we reared or sampled 67 phytophagous organisms, only seven of which have previously been
recorded from PPW. Although plants in Kazakhstan showed more obvious signs of damage than in
Turkey, a similar number of phytophagous organisms were collected. Only three species were found
in both countries, indicating two distinct herbivore communities. At least six potential biological
control agents were found during these first surveys: one root-mining weevil, Melanobaris sp. pr.
semistriata Boheman, and an eriophyid mite, Aculops sp., in Turkey; one gall-forming weevil, Ceutorhynchus marginellus Schultze, a shoot-mining flea beetle, Phyllotreta reitteri Heikertinger, and a
fungal leafspot pathogen, Septoria lepidii Desmazires, in Kazakhstan; and a shoot-mining chloropid
fly, Lasiosina deviata Nartshuk, in both countries. The shoot-mining flea beetle in particular was
found to be very damaging, causing the die-back of shoots. For M. sp. pr semistriata and P. reitteri,
we have established a colony in quarantine at CABI and have started to investigate their biology.
Keywords: Lepidium latifolium, Turkey, Kazakhstan, field surveys.
Introduction
Perennial pepperweed (PPW), Lepidium latifolium L.
(syn.: Cardaria latifolium) (Brassicaceae), is a herbaCABI Europe-Switzerland, Rue des Grillons 1, 2800 Delmont,
Switzerland.
2
BBCA and ENEA-Casaccia, Via del Bosco 10, 00060 Sacrofano
(Rome), Italy.
3
BBCA, Via del Bosco 10, 00060 Sacrofano (Rome), Italy.
4
CABI UK Centre, Silwood Park, Buckhurst Road, Ascot, Berkshire
SL5 7TA, UK.
5
Russian Academy of Sciences, Zoological Institute, 199034 St. Petersburg, Russia.
6
Atatrk University, Department of Plant Protection, 25240 Erzurum,
Turkey.
7
USDA, ARS, Exotic and Invasive Weeds Research Unit, 920 Valley
Road, Reno, NV, USA.
8
University of Idaho, Department of Plant, Soil and Entomological
Sciences, Moscow, ID 83844-2339, USA.
Corresponding author: H.L. Hinz <[email protected]>.
CAB International 2008
1
ceous, semi-woody perennial that typically reaches 0.5
to 1.5 m in height and reproduces vegetatively and by
seed (Renz, 2000). Plants regrow early each year from
a dense network of creeping, horizontal roots, flower
in June/July and set seeds in July/August. PPW is a
prolific seed producer, capable of producing more than
six billion seeds per acre of infestation (Miller et al.,
1986).
PPW is native to central Asia and is believed to have
been introduced into the United States through California around 1900 as a contaminant of sugar beet seeds
(Renz and DiTomaso, 1998). It is now widespread
across North America but especially prevalent in Nevada, Oregon, Utah and California. Perennial pepperweed is declared noxious or prohibited in 14 US states
and Canadian provinces (US Department of Agriculture (USDA)Natural Resources Conservation Service
(NRCS) Plants National Database, http://plants.usda.
gov; Invaders Database, http://invader.dbs.umt.edu).
154
All against one: first results of a newly formed foreign exploration consortium
PPW is often associated with mesic habitats, such
as riparian areas, drainage ditches and subirrigated
pastures and hay meadows. However, it can invade a
wide range of habitats including pastures, open fields,
roadsides and residential areas (Young et al., 1998).
Downstream movement of seeds and root fragments
along waterways and irrigation systems is the primary
mode of spread for this weed. PPW is highly competitive, and invasions result in dense monocultures and
subsequent loss of biodiversity through the exclusion
of native, riparian vegetation (Trumbo, 1994; Young et
al., 1995; Blank and Young, 1997). In areas that are
not mown annually, semi-woody stems can accumulate
over the years, negatively impacting nesting habitat
for birds and hindering the grazing and movement of
livestock and wildlife (Renz, 2000). In addition, like
Tamarix, PPW alters habitat by increasing soil salinity
(Blank and Young, 1997), thereby hindering regeneration of native flora. Conventional control methods
are not necessarily cost-effective or environmentally
safe and must need to be repeated over several years
to be successful. Therefore, CABI Europe-Switzerland
(CABI E-CH), Biotechnology and Biological Control
Agency (BBCA) and associated collaborators formed
a foreign exploration consortium in 2005 to investigate
the feasibility for biological control of PPW.
Literature surveys
The native range and area of origin of PPW are not clearly
understood because of high intraspecific variability and
the fact that the plant has been cultivated as a spice and
vegetable since the twelfth century (Hegi, 1986). The
plant is probably native to central Asia, with a centre of
species diversity in Tajikistan, Kazakhstan and western
Mongolia (E. Jger, personal communication).
During a literature search for phytophagous organ
isms associated with PPW, we found 23 insect species
(Hinz et al., 2007). Twelve of these are beetles, and
nine are heteropterans. In addition, the list contains one
mite and eight fungal pathogens. Overall, surprisingly
little information is available on phytophagous organ
isms and diseases associated with PPW, especially com
pared to the closely related hoary cress, Lepidium draba
L., another invasive perennial mustard, for which we
found nearly 200 associated phytophagous organisms
(Cripps et al., 2006). Lack of detailed investigations
on PPW in its area of origin may reflect its more
eastern distribution and relative scarcity. This has been
confirmed by results of our first field surveys that have
already revealed new host associations with PPW (see
below).
Field surveys
In 2006, five field trips were conducted, three to central and northeastern Turkey, one to southeastern Kazakhstan and one to Bulgaria and Romania. During the
three field trips to Turkey in April, June and October,
ten PPW sites were found, while in Kazakhstan (1431
May), 19 sites of PPW were found (Fig. 1). Literature
records indicate that PPW occurs in northeastern Bulgaria and southern Romania; however, we were unable
to find PPW at several locations mainly along the Danube River.
At each field site, adult insects were sampled, plants
were dissected, and eggs and immature larvae were
transferred into artificial diet for rearing. In addition,
plant parts (leaves, shoots and roots) with immature
stages or signs of mite or pathogen attack were collected and transported to the lab. Immature stages were
reared to adult and sent, together with field-collected
adults, to taxonomists for identification. Based on identifications available thus far and combined with data of
opportunistic surveys during previous years (see Hinz
and Cristofaro, 2005), we reared or sampled 65 species,
59 of which are probably associated with plants in the
family Brassicaceae.
So far, 28 species have been identified from Kazakhstan and 25 from Turkey. Interestingly, only three
insect species, based on identifications available thus
far, were recorded from both countries, indicating the
existence of distinct herbivore communities in both areas and endorsing the importance of conducting field
surveys for PPW over a large geographic area. Similar
to results of surveys on other brassica weeds (e.g. hoary
cress), two thirds of insects sampled and reared are
weevils (Curculionidae; n = 30) and leaf beetles (Chrys
omelidae; n = 16). However, not all identifications have
been completed.
Kazakhstan
In terms of potential biological control agents, Kazakh
stan is a very interesting country, presumably because
it forms part of the area of origin of PPW. Sites were
relatively easy to find but never very large. PPW growing along the roadside was usually relatively healthy or
attacked by oligophagous or polyphagous insects, such
as heteropterans. Plants with heavy insect and pathogen damage were mostly located in natural, relatively
wet areas, such as the Ili River corridor, where PPW
was often growing in close vicinity or intermixed with
common reed, saltcedar and/or Russian olive.
Of particular interest are the flea beetle Phyllotreta
reitteri Heikertinger, two weevil species Lixus myagri
Olivier and Ceutorhynchus marginellus Schultze and
the chloropid fly Lasiosina deviata Nartshuk (Table 1).
Plants at several sites in Kazakhstan were attacked by
a white rust, identified as Albugo candida (Pers.). The
species is known to attack several crops in the mustard family and is already present in several areas in
the western USA. Also identified were the pycnial and
aecial stages of an unknown, possibly heteroecious
rust, as well as the coelomycete fungus Septoria lepidii Desmazires, reported from several species within
155
XII International Symposium on Biological Control of Weeds
156
Figure 1.
Overview of field surveys conducted for potential biological control agents for Lepidium latifolium, perennial pepperweed, in 2006.
All against one: first results of a newly formed foreign exploration consortium
Table 1.
Potential biological control agents for Lepidium latifolium, perennial pepperweed.
Species
Family
Plant part attacked
Country found in
Ceutorhynchus marginellus
Curculionidae
Kazakhstan
Melanobaris sp.n. semistriata
Phyllotreta reitteri
Lasiosina deviata
Aculops sp.
Septoria lepidii
Curculionidae
Chrysomelidae
Chloropidae
Eriophyidae
Coelomycete
Petioles, leaves, stems
(gall inducer)
Root, root crown
Stem
Stem
Inflorescences and leaves (?)
leaf
the genus Lepidium (Cardaria). Identification of other
disease symptoms is still under way. Small flea beetles occurring in large numbers on PPW were also
collected. However, most of these were identified as
oligophagous or polyphagous species. In addition, two
weevil larvae were found mining in the root crown of
PPW at one site, and leaf-mining flies were collected
at several sites. Unfortunately, none of these could be
reared through to adult.
Turkey
We concentrated our surveys in eastern Turkey and
in Central Anatolia and the Ankara region. As in Kazakhstan, most sites were in relatively wet areas, i.e.
along rivers, irrigation channels and sewage ditches, in
both natural and urban areas.
Compared to Kazakhstan, plants in Turkey showed
less obvious signs of damage. Nevertheless, a similar
number of phytophagous organisms was collected (see
above). Three potential biological control agents were
found: the weevil M. sp. near semistriata Boheman,
the chloropid fly L. deviata and an eryophyid mite,
preliminarily identified as Aculops sp. near lepidii.
Two additional species of potential interest are an as
yet unidentified flea beetle, Psylliodes sp., that probably mines in the shoots of PPW, and the root-mining
weevil L. myagri. However, the latter is reported to be
oligophagous (Dieckmann, 1983).
Potential biological control agents
Melanobaris. sp. pr. semistriata
(Coleoptera: Curculionidae)
This root-mining weevil is recorded from several
PPW sites in Turkey, and data on its biology and phenology have been collected since 2002 (Gltekin et
al., unpublished data). Adults usually become active
in mid-May and feed on young petioles and the base
of stems. In mid-June, the first larvae were found feeding in the petioles. Larvae then move down to the root
crown, where they also pupate. New generation adults
emerge in the autumn. During the field trip to Turkey at
the beginning of April 2006, 28 adults were collected
and taken back to Switzerland to establish a colony in
quarantine and develop methods for host-specificity tests.
Turkey
Kazakhstan
Turkey, Kazakhstan
Turkey
Kazakhstan
Weevils were either released onto potted plants of
PPW or transferred into cylinders and offered cut plant
material inserted in wet florist foam blocks. Oviposition occurred in the foam material, which may suggest
that in the field, M. sp. pr. semistriata females lay eggs,
not as most other weevil species into the plant tissue
but into the soil close to the root crown. Host-specificity tests will start in 2007.
The weevil is very close to M. semistriata Boheman,
normally developing on L. draba (hoary cress), another
invasive Brassicaceae studied at CABI, but may well
represent a distinct undescribed species (Korotyaev,
unpublished data).
Ceutorhynchus marginellus
(Coleoptera: Curculionidae)
During the field trip to Kazakhstan in May 2006,
we found several sites, at which PPW plants were attacked by this gall-inducing weevil. Not much information is available on the species, except that it has
been collected previously on PPW in Russia (Isaev,
1994) and also occurs in Kazakhstan. Galls are formed
on leaf midribs, leaf stalks and stems, and weevil larvae were mining inside. Gall induction appears to stunt
shoot growth. At one site, we also found adults of
C. marginellus. Infested material and adult weevils
were returned to CABIs quarantine facility in Switzerland. Mature larvae left the material to pupate in
the soil, and adults emerged between 12 and 27 June.
Weevils were artificially overwintered in an incubator
at 3C and started to lay eggs in March 2007.
Phyllotreta reitteri (Coleoptera:
Chrysomelidae)
At two field sites in Kazakhstan in May, we found
PPW plants that were heavily attacked by larvae of P.
reitteri. The species is also recorded from Uzbekistan
and the Crimea Peninsula (Lopatin, 1977). No information is available on the biology of this species. We
observed larvae mining in the stems of PPW. At one
site, nearly 100% of plants were attacked, and heavy
attack caused the die-back of shoots. We took infested
material back to CABIs quarantine facility in Switzerland, and between 16 June and 7 July, adults emerged.
Beetles were successfully overwintered and started to
157
XII International Symposium on Biological Control of Weeds
lay eggs in March 2007. We are currently transferring
larvae onto potted PPW plants to establish a colony and
conduct preliminary host specificity tests.
Lasiosina deviata (Diptera: Chloropidae)
This chloropid stem-mining fly was reared for the
first time from PPW in June 2005 (Gltekin, unpub
lished data) and later described by Nartshuk (2006).
During surveys in 2006, we mainly found it in central
Turkey but presumably also at one field site in Kazakh
stan. However, whether the flies from Kazakhstan are
exactly the same species as the ones collected in Turkey still needs to be confirmed. The fly can be quite
frequent; for example, at one site 60% of shoots were
attacked, and up to 11 larvae were found mining in one
stem (mean SE: 3.8 1.1). In June, mature larvae and
pupae were found. Unfortunately, most adults emerged
during the field trip and died, so no further studies
could be initiated.
The genus Lasiosina comprises 34 species in the Palaearctic (Nartshuk, 2006). This is the first record of a
Lasiosina species from a dicotyledon plant. The flies
usually develop on cereals and grasses (Poaceae). We
are therefore planning to check monocot plants growing intermixed with PPW at L. deviata field sites for attack and conduct preliminary host-specificity tests concentrating on one or two important cereals (e.g. wheat)
and grass species. Depending on results, we will make
a decision whether L. deviata can truly be regarded as a
potential biocontrol agent for PPW.
Aculops sp. (Aceria: Eriophyidae)
This eriophyid mite was only found at one field
site in Central Turkey in June 2006. Plants attacked by
the mites appeared to have stunted inflorescences and
showed discoloration. The species resembles A. lepidii
Roivainen, which was originally described from a single location within the Sierra Nevada, Spain (Roivainen, 1953). It was only found associated with Lepidium
stylatum (Lag. et Rodr.) and is reported to attack leaves,
causing abnormal pilosity, leaf margin rolling and curl
ing. The differences in symptoms, host plant and habitat conditions between A. lepidii and the mite found in
Turkey indicate that it is most probably a new species
or subspecies. During a field trip in April to Turkey,
we found PPW plants with curled leaves, symptoms resembling those described for A. lepidii. More material
will be collected and sent to taxonomists in 2007 to
resolve the taxonomic position of the mites found and
to make a complete morphological description.
Septoria lepidii
(Mycosphaerellaceae: Dothideales)
The coelomycete Septoria lepdii was also only
found at one PPW site in Kazakhstan, causing leafspots on infected host plants. Although it is reported
from a number of Lepidium species in Europe and Asia,
specific host strains might exist. Septoria species, consti
tuting the asexual stage of Mycosphaerella sp., have
previously been used successfully as biological control
agents, i.e. Septoria passiflora Syd. against the alien invasive Passiflora tripartita (Juss.) Poir. (Banana Poka)
in Hawaii (Julien and Griffiths, 1998). We are planning
to collect more material in 2007 and start preliminary
pathogenicity and host-specificity tests.
Conclusions and outlook
Results of these first field surveys are very promising
and have revealed several insects and one fungal pathogen with potential for biological control of perennial
pepperweed. We have already established colonies of
two species, one root-mining weevil and one stem-mining
flea beetle and started preliminary host-specificity tests.
In 2007, we will collect material of at least two addi
tional species. The identification of all remaining material collected in 2006 will be concluded, and depending
on results, a revised list of potential agents prepared. In
addition, we will continue and extend field surveys to
regions not yet covered, such as western China, Mongolia, southern Russia, and southern Ukraine. Finally,
we will collect herbarium specimens and material for
molecular analysis to better understand the genetic vari
ability of PPW in Eurasia and the origin of invasive
populations in North America.
Acknowledgements
We thank Ghislaine Cortat, Bethany Muffley and Cristobal Tostado for additional technical assistance in
the lab; Florence Willemin and Christian Leschenne
for plant propagation; Dr Roman V. Jashenko and
Dr Anna Ivashenko for facilitating our field trip to
Kazakhstan and providing access to the herbarium collection in Almaty; and the following taxonomists for
species identifications: Prof Paolo Audisio (Nitidulidae),
Dr Enrico De Lillo (Acari), Dr Alexander S. Konstantinov (Alticinae) and Prof. Vera A. Richter (Syrphidae).
The following people kindly provided additional information on the distribution of L. latifolium: Dr Roman V.
Jashenko, Dr Ilhan Kaya and colleagues, Prof. Sergey
Mosyakin, Dr Rumen Tomov, Margarita Yu Dolgov
skaya and her group and Prof. Vladimir I. Dorofeyev.
Margarita Dolgovskaya and her group conducted a survey of the Russian literature on insects associated with
PPW, and Prof. Dr Eckehart Jger provided valuable
information on the distribution and taxonomy of PPW.
Dr John Gaskin volunteered to investigate the genetics
of L. latifolium. Financial support for this project was
provided by the Wyoming Biological Control Steering
Committee, the Idaho State Department of Agriculture
and the Bureau of Indian Affairs through the University of Idaho, the Californian Department of Food and
Agriculture and the USDA-ARS Western Region Re-
158
All against one: first results of a newly formed foreign exploration consortium
search Center, Reno, NV, and would not have been
possible without the continuing effort and dedication
of Nancy Webber (PPW Consortium chair, WY), Mike
Pitcairn (CDFA), and Lincoln Smith and Ray Carruthers (both USDA, ARS, Albany, CA). The study by
B. Korotyaev was supported by the Russian Foundation for Basic Research (07-04-00482-a and 04-0481026 Bel2004a).
References
Blank, R. and Young, J.A. (1997) Influence of invasion of perennial pepperweed on soil properties. USDA Agricultural
Experiment Station, Oregon State University, Special Report 972, 1113.
Cripps, M.G., Hinz, L.H., McKenney, J.L., Bradley, L.,
Harmon, B.L., Merickel, F.W. and Schwarzlaender, M.
(2006) Comparative survey of the phytophagous arthropod faunas associated with Lepidium draba in Europe and
the western United States, and the potential for biological weed control. Biocontrol Science and Technology 16,
10071030.
Dieckmann, L. (1983) Beitrge zur Insektenfauna der DDR.
Coleoptera-Curculionidae (Tanymecinae, Raymondionyminae, Bagoinae, Tanysphyrinae). Beitrge zur Entomologie 33, 128 pp.
Hegi, G. (1986) Illustrierte Flora von Mitteleuropa. Spermatophyta, Band IV Teil 1. Angiospermae, Dicotyledones 2.
Paul Parey, Berlin, Germany, pp. 126131.
Hinz, H.L., and Cristofaro, M. (2005) Prospects for the Biological Control of Perennial Pepperweed, Lepidium latifolium. A joint report prepared by CABI Bioscience Switzerland Centre and BBCA, Rome, Italy, 23 pp.
Hinz, H.L., Gerber, E., Cristofaro, M. and Tronci, C. (2007)
Biological Control of Perennial Pepperweed, Lepidium
latifolium. A joint report prepared by CABI EuropeSwitzerland and BBCA, Rome, Italy, 23 pp.
Isaev, A.Yu. (1994) Ecological-faunistic review of weevils
(Coleoptera: Apionidae, Rhynchophoridae, Curculionidae) of Ulyanovsk Province. Priroda Ulyanovskoi Oblasti
(Nature of Ulyanovsk Province) 4, 77 pp. (in Russian).
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds. A World Catalogue of Agents and their Target
Weeds. Fourth Edition. CABI Publishing, Wallingford,
UK, 223 pp.
Lopatin, I.K. (1977) Zhuki-listoyedy Srednei Azii i Kazakh
stana: (Opredelitel) (Leaf beetles of Middle Asia and Kazakhstan: (A key)) Leningrad, 270 pp. (in Russian).
Miller, G.K., Young, J.A. and Evans, R.A. (1986) Germination of seeds of perennial pepperweed (Lepidium latifolium). Weed Science 34, 252255.
Nartshuk, E.P. (2006) A new species of grassflies of the genus
Lasiosina Becker from Turkey (Diptera: Chloropidae).
Zoosystematica Rossica 14, 289291.
Renz, M.J. (2000) Element Stewardship abstract for Lepidium latifolium L. perennial pepperweed, tall whitetop.
The Nature Conservancy, Arlington, VA, 22 pp.
Renz, M.J. and DiTomaso, J.M. (1998) The effectiveness
of mowing and herbicides to control perennial pepperweed in rangeland and roadside habitats. Proceedings
from the 1998 California Weed Science Conference 50,
178.
Roivainen, H. (1953) Some gall mites (Eriophyidae) from
Spain. Archivos do Instituto de Aclimatacion 1, 941.
Trumbo, J. (1994) Perennial pepperweed: a threat to wildland
areas. CalEPPC Newsletter 2, 45.
Young, J.A., Turner, C.E. and James, L.F. (1995) Perennial
pepperweed. Rangelands 17, 121123.
Young, J.A., Palmquist, D.E. and Blank, R. (1998) The ecology and control of perennial pepperweed (Lepidium latifolium L.). Weed Technology 12, 402405.
159
Potential biological control agents for
fumitory (Fumaria spp.) in Australia
M. Jourdan,1 J. Vitou,1 T. Thomann,1
A. Maxwell2,3 and J.K. Scott2
Summary
Fumaria species are increasingly problematic in the cropping regions of southern Australia, and one
fumitory, Fumaria densiflora DC, has developed populations with herbicide resistance. Consequently,
the potential for biological control was assessed. Nine species of fungi were found associated with
Fumaria spp. in a survey of 33 sites in southern France. According to the literature, species potentially
host specific to fumitory include Cladosporium brachormium Berk. and Broome, Entyloma fumariae
J. Schrt. and Peronospora affinis Rossmann. Of the insects detected on Fumaria spp. in France, the
stem weevil, Sirocalodes mixtus Mulsant and Rey has potential as a biological control agent because
it is thought to be host specific. None of these species were detected amongst the six pathogen species found during surveys of 64 locations in southeastern and southwestern Australia. The absence of
pathogens and insects associated with Fumaria species in Australia, the lack of Fumaria spp. native
to Australia, and few closely related crops or ornamental species, indicate that there are opportunities
for research into the development of natural enemies for the biological control of fumitory.
Keywords: Fumaria species, biological control, field surveys, fungal pathogens,
arthropods.
Introduction
Fumitory species are weeds of many parts of the world,
mainly in cereal and legume cultivation, vineyards,
wastelands and gardens, but have not been considered
as a target for biological control. In Australia, fumitory
is a problem in canola and pulse crops (particularly
peas and lupins; Holding and Bowcher, 2004). Lemerle
et al. (1996) found Fumaria species in 37% of 86 cereal
crops in southern New South Wales during a survey
of weeds conducted in spring 1993. Agronomists and
farmers from this region ranked fumitory in the top ten
of 50 species in terms of potential threat. Fumaria densiflora DC first evolved resistance to group K1/3 herbicides (Ditroanilines and others) in New South Wales
(NSW) (Heap, 2007), underlining that this species is the
most important weed of the genus in Australia (Norton
et al., 2004). In Australia, all Fumaria species (seven
species and two subspecies: Fumaria bastardii Boreau,
Fumaria capreolata L., Fumaria capreolata L. subsp.
capreolata, F. densiflora, Fumaria indica (Hausskn.)
Pugsley, Fumaria muralis Sond. ex W.D.J.Koch, Fumaria muralis Sond. ex W.D.J.Koch subsp. muralis,
Fumaria officinalis L., Fumaria parviflora Lamarck)
are introduced, mostly from Europe (Australian National Botanical Garden, 2007 a, b; Norton, 2003). Fumaria species were found in all States of Australia and
are abundant only in southern regions, mostly in New
South Wales, Victoria, South Australia and Western
Australia. The increasing importance of Fumaria spp.
and the rise of herbicide resistance point to the need
to investigate alternative means of control. This paper
reports on preliminary surveys in France and Australia
to find potential biological control agents.
Methods
Surveys in France
CSIRO European Laboratory, Campus International de Baillarguet,
34980 Montferrier-sur-Lez, France.
2
CRC for Australian Weed Management and CSIRO Entomology, Private Bag 5, PO Wembley, WA 6913, Australia.
3
AQIS, PO Box 606, Welshpool, WA 6986, Australia.
Corresponding author: M. Jourdan <[email protected]>.
CAB International 2008
1
Surveys were conducted in France from April 2004
to May 2005. Populations of Fumaria were found at
33 sites out of 79 inspected, either in vegetable culture
or vineyards or on areas of freshly disturbed soil. The
sites were distributed along five transects covering the
Riviera between Hyres, St Tropez and Lake Ste Croix
160
Potential biological control agents for fumitory (Fumaria spp.) in Australia
(circa 270 km), the Cvennes between Mount Aigoual
and Montpellier (ca. 110 km), the Pyrnes between
The Boulou and Prades (ca. 130 km), the Atlantic coast
between Tarbes and St Jean de Luz (ca. 200 km) and
Bordeaux to Agen (ca. 150 km) and one longer transect
through agricultural regions (ca. 1580 km).
At each site, plants were searched visually for evidence of disease. Samples of diseased tissue were collected and examined in the laboratory. When a pathogen
was found associated with a symptom, after its fructification in a humid chamber, a single spore was collected
with a sterile needle under a stereomicroscope, depos
ited on a potato dextrose agar (PDA; Difco) Petri dish
and placed in a controlled environment (20C and 12h
photoperiod). Fungi were identified morphologically
by microscopic observations.
Plants were searched for evidence of insects at each
site, and from four to ten entire plants were collected
then taken back to the laboratory in transparent plastic
bags for dissection. Plants with evidence of insect damage were cut up and put in Petri dishes, on wet paper, to
allow eggs or larvae to develop to the adult.
Surveys in Australia
Transects were made through grain-growing regions in southeastern Australia (NSW and Victoria)
and southwestern Australia between April and October
2005. The transects were made through districts where
Fumaria species have been previously recorded, and
crops and other roadside habitats were briefly searched
for the presence of fumitory at intervals of approximately 25 km along the length of each transect. The
transect distance travelled was approximately 1200 km
in each of NSW and Victoria and 1800 km in southwestern Australia. At sites where fumitory was detected, at least 20 plants from that site were examined
closely over a 20- to 30-min period for symptoms of
disease and insect attack. Where arthropods were detected or signs of arthropod attack were observed, this
was noted. When fruits were present, these were examined from at least 20 plants using a 10 hand lens for
evidence of insect-caused galls.
Representative samples from each disease category
were collected from each site, placed into polyethylene
bags and stored in an insulated cool box with ice until
they were brought to the laboratory. Leaf symptoms
were recorded, and pathogens were isolated and identified as described below.
Fungal isolation and species identification: Subsamples of diseased material were placed into humid
chambers for up to 14 days at 15C and 20C under a
12h near-UV light: 12 h dark photoperiod to encourage fungal sporulation. The humid chambers consisted
of two to three layers of moistened sterile filter paper
in sterile 90 mm Petri dishes or plastic take-away
containers (150 mm 80 mm 50 mm) sealed with
Parafilm wrap or plastic cling film (Gladwrap). In addition, diseased shoot and root material was surface
disinfested by soaking in 2.5% sodium hypochlorite
(v/v) for 1min, rinsed in three changes of sterile tap
water, blotted dry and plated onto 2% malt extract agar
(MEA; Difco) or water agar (20 g Difco agar made
up to 1 l with deionised water). Pure fungal cultures
were obtained by transferring single spores or single
hyphal tips onto fresh media and maintained on MEA,
PDA or V8 juice agar (200 ml Campbells V8 juice, 3 g
CaCO3, 15 g Difco agar made up to 1 l with deionised
water). Cultures were grown under white fluorescent
and near-UV light in a 12 h light:12 h dark photoperiod
at 20C in order to induce fruiting structures to facilitate species identification.
Small pieces of culture were mounted under acidified glycerol blue [0.05 % aniline blue (Gurr) in lacto
glycerol] and investigated under the light microscope
for the formation of fructifications. Fungal structures
were investigated on normal or phase-contrast settings
(1001000), and 30 measurements were made of conidial dimensions. Fungi were identified according to
standard mycological texts, and original descriptions
where indicated.
Results
Twelve genera including 15 identified species of fungal
pathogens have been reported in the literature as associated with Fumaria spp. (results not shown). Three
species are considered to be potentially host specific
at the level of the plant family and genus (Cladosporium brachormium Berk. and Broome (anamorphic
Mycosphaerellaceae), Entyloma fumariae J. Schrot.
(Entylomataceae) and Peronospora affinis Rossmann
(Peronosporaceae)).
Surveys in France
Pathogens: Nine species of fungi were found associated with Fumaria species. Diseased leaves of Fumaria sp. with symptoms of leaf smut, E. fumariae,
were collected from six sites in November, January
and April. Symptoms of leaf smut appear as whitish
spots on both sides of Fumaria leaves and occasionally petioles. Lesions become dark brown and entire leaves shrivel and eventually die. The three kinds
of spores were observed, two external: primary sporidia (filiformes) and secondary sporidia on sterigma
(ballistospores). Microscopic examination of internal
leaf tissues revealed masses of round, double-walled,
pale green-to-yellow spores typical of the ustilospores of Entyloma. E. fumariae was grown on artificial
culture medium (PDA) and formed ballistospores that
161
XII International Symposium on Biological Control of Weeds
have a similar structure and form to those produced on
plants.
P. affinis was found on Fumaria spp. in ten sites of
the 33 sites surveyed in France. The first symptoms
observed were small faint yellow spots on the upper
foliage. As they matured, fungus fruiting structures
were visible on the lower leaf surface as conidiophores
bearing conidia. Two kinds of spores were observed in
field material: conidia on conidiophores and oospores
in infected leaves.
Diseased plants of Fumaria spp. with symptoms of a
Cladosporium species were collected from 13 sites. Extensive sporulation on leaves, inflorescences and stems
were observed.
Two Alternaria spp. (anamorphic Pleosporaceae),
a Botrytis sp. (anamorphic Sclerotiniaceae), a Colletotrichum sp. (anamorphic Phyllachoraceae), an
Oidium sp. (anamorphic Erysiphaceae) and Phomopsis leptostromiformis (J.G. Khn) Bubk (Valsaceae)
were also found associated with diseased plants of
Fumaria spp.
Arthropods: The only two polyphagous aphid species
cited in the literature on Fumaria spp. were observed
at three sites in France, Aphis fabae Scopoli and Macrosiphum euphorbiae (Thomas). A polyphagous lepidoptera, Plusia sp. (Noctuidae), was also observed at
five sites. Otherwise, Sirocalodes mixtus Mulsant and
Rey (Curculionidae), with stem-mining larvae, was the
most common insect, being found at 13 of 33 sites.
Surveys in Australia
Pathogens: The survey found fumitory at 64 of the
160 sites visited throughout the grain-growing districts
of southern Australia. Six fungal species were identified
from diseased fumitory plants: Alternaria alternata (Fr.)
Keissl. (anamorphic Pleosporaceae), Alternaria dauci
(J.G. Khn) J.W. Groves and Skolko (anamorphic Pleosporaceae), Davidiella tassiana (De Not.) Crous and U.
Braun (Mycosphaerellaceae), Phoma sp. (anamorphic
Pleosporaceae), Pleospora sp. (Pleosporaceae) and Sclerotinia sp. (Sclerotiniaceae). In addition, several saprophytic and unidentified fungi that failed to sporulate were
isolated from diseased material (data not shown). Olpidium brassicae (Woronin) P.A. Dang. (Olpidiaceae), previously recorded on fumitory in Western Australia (WA),
was not recorded in the current survey. No pathogenic
bacteria or nematodes were detected on fumitory.
The fungi were identified from six different fumitory species. F. muralis and F. capreolata were the only
two species present in the WA surveys. In addition to
these two species, F. bastardii, F. densiflora, F. officinalis and F. parviflora were observed in the surveys
of eastern Australia. The plants were not identified to
subspecies rank. F. indica was not detected in the current survey.
Arthropods: No arthropods were found on Fumaria
in Australia during the surveys. There was evidence
of possible sap-feeding insects on some plants at
a few sites; however, no causative insects were pre
sent on the plants. No arthropods associated with
fumitory have been reported from Australia in the
literature.
Discussion
The genus Fumaria is placed in the Subfamily Fumarioideae in the Family Papaveraceae (Lidn, 1986; Stevens, 2007). In earlier taxonomic treatments, the subfamily is treated as the Family, Fumariaceae. Fumaria
is chiefly a Mediterranean group of 55 species with
few species reaching India and east Africa (Lidn,
1986). There are no native Australian species in the
Family Papaveraceae; the only genera of this family in Australia apart from Fumaria are introduced
species of Argemone (three species), Dicentra (one
species), Eschsholzia (one species), Glaucium (two
species), Hypecoum (one species), Papaver (six species), Platycapnos (one species), Pseudofumaria
(one species), Roemeria (one species) and Romneya
(two species) (Australian National Botanical Gardens,
2007a,b). The only closely related crop to Fumaria is
Papaver somniferum L. (poppy) which is grown in
Tasmania for production of pharmaceutical chemicals (Laughlin et al., 1998). Fumaria species are also
weeds in this crop (Baldwin, 1977). Related ornamentals grown in Australia include species of Chelidonium, Corydalis, Dicentra, Eschscholzia, Glaucium,
Macleaya, Meconopsis, Romneya and Pseudofumaria
(Spencer, 1997). Thus the absence of related native
species and few related commercial species indicates
that weedy Fumaria species could be ideal targets for
biological control.
Surveys
The identification of pathogens is provisionary, and
further work is needed to confirm identifications with
molecular techniques or by appropriate experts. The
sampling was carried out in France during an exceptionally dry year (2005). Even so, up to five pathogens
were found per site, and what appeared to be sites with
damaging infestations were found for P. affinis, E. fumariae and P. leptostromiformis. A more complete picture of the number of pathogens could be obtained by
extending the survey to other western and southern European areas where there is an abundance of Fumaria
species (Tutin et al. 1993). However, based on the literature, the survey in France has found most organisms
of interest. This is also the first report of presence of
E. fumariae in France. The sampling in Australia shows
that this fungus and the other potential agents are highly
likely to be absent.
162
Potential biological control agents for fumitory (Fumaria spp.) in Australia
Pathogens
E. fumariae appears to be the most promising potential agent especially as it attacks several Fumaria
species and is not present in Australia. E. fumariae has
been found on F. muralis in Madeira Islands, Fumaria
rostellata Knaf in Romania and Fumaria vaillantii
Loisel. in Germany and Sweden (Vanky, 1994) and on
F. parviflora in India (Zundel, 1953). Vanky (1994)
noted about E. fumariae in Europe Rarely reported,
almost certainly because E. fumariae is very inconspicuous.
Despite this, recent experience shows that Entyloma
species can be successful agents. Barton et al. (2007)
report that Entyloma ageratinae R.W. Barreto and H.C.
Evans was introduced in 1998 with success for the biological control of mist flower, Ageratina riparia (Regel) R.M. King and H.Robinson, in New Zealand. By
2004, the proportion of leaves infected by the fungus
increased to 60%. The estimated cover of A. riparia
in heavily infested plots decreased from 81% to 1.5%
with a corresponding recovery of the diversity of native
vegetation (Barton et al., 2007).
P. affinis, an obligate plant parasite, was abundant
at three sites in France and not reported from Australia. It is associated with Fumaria species making it a
candidate for further consideration for biological control. However, there is a report in Farr et al. (2007) of
a 1902 herbarium specimen from Italy with P. affinis
on the leaves of Chenopodium album L. (Chenopodiaceae). Host range tests are needed to establish if this is
a valid host.
The third species that could be considered is C. brachormium since it is only known from F. officinalis.
It has not been reported from Australia. Cladosporium
species are known generally as saprophytes, but they
can be also pathogens. Otherwise, there appears to be
very little known about this fungus.
Insects
Few insect species were found during the surveys
and most of them are polyphagous. Little information
on fumitory and insects is available in the literature.
Three curculionids are cited on Fumaria (Hoffmann,
1954). Sirocalodes quercicola Paykull is a gall former
developing in crown of F. officinalis and F. capreolata.
This species is found in Europe but is not common.
Sirocalodes nigrinus Marsham is common in France
particularly in the southeast coastal region. The larvae
develop in stems on F. officinalis and F. parviflora. The
adult was also recorded on F. vaillantii, F. capreolata,
Platycapnos spicata (L.) Bernh. (Hoffmann, 1954) and
Papaver rhoeas L. (Campobasso et al., 1999).
During our surveys, we only found the third species, S. mixtus. It occurs throughout France, but it is
more common in the Mediterranean region (Hoffmann,
1954). This species could be considered as a poten-
tial agent, and further studies are needed to define the
host-plant specificity of this insect. It has only been associated with F. officinalis and F. parviflora in France
(Hoffmann, 1954). Adults have been found on Papaver
hybridum L. in Portugal (Campobasso et al., 1999).
A second arthropod that is a potential agent is the
cynipid wasp, Neaylax versicolor (Nieves-Aldrey).
The larvae of this wasp form galls in the fruits of Fumaria species. It is the only cynipid found in Fumaria
spp. (Nieves-Aldrey, 2003; Askew and Nieves-Aldrey,
2005) and has a distribution including southern France,
Greece and Spain. We did not find it in our survey as
most of the plants were just flowering, and the few
green fruits did not show any obvious wasp damage.
Conclusions
We have identified suitable organisms to be considered
for a biological control project in Australia. Any future
project will require relevant authorizations, and this
will be helped by a clearer understanding of the taxonomy, identification and importance of Fumaria species from Australia and the costbenefits of control in
cropping systems.
Acknowledgements
We thank the Grains Research and Development Corporation, the CRC for Australian Weed Management and
CSIRO for funding and facilities for this work. We also
thank Steve Walker for comments on the manuscript.
References
Askew, R.R. and Nieves-Aldrey, J.L. (2005) A new genus
and species of pteromalid (Hymenoptera, Chalcidoidea)
from Spain, parasitic in cynipid galls on Fumaria. Journal
of Natural History 39, 23312338.
Australian National Botanical Gardens (2007a) http://www.
anbg.gov.au/chah/apc/interim/Fumariaceae.pdf.
Australian National Botanical Gardens (2007b) http://www.
anbg.gov.au/chah/apc/interim/Papaveraceae.pdf.
Baldwin, B.J. (1977) Chemical weed control in oil-seed poppy
(Papaver somniferum). Australian Journal of Experimental Agriculture and Animal Husbandry 17, 837841.
Barton, J., Fowler, S.V., Gianotti, A.F., Winks, C.J., de Beurs,
M., Arnold, G.C. and Forrester, G. (2007) Successful biological control of mist flower (Ageratina riparia) in New
Zealand: agent establishment, impact and benefits to the
native flora. Biological Control 40, 370385.
Campobasso, G., Colonnelli, E., Knutson, L., Terragitti, G.
and Cristofaro, M. (1999) Wild plants and their associated
insects in the Palearctic region, primarily Europe and
Middle East. US Department of Agriculture, Agricultural
Research Service, ARS-147. 249 p.
Farr, D.F., Rossman, A.Y., Palm, M.E. and McCray, E.B. (2007)
Fungal Databases, Systematic Botany and Mycology Laboratory, ARS, USDA. Available at: http://nt.arsgrin.gov/
fungaldatabases/ (accessed March 15, 2007).
163
XII International Symposium on Biological Control of Weeds
Heap, I. (2007) The International Survey of Herbicide Resistant Weeds. Available at: www.weedscience.com (accessed 7 Jan 2007).
Hoffmann, A. (1954) Coloptres Curculionides (Deuxime
partie). Faune de France 59, 879897.
Holding, D. and Bowcher, A. (2004) Weeds in winter pulses
- integrated solutions. CRC Australian Weed Management
Technical Series 9, 111.
Laughlin J.C., Chung, N. and Beattie, B.M. (1998) Poppy
cultivation in Australia. In: Bernath, J. (ed.) Poppy
the Genus Papaver. Harwood Academic, Amsterdam,
pp. 249277.
Lemerle D., Yuan, T.H., Murray, G.M. and Morris, S. (1996)
Survey of weeds and diseases in cereal crops in the southern wheat belt of New South Wales. Australian Journal of
Experimental Agriculture 36, 545554.
Lidn, M. (1986) Synopsis of Fumarioideae (Papaveraceae)
with a monograph on the tribe Fumarieae. Opera Botanica 88, 5133.
Nieves-Aldrey, J.L. (2003) Descubrimiento de la agalla y
ciclo biolgico de Neaylax versicolor (Nieves-Aldrey)
(Hymenoptera, Cynipidae): primer registro de un cinpido
asociado a plantas papaverceas del gnero Fumaria. Boletin de la Sociedad Entomolgica Aragonesa 32, 111
114.
Norton, G.M. (2003) Understanding the success of fumitory
as a weed in Australia. PhD thesis. Charles Sturt University, Australia, 264 p.
Norton, G.M., Lemerle, D. and Pratley, J.E. (2004) Persistence of Fumaria densiflora DC. seed in the field. In:
Sindel, B.M. and Johnson, S.B. (eds) Proceedings of the
14th Australian Weeds Conference. Weed Science Society
of New South Wales, Australia, pp. 519522.
Spencer, R. (1997) Horticultural Flora of South-Eastern
Australia. Volume 2 Flowering Plants Dicotyledons Part
1. The Identification of Garden and Cultivated Plants.
University of New South Wales Press, Sydney.
Stevens, P.F. (2007) Angiosperm Phylogeny Website. Version
7. Available at: http://www.mobot.org/MOBOT/research/
APweb/ (accessed May 2006).
Tutin, T.G., Burges, N.A., Chater, A.O., Edmonson, J.R.,
Heywood, V.H., Moore, D.M., Valentine, D.H., Walters,
S.M. and Webb, D.A. (1993) Flora Europaea. Cambridge University Press vol.1: Psilotaceae to Platanaceae.
pp. 306311.
Vanky, K. (1994) European smut fungi. Gustav Fischer
Verlag, Stuttgart, 570 p.
Zundel, G.L. (1953) The Ustilaginales of the World. Pennsylvania State College School of Agriculture Department of
Botany Contribution 176, 1410.
164
Expanding classical biological control
of weeds with pathogens in India:
the way forward
P. Sreerama Kumar,1 R.J. Rabindra1 and C.A. Ellison2
Summary
Invasive alien weeds are a major constraint in agriculture, forestry and the environment in India. Classical biological control (CBC) of these exotic weeds through deliberate introduction of arthropods is
almost a century-old practice. History has recently been made with the successful introduction of the
first plant pathogen, Puccinia spegazzinii de Toni, against mikania weed (Mikania micrantha H.B.K.)
in India. With the mechanism in place for the importation, quarantining and release of pathogens, it
is envisaged that more introductions will be made in the future. The agent under immediate consideration is Puccinia abrupta Diet. and Holw. var. partheniicola (Jackson) Parmelee against parthenium
weed (Parthenium hysterophorus L.), generally considered as the worst terrestrial social weed in India. Down the line, other terrestrial weeds such as Chromolaena odorata (L.) R. King and H. Robinson and Lantana camara L. and aquatic species like Eichhornia crassipes (Martius) Solms-Laubach
could be targeted for pathogens. This article, besides presenting an overview of the research that has
gone into selection of candidate fungi for CBC of M. micrantha and P. hysterophorus, also analyses
the infrastructure and expertise requirements for further expanding the target list.
Keywords: Mikania micrantha, Parthenium hysterophorus, invasive alien weeds,
Indian infrastructure for biological control, rust pathogens.
Introduction
The impact of invasive alien weeds on agriculture, horticulture, forestry and the environment has been felt for
centuries in India. History is replete with examples of
reports and records of new and emerging or invading weeds. In India, traditionally an aggressive trading nation, movement of unwanted plants into and out
of the country was probably widespread before the
government-run quarantine system came into existence
with the promulgation of the Destructive Insects and
Pests Act in 1914.
Classical biological control (CBC) of these exotic weeds through deliberate introduction of natural
enemies, principally arthropods, has been in practice
for almost a century in India. Surprisingly, however,
pathogens have not received much attention in India
Project Directorate of Biological Control, PB 2491, H.A. Farm Post,
Hebbal, Bellary Road, Bangalore 560 024, India.
2
CABI Europe-UK, Bakeham Lane, Egham, Surrey TW20 9TY, UK.
Corresponding author: P. Sreerama Kumar <psreeramakumar@yahoo.
co.in>.
CAB International 2008
though the approach of manipulating plant pathogens
for suppressing troublesome weeds has been known to
science for more than a century (Wilson, 1969).
Nevertheless, India has now caught up with the
rest of the pioneers in the field by recently introducing a host-specific plant pathogen, Puccinia spegazzinii de Toni, against mikania weed (Mikania micrantha
H.B.K.), and thereby became the eighth country in the
world to practise CBC of weeds with plant pathogens
(Kumar et al., 2005; Ellison et al., 2006). It took more
than three decades for India to adopt this strategy since
the first successful use of an introduced pathogen elsewhere in the world, i.e. control of the skeleton weed,
Chondrilla juncea L., in south-east Australia with the
rust fungus, Puccinia chondrillina Bubak and Sydenham, was successfully implemented in the early 1970s
(Cullen et al., 1973).
A brief history of CBC of
weeds in India
India has traditionally been one of the early-adopters
of CBC of insect pests and weeds alike (see Table 1).
165
XII International Symposium on Biological Control of Weeds
Table 1.
Exotic natural enemies field-released for CBC of weeds in India.
Weed
(purported year of introduction)
Terrestrial weeds
Ageratina adenophora (Sprengel) R. King and H. Robinson
(1900)
Chromolaena odorata (L.)
King and H. Robinson (1914)
Lantana camara L.(1809)
Mikania micrantha H.B.K
(1914)
Opuntia spp. (unknown)
Parthenium hysterophorus L.
(1955)
Aquatic weeds
Eichhornia crassipes (Martius)
Solms-Laubach (1900)
Salvinia molesta Mitchell
(1955)
Agents released (year)a
Establishment in the field and impact
Procecidochares utilis Stone (1963)
Established - minimal control due to
parasitoids
Apion brunneonigrum Bguin Billecocq (1972)
Pareuchaetes pseudoinsulata Rego Barros (1973
and 1984)
Cecidochares connexa (Macquart) (2005)
Ophiomyia lantanae (Froggatt) (1921)
Teleonemia scrupulosa Stl (1941)
Not established
Recently reappeared
Diastema tigris Guene (1971)
Salbia haemorrhoidalis Guene (1971)
Octotoma scabripennis Gurin-Mneville (1972)
Uroplata girardi Pic (1972)
Puccinia spegazzinii de Toni (rust pathogen,
2005 Assam and 2006 Kerala)
Dactylopius ceylonicus (Green) against Opuntia
vulgaris Miller (1795)
Dactylopius confusus (Cockerell) against
O. vulgaris (1836)
Dactylopius opuntiae (Cockerell) against
Opuntia elatior Miller and Opuntia stricta
(Haworth) Haworth var. dillenii (Ker Gawler)
L. Benson (1926)
Zygogramma bicolorata Pallister (1984)
Neochetina eichhorniae Warner (1983)
Neochetina bruchi Hustache (1984)
Orthogalumna terebrantis Wallwork (1986)
Paulinia acuminata (Degeer) (1974)
Cyrtobagous salviniae Calder and Sands (1983)
Established - too early to assess
Established - not effective
Established - provides minimal
control
Not established
Not established
Established - not effective
Established - not effective
Established in Kerala - too early
Established and provided excellent
control
Not established
Established and provided complete
control of both species
Excellent control in some areas
Established - provides good to
variable control
Established - provides good to
variable control
Established - alone not very effective
Established - uncertain control
Established - spectacular control
All agents are arthropods except where indicated otherwise.
Although the first exceptional success in CBC of a
weed was in fact achieved with an erroneous introduction in 1795, that laid the foundation for further imports
of specific natural enemies as a result of the realization
of the potential of the approach. The agent in question
was the mealybug Dactylopius ceylonicus (Green) introduced from Brazil in place of Dactylopius coccus
Costa, the species intended for commercial production
of cochineal dye. D. ceylonicus dramatically brought
down the population of the prickly pear cactus, Opuntia vulgaris Miller, within 5 to 6 years in central and
north India (Singh, 1989).
The same episode gave a lesson on the significance
of host specificity as well. D. ceylonicus, when tried
against Opuntia stricta (Haworth) Haworth var. dillenii
(Ker Gawler) L. Benson [=Opuntia dillenii (Ker Gawler) Haworth], could not suppress the weed in south
India. Subsequently, the intentional introduction of a
North American species, Dactylopius opuntiae (Cock-
erell), from Sri Lanka in 1926 into India resulted in
impressive control of O. stricta and the related Opuntia
elatior Miller. This was the first successful intentional
use of an insect to control a weed in India, and more
than 40,000 ha area was thus cleared (Singh, 1989; Julien and Griffths, 1998).
The opuntia experience resulted in a series of introductions of phytophagous insects such as Ophiomyia
lantanae (Froggatt) (ex Mexico, via Hawaii in 1921)
against lantana weed (Lantana camara L.) Procecidochares utilis Stone (ex Mexico, via New Zealand
in 1963) against crofton weed [Ageratina adenophora
(Sprengel) R. King and H. Robinson] and Pareuchaetes
pseudoinsulata Rego Barros (ex Trinidad in 1973)
against Siam weed [Chromolaena odorata (L.) King
and H. Robinson] (Julien and Griffiths, 1998).
In the post-independence era, CBC became more
systematic and scientific with specific programmes
managed by the erstwhile Indian Station of the Com-
166
Expanding classical biological control of weeds with pathogens in India: the way forward
monwealth Institute of Biological Control (CIBC)
based in Bangalore. Later, country-specific programmes
came under the purview of the All-India Coordinated
Research Project (AICRP) on Biological Control of
Crop Pests and Weeds, which was launched in 1977.
This programme eventually came under the auspices of
the Project Directorate of Biological Control (PDBC),
which was formed in October 1993 under the Indian
Council of Agricultural Research (ICAR). For weed
control, only an insect and a plant pathogen have been
introduced since the formation of PDBC. Table 1 provides the list of natural enemies other than fish imported
for biological control of weeds in India and the current
status of their impact.
An overview of two pathogen-based
weed CBC projects in India
Target weed 1: Parthenium hysterophorus
The worst terrestrial social weed in India, Parthenium hysterophorus L., in general referred to as parthenium weed or congress grass, has been the primary
target for possible biological control using both insects
and pathogens.
Under the UK Department for International Development (DFID)-sponsored collaborative project between CABI Europe-UK (formerly CABI Bioscience)
and the ICAR, between 1996 and 2000, development
of both classical and the bioherbicide approaches were
given prominence. Research in India culminated in the
identification of a range of fungal pathogens of parthenium weed in Karnataka and Tamil Nadu in the south,
Madhya Pradesh, Haryana, Punjab, Himachal Pradesh,
Delhi and Uttar Pradesh in the north (Evans et al.,
2000). A few of these pathogens, despite possessing
some potential as mycoherbicides, did not warrant the
significant costs involved in further product development (Kumar and Evans, 2005).
In parallel research, two rust species were considered as options to be CBC agents, Puccinia melampodii Diet. and Holw. and Puccinia abrupta Diet. and
Holw. var. partheniicola (Jackson) Parmelee. Both of
these damaging rusts originate from Mexico and have
already been fully screened and released in Australia
Table 2.
for the control of parthenium weed. A comparison between these two rust species is given in Table 2.
P. abrupta var. partheniicola was found to severely
reduce both the vegetative growth of young plants and
the seed production of older plants under glasshouses
conditions (Evans, 1987a, b). This rust is also known to
be present in its exotic range, including India, though
the strains do not appear to be widespread or aggressive as they are in their native range (Kumar and Evans,
2005). In India, the rust was first reported from an elevated site (930 m) (Evans and Ellison, 1987, cited by
Parker et al., 1994).
Unconfirmed reports suggest that P. abrupta var.
partheniicola also occurs at lower elevations in India, but it is not a common pathogen of the weed (Kumar and Evans, 2005). A Mexican isolate (CABI no.
W1905) of this rust was, however, found to be virulent
and damaging to 12 P. hysterophorus collections from
across India (Evans et al., 2000).
Mexican isolates of P. melampodii (CABI nos.
W1496 and W1500) were also found to be highly
virulent towards the 12 Indian collections of P. hysterophorus producing high infection level and sporulation (Evans et al., 2000). This rust was considered for
introduction in India under the DFID project. However,
the ability of the rust to infect calendula (Calendula officinalis L.) under glasshouse conditions could not be
tolerated in India (as it was in Australia). Thus, fieldbased host-range testing was undertaken in Australia
to see if Indian varieties of calendula could be infected
under natural conditions. Unfortunately, they were
susceptible, and consequently the rust was not released
in India.
Target weed 2: Mikania micrantha
The neotropical vine mikania weed is an increasing
threat to natural and man-made forests as well as to
several agricultural and horticultural ecosystems in India. Although in its native range M. micrantha is rarely
weedy, in its exotic range, especially in south and
south-east Asian countries, it has become an intractable weed over the past several decades. Because of its
rapid growth habit, the plant, which can smother even
such hardy trees as teak, eucalyptus, rubber, oil palm
A comparison between two rust species used for the control of parthenium weed.
Puccinia melampodii
Microcyclic, autoecious rust-producing telia and
basidiospores
Puccinia abrupta var. partheniicola
Macrocyclic, autoecious rust-producing uredinia and telia in
the field. Pycnia and aecia have been induced in glasshouse
conditions (Evans, 1987b)
Winter rust - found predominantly in the semi-arid,
uplands of northern Mexico
Strain found in India (Evans and Ellison, 1987, cited in
Parker et al., 1994)
Highly host-specific
Summer rust - found in the humid and warmer lowland
plains of the Caribbean coast of Mexico (Evans, 1997)
Not present in India
Infects calendula
167
XII International Symposium on Biological Control of Weeds
and cocoa, has acquired one of its common names,
mile-a-minute weed. Whereas in north-east India it is
a great problem particularly in tea, it is an equally big
problem in plantation crops in south-west India. The
common control measures that are prevalent in tea
gardens and plantations are either cultural or chemical
means. These methods are expensive and impracticable.
Moreover, chemical herbicides can be very harmful to
the non-target plants, people and the environment.
Between 1996 and 1999, surveys were conducted
throughout the tropical and sub-tropical American native range of M. micrantha for pathogens having potential for CBC in India. Three microcyclic rust species,
P. spegazzinii, Dietelia portoricensis (Whetzel and Olive) Buritic and J.F. Hennen and Dietelia mesoamericana H.C. Evans and C.A. Ellison, were found to occur
in association with the plant (Evans and Ellison, 2005).
However, P. spegazzinii was selected for use as a CBC
agent in India after an extensive host-range screening
and studies on the environmental requirements for the
fungus (Ellison, 2001). This rust is a microcyclic, autoecious species that infects all aerial parts of the plant
causing necrosis and cankering, leading to plant death.
These studies were carried out in the CABI Europe-UK
quarantine in Ascot with funding from DFID, under a
collaborative project between CABI and research institutions in India between 1996 and 2000.
P. spegazzinii was imported into the National Containment-cum-Quarantine Facility (CQF) forTrans
genic Planting Material of the National Bureau of
Plant Genetic Resources (NBPGR) in New Delhi during 2003 and 2004. After establishing the fungus at
NBPGR, an additional host-specificity screening was
undertaken during 2004 and 2005. This involved 74
plant species, including 18 species that were earlier
tested in the UK, and reconfirmed the results from the
UK: the rust is totally specific to M. micrantha. At
NBPGR, the rust was found to be pathogenic to populations of mikania weed from several locations within
Kerala and Assam, which indicated that P. spegazzinii
has considerable potential as a CBC agent for mikania
weed in India.
A Supplementary Dossier on the additional hostspecificity tests provided the basis for obta ining the
permit for release of P. spegazzinii from the Plant Protection Advisor to the Government of India, Ministry of
Agriculture, in June 2005 (Kumar et al., 2005). Limited field releases of the rust have been made since 2005
Table 3.
in both Assam and Kerala (Ellison et al., 2006; Sankaran
et al., in this volume). India has thus become the eighth
country in the world to have released a plant pathogen
for the CBC of a weed. This is also the first time that
a fungal pathogen is being used as a CBC for mikania
weed. An estimate made in 2004 indicates that more
than 26 species of fungi originating from 15 different
countries have been used as CBC agents against more
than 26 weed species in seven countries (Barton, 2004).
The mikania weed CBC project in India has become
a flag-ship project. Other Asian countries, including
China, are following the Indian example for future
management of M. micrantha using P. spegazzinii.
The contrasts between the parthenium and mikania
weeds fungal-CBC projects are presented in Table 3.
Future strategies
Parthenium weed
India is continuing work on both off-the-shelf
pathogens and arthropod natural enemies to increase
the suppression of parthenium weed, already achieved
by Zygogramma bicolorata Pallister. The seed-feeding weevil, Smicronyx lutulentus Dietz, is planned to
be imported for the second time once the new quarantine facility being constructed at PDBC is functional.
In addition, project funding will be sought to undertake
strain selection studies of both rusts. For P. abrupta var.
partheniicola, the aim is to identify a virulent strain
that will be efficacious under Indian conditions and for
P. melampodii, other strains need to be tested to see if
there exists a strain that does not infect calendula.
There is also the option to investigate the potential
of new agents, for example the white smut fungus, Entyloma compositarum de Bary. This fungus, capable of
provoking severe leaf necrosis through the coalescing
of grey, senescing lesions, was found in upland, humid,
subtropical areas in Mexico and in semi-arid rangelands in Argentina by Evans (1997). Though this pathogen has not been evaluated as a CBC agent, it seems to
have considerable potential (Kumar and Evans, 2005),
especially in the light of the spectacular success of
the closely related species, Entyloma ageratinae Barreto and Evans, against the highly invasive upland and
cloud forest ecosystems weed mist flower [Ageratina
riparia (Regal) R. King and H. Robinson] in Hawaii
(Barreto and Evans, 1988), and more recently in both
Contrasts between the parthenium and mikania weeds fungal-CBC projects.
Parthenium weed project
Mikania weed project
Off-the-shelf agents available (previously released in Australia)
Arthropod CBC agent already released in India
Puccinia abrupta var. partheniicola already present in India
Non-target risk with P. melampodii
Importation of rusts put on hold
New agent identified and screened
No previous CBC attempt in India
No coevolved agents present in India
No non-target risk identified
Project led to release of rust in India
168
Expanding classical biological control of weeds with pathogens in India: the way forward
South Africa and New Zealand (Evans et al., 2001;
Barton, 2004).
Overall, it is considered that an integrated approach
is required for this weed, depending on the habitat it
is invading and the level of control needed. For example, in peri-urban areas, a high level of control is
necessary due to its toxicity to humans. Kumar and Evans (2005) stated: it is our considered opinion that
management of parthenium weed in India will only be
achieved through an integrated strategy based on biological control, specifically the classical approach with
the introduction of host-specific or coevolved natural
enemies from the plants centre of origin/diversity in the
Neotropics.
Mikania weed
The CBC project for mikania weed using P. spegazzinii is still in its early days since the release of the rust.
However, there is a clear need to improve the rust release strategy to elicit an epidemic of the rust so it is in
high enough concentrations to survive the dry season.
CABI has seven strains of the rust under quarantine in
the UK, and strains other than the one released in India
may prove to be more aggressive under field conditions. In the future, it may be worth considering previously untried rust pathogens such as D. portoricensis
and D. mesoamericana in new areas such as the Andaman and Nicobar Islands, or even in Assam and Kerala,
if P. spegazzinii does not give substantial control of the
weed. Finally, it should not be forgotten that substantial work has been undertaken on the arthropod natural
enemies of mikania weed; the CBC potential of many
were not fully investigated (Cock et al., 2000).
The way forward in India
Infrastructure
A brand-new quarantine facility of international
standards is being constructed on the PDBC campus in
Bangalore, funded by ICAR. This two-storey containment facility will allow for an increase in entomological work on introduced natural enemies at PDBC. It
will also have a pathogen-safe unit of level-4 containment (CL-4), to allow for the importation and screening
of pathogens for the control of weeds and other pests.
The upper floor or the Pathology quarantine cell has
two dedicated laboratories and a large greenhouse with
three individual bays complete with cement platforms
for plant propagation and handling. The entry to these
is routed through a shower room sandwiched between
two changing rooms to safeguard from entry of unwanted organisms and exit of organisms under quarantine.
International standard air and water handling systems
for quarantine facilities and equipment for waste disposal, viz. a double-ended autoclave and an incinerator,
are integrated into the overall configuration. This facility is due to be finished and operational by early 2008.
Although the original aim was to undertake the mikania weed work in the PDBC facility, the CL-4 CQF
on the NBPGR campus in New Delhi had to be used as
an interim facility by PDBC for both the quarantining
and host-specificity screening of P. spegazzinii. This facility also includes features such as outer and inner decontamination rooms provision for safe effluent treatment, a large incinerator and a dedicated generator as a
stand-by for uninterrupted power supply are available.
With funds from the DFID mikania weed project,
rust propagation units were constructed at the Assam
Agricultural University (AAU, Jorhat) and the Kerala
Forest Research Institute (KFRI, Peechi). These facilities have an area for plant propagation, an inoculation
chamber and an area for rust infected plants to develop
symptoms, prior to being placed in the field.
Expertise
One of the major outcomes of the DFID-funded
collaborative projects on P. hysterophorus and M. micrantha has been the development of local expertise
in handing and quarantining exotic weed pathogens for
biological control. The involvement of a host of institutes across the country in these projects has resulted in
invaluable know-how and do-how expertise in India.
Funding
Both the CBC projects on parthenium and mikania
weeds were funded by DFID. Similarly, Indian government agencies, including ICAR and the Department of
Biotechnology (DBT), have supported several research
projects on biological control of weeds with pathogens,
principally the mycoherbicide approach. Other international aid agencies operating in the Indian region may
provide funding in future, e.g. the Australian Centre for
International Agricultural Research (ACIAR). However, India is likely to have to look inward to national
and regional funding in the future, as its economy continues to grow and the country becomes less dependent
on external donor support.
Process
Although India has a long history of importing CBC
agents for the control of invasive alien weeds, until the
mikania weed project, all the natural enemies had been
arthropods or fish. The mikania weed project is considered a flagship project in India through which policy
and procedures for the import and release of fungal
CBC agents have been developed. This should enable
easy passage of future agents through the regulatory
system and into the field (Ellison et al., 2005).
Selection of future target weeds
Environmental weeds, both terrestrial and aquatic,
should be the main targets of control through the classical
169
XII International Symposium on Biological Control of Weeds
strategy (Kumar, 2005), although it is important to note
that most environmental weeds also impact on agroecosystems and/or agroforestry.
The use of fungal pathogens in weed CBC is a relatively young technology compared to employing arthropods. Thus, it is not surprising that most weed targets
for CBC, using fungal pathogens, are those where arthropods have not been effective, and this holds true for
the major invasive alien plants in India.
Ageratina adenophora
Crofton weed is not under control in India a gall fly
has been released, but a suite of native parasitoids have
prevented it from being effective. Other arthropods and
pathogens have been identified, in the native range of
the plant (Mexico), with good biocontrol potential, e.g.,
a lepidopterous and a curculionid stem borer and a rust
fungus Baeodromus eupatorii (Arthur) Arthur. These
agents have yet to be even established in the laboratory
and thus are unlikely to be considered in the near future
by India.
Chromolaena odorata
A suite of fungal natural enemies have been documented from C. odorata in its native range (Evans,
1987a; Barreto and Evans, 1994), some of which have
been partially assessed in the glasshouse (Elango et al.,
1993). However, the recent release of the stem gall fly,
Cecidochares connexa (Macquart), against this weed
in India (Bhumannavar et al., 2007) means that no further agents will be considered in the short term.
Cyperus rotundus
The grassy weed, purple nutsedge or nutgrass, C.
rotundus L., which is broadly considered to be the
worlds worst weed (Holm et al., 1977), hinders vegetable cultivation and is a huge problem in crops such as
maize and sugarcane across India. It has an Old World
centre of origin, possibly India, but the natural enemies
do not seem to exert sufficient pressure to keep it under
check in India, suggesting that its true centre of origin
may be elsewhere.
Currently, a specific sub-project within the ICARfunded network programme on Application of Microorganisms in Agriculture and Allied Sectors is being
undertaken to develop a mycoherbicide-based strategy
for its control. Puccinia canaliculata (Schwein) Lagerh (reported as Puccinia romagnoliana Maire and
Sacc.), which is widely prevalent across India during
winter, has been evaluated for augmentative use (Bedi
and Sokhi, 1994). Puccinia cyperi Arthur and Puccinia
cyperi-tegetiformis (Henn.) F. Kern, recorded in the
Neotropics, and the Uredo spp., reported from the Old
World (Barreto and Evans, 1995), are the pathogens
that need to be explored for within India or considered
for importation and evaluation.
Eichhornia crassipes
Water hyacinth [Eichhornia crassipes (Martius) Solms-Laubach] was successfully controlled in many areas
in India in the 1980s, but there have been some resurgence problems due in part to eutrophication of water
bodies. There is a wealth of unexploited pathogens in
its native range (Upper Amazon) (Evans, 1987a). Charudattan (1996) advocated research on the lifecycle,
host-range and biocontrol efficacy of the rust Uredo
eichhorniae Gonz. Frag. and Cif., which is found only
in South America, as a high-priority area. This will of
course necessitate significant investment, since a full
(5 years +) project will be required.
Invasive grasses
There are a number of grassy weeds of growing importance in India, for example littleseed canarygrass
(Phalaris minor Retz.), originating from the Mediterranean, is a serious weed in wheat, and is developing
resistance to certain herbicides. In addition, two species of Echinochloa, barnyard grass [Echinochloa crusgalli (L.) Beauv.] and jungle rice [Echinochloa colona
(L.) Link.], are problematic in rice. Grasses are notoriously difficult targets for CBC because of their close
relationship to staple crop species. However, some
biotrophic pathogens are known to be species specific,
within the Poaceae, and could be investigated for these
two genera. For example, smuts and/or rusts have been
recorded to infect species from both genera.
Lantana camara
Lantana weed has had a wide range of natural enemies released to control it throughout its invasive
range. Some have been more successful than others
at suppressing the weed (Broughton, 2000; Day et al.,
2003). India is yet to invest significantly in studying
the most successful agents and considering them for
introduction. This included a leaf rust, Prospodium tuberculatum (Speg.) Arthur (ex Brazil) that was released
in Australia in 2001 (Ellison et al., 2006; Thomas et al.,
2006) and Puccinia lantanae Farl. (ex Peru) that attacks
leaves, petioles and stems, which has been partially
screened and would seem to have potential for control
of lantana weed in India (Renteria and Ellison, 2004).
Mimosa diplotricha
The giant sensitive plant, Mimosa diplotricha C.
Wright, native to South America, has been selected by
India for CBC using an off-the-shelf agent Heteropsylla spinulosa Muddiman, Hodkinson and Hollis. This
psyllid is having a high impact on populations of this
weed in Papua New Guinea and Australia and is soon
to be imported into India (Kuniata and Korowi, 2001).
Pathogens, such as the rust Uredo mimosae-invisae
Vigas, may be worth future assessment.
170
Expanding classical biological control of weeds with pathogens in India: the way forward
Phanerogamic parasitic weeds
Both hemi-parasitic and holo-parasitic weeds continue to cause enormous losses in several crops in India.
Striga spp. interfere with cereals and legumes, whereas
Orobanche spp. parasitise roots of solanaceous (particularly, tobacco) and asteraceous (e.g. sunflower) crops.
Cuscuta spp. are equally problematic to ornamentals
and trees. The dipteran, Phytomyza orobanchia Kaltenbach, imported into India from the former Yugoslavia,
could not be field-released against Orobanche spp. because of problems in rearing of the fly. Both soil-borne
and air-borne fungal pathogens described on Striga and
Orobanche species are broad-range pathogens, excepting a few varieties or formae speciales of Fusarium
species (Kroschel and Mller-Stver, 2003). Highly
host-specific pathogens still need to be collected in
the centres of origin of these phanerogamic parasitic
weeds.
Conclusions
India has a fast-developing CBC of weeds programme
that has now branched out and embraced pathogens as
natural enemies. This strategy, therefore, finds an important place in the Perspective Plan (Vision 2025)
document of PDBC. The expertise, infrastructure and
now the precedent set by importing and releasing the
M. micrantha rust means that future projects are set to
roll, with the process fully in place. There is a wealth of
off-the-shelf arthropod and pathogen agents that could
be fast-tracked for release over the next decade, targeting the most noxious weed species in India. In addition, many more pathogen agents have been identified
that could be considered in the longer-term that require
full assessment. However, significant financial support
must be invested into this proven technology if the true
potential is to be realized.
Acknowledgements
The authors are grateful to Dr. Harry C. Evans, for
reviewing the manuscript. This publication is an output from the research projects funded by the United
Kingdom Department for International Development
(DFID) for the benefit of developing countries (R6695,
R6735 and R8228 Crop Protection Programme). The
views expressed are not necessarily those of DFID.
References
Barreto, R.W. and Evans, H.C. (1988) Taxonomy of a fungus
introduced into Hawaii for biological control of Ageratina
riparia (Eupatorieae: Compositae), with observations on
related weed pathogens. Transactions of the British Mycological Society 91, 8197.
Barreto, R.W. and Evans, H.C. (1994) The mycobiota of the
weed Chromolaena odorata in southern Brazil, with par-
ticular reference to fungal pathogens for biological control. Mycological Research 98, 11071116.
Barreto, R.W. and Evans, H.C. (1995) Mycobiota of the weed
Cyperus rotundus in the state of Rio de Janeiro, with an
elucidation of its associated Puccinia complex. Mycological Research 99, 407419.
Barton, J. (2004) How good are we at predicting the field
host-range of fungal pathogens used for classical biological control of weeds? Biological Control 31, 99122.
Bedi, J.S. and Sokhi, S.S. (1994) Puccinia romagnoliana rust
- a possible biological control agent for purple nutsedge,
Cyperus rotundus. Indian Journal of Plant Protection 22,
217218.
Bhumannavar, B.S., Ramani, S. and Rajeshwari, S.K. (2007)
Field release and impact of Cecidochares connexa (Macquart) (Diptera: Tephritidae) on Chromolaena odorata
(L.) King and Robinson. Journal of Biological Control
21, 5964.
Broughton, S. (2000) Review and evaluation of lantana biocontrol programs. Biological Control 17, 272286.
Charudattan, R. (1996) Pathogens for biological control of
water hyacinth. In: Strategies for Water Hyacinth Control
- Report of a Panel of Experts Meeting, 1114 September
1995, Florida, USA. Food and Agriculture Organization
of the United Nations, Rome, Italy, pp. 189196.
Cock, M.J.W., Ellison, C.A., Evans, H.C. and Ooi, P.A.C.
(2000) Can failure be turned into success for biological
control of mile-a-minute weed (Mikania micrantha)? In:
Spencer, N.R. (ed.) Proceedings of the X International
Symposium on Biological Control of Weeds. Boseman,
MT, pp. 155167.
Cullen, J.M., Kable, P.F. and Catt, M. (1973) Epidemic
spread of a rust imported for biological control. Nature
244, 462464.
Day, M.D., Wiley, C.J., Playford, J. and Zalucki, M.P. (2003)
Lantana: Current Management Status and Future Prospects. ACIAR, Canberra, Australia. 128 p.
Elango, D.E., Holden, A.N.G. and Prior, C. (1993) The potential of plant pathogens collected in Trinidad for biological control of Chromolaena odorata (L.) King and
Robinson. International Journal of Pest Management 39,
393396.
Ellison, C.A. (2001) Classical biological control of Mikania
micrantha. In: Sankaran, K.V., Murphy, S.T. and Evans,
H.C. (eds) Alien Weeds in Moist Tropical Zones, Banes
and Benefits, Workshop Proceedings, 24 November
1999. Kerala Forest Research Institute, Peechi, India and
CABI Bioscience, UK Centre (Ascot), Berkshire, UK,
pp. 131138.
Ellison, C.A., Murphy, S.T. and Rabindra, R.J. (2005) Facilitating access for developing countries to invasive alien
plant classical biocontrol technologies: the Indian experience. In: Harris, D., Richards, J.I., Silverside, P., Ward,
A.F. and Witcombe, J.R. (eds) Aspects of Applied Biology
75, Pathways Out of Poverty. Association of Applied Biologists, c/o Warwick HRI, Wellesbourne, Warwick, UK,
pp. 7180.
Ellison, C.A., Puzari, K.C., Kumar, P.S., Usha Dev, Sankaran,
K.V., Rabindra, R.J. and Murphy, S.T. (2006) Sustainable
control of Mikania micrantha - implementing a classical
biological control strategy in India using the rust fungus
Puccinia spegazzinii. In: Program and Abstracts, Seventh International Workshop on Biological Control and
171
XII International Symposium on Biological Control of Weeds
Management of Chromolaena odorata and Mikania micrantha, 1215 September 2006. National Pingtung University of Science and Technology, Pingtung, Taiwan,
Republic of China, pp. 34.
Evans, H.C. (1987a) Fungal pathogens of some subtropical
and tropical weeds and the possibilities for biological control. Biocontrol News and Information 8, 730.
Evans, H.C. (1987b) Life-cycle of Puccinia abrupta var.
partheniicola, a potential biological control agent of
Parthenium hysterophorus. Transactions of the British
Mycological Society 88, 105111.
Evans, H.C. (1997) The potential of neotropical fungal pathogens as classical biological control agents for management of Parthenium hysterophorus L. In: Mahadevappa,
M. and Patil V.C. (eds) Proceedings of the First International Conference on Parthenium Management, 68 October 1997, vol. 1. University of Agricultural Sciences,
Dharwad, India, pp. 5562.
Evans, H.C. and Ellison, C.A. (2005) The biology and taxonomy of rust fungi associated with the neotropical vine
Mikania micrantha, a major invasive weed in Asia. Mycologia 97, 935947.
Evans, H.C., Greaves, M.P. and Watson, A.K. (2001) Fungal
biocontrol agents of weeds. In: Butt, T.M., Jackson, C.W.
and Magan, N. (eds) Fungi as Biocontrol Agents. CABI
Publishing, Wallingford, UK, pp.169192.
Evans, H.C., Seier, M., Harvey, J., Djeddour, D., Aneja, K.R.,
Doraiswamy, S., Kauraw, L.P., Singh, S.P. and Kumar,
P.S. (2000) Final technical report: Developing strategies
for the control of parthenium weed in India using fungal
pathogens. Unpublished report submitted to the Department for International Development, UK. 191 p.
Holm, L.G., Plucknett, D.L., Pancho, J.V. and Herberger, J.P.
(1977) The Worlds Worst Weeds, Distribution and Biology. University Press of Hawaii, Manoa, Honolulu, HI.
609 p.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A World Catalogue of Agents and their Target
Weeds. Fourth Edition. CABI Publishing, Wallingford,
UK. 223 p.
Kroschel, J. and Mller-Stver, D. (2003) Biological control
of root parasitic weeds with plant pathogens. In: Inderjit
(ed.) Weed Biology and Management. Kluwer Academic
Publishers, Amsterdam, The Netherlands, pp. 423438.
Kumar, P.S. (2005) Scope of fungal pathogens in weed control in India. In: Rabindra, R.J., Hussaini, S.S. and Ra-
manujam, B. (eds) Microbial Biopesticide Formulations
and Application - Proceedings of the ICAR-CABI Workshop on Microbial Biopesticide Formulations and Application, 913 December 2002. Technical Document No.
55, Project Directorate of Biological Control, Bangalore,
India, pp. 203211.
Kumar, P.S. and Evans, H.C. (2005) The mycobiota of
Parthenium hysterophorus in its native and exotic ranges:
opportunities for biological control in India. In: Prasad,
T.V.R., Nanjappa, H.V., Devendra, R., Manjunath, A.,
Subramanya, Chandrashekar, S.C., Kiran Kumar, V.K.,
Jayaram, K.A. and Prabhakara Setty, T.K. (eds) Proceedings of the Second International Conference on Parthenium Management, 57 December 2005. University of
Agricultural Sciences, Bangalore, India, pp. 107113.
Kumar, P.S., Rabindra, R.J., Usha Dev, Puzari, K.C., Sankaran, K.V., Khetarpal, R.K., Ellison, C.A. and Murphy,
S.T. (2005) India to release the first fungal pathogen for
the classical biological control of a weed. Biocontrol News
and Information 26, 71N72N.
Kuniata, L.S. and Korowi, K.T. (2001) Biological control
of the giant sensitive plant with Heteropsylla spinulosa
(Hom.: Psyllidae) in Papua New Guinea. In: Proceedings
of the VI Workshop for Tropical Agricultural Entomology,
Darwin, Australia. Technical Bulletin Department of Primary Industry and Fisheries, Northern Territory of Australia No. 288, pp.145151.
Parker, A., Holden, A.N.G. and Tomley, A.J. (1994) Host
specificity testing and assessment of the pathogenicity of
the rust, Puccinia abrupta var. partheniicola, as a biological control agent of parthenium weed (Parthenium hysterophorus). Plant Pathology 43, 116.
Renteria B., J.L. and Ellison, C.A. (2004) Potential biological
control of Lantana camara in the Galapagos using the rust
Puccinia lantanae. SIDA 21, 10091017.
Singh, S.P. (1989) Biological Suppression of Weeds. Technical Bulletin No. 1, Biological Control Centre, National
Centre for Integrated Pest Management, Bangalore, India,
27 p.
Thomas, S.E., Ellison, C.A. and Tomley, A.J. (2006) Studies on the rust Prospodium tuberculatum, a new classical biological control agent released against the invasive
alien weed Lantana camara in Australia, II: host range.
Australasian Plant Pathology 35, 321328.
Wilson, C.L. (1969) Use of plant pathogens in weed control.
Annual Review of Phytopathology 7, 411434.
172
Explorations in Central Asia and
Mediterranean basin to select biological
control agents for Salsola tragus
F. Lecce,1 A. Paolini,1 C. Tronci,1 L. Gltekin,2 F. Di Cristina,1
B.A. Korotyaev,3 E. Colonnelli,4 M. Cristofaro5 and L. Smith6
Summary
Russian thistle, Salsola tragus L., (Chenopodiaceae), a troublesome weed complex in the drier regions
of western North America, is native to Central Asia and widely distributed throughout the Palaearctic
Region. Since 2003, several exploration and survey trips to discover new potential biological control
agents for the weed were carried out in Italy, Russia, Turkey, Tunisia, Kazakhstan, Greece, Egypt and
Morocco. Twenty-five arthropod species (one gall midge, one heteropteran, two flea beetles, three
moths and 18 weevils) and two fungi were preliminarily selected during the surveys. Among them, four
arthropod species (two weevils and two moths) were selected for further biology and host-specificity
studies. At the moment, the most promising among these is the stem-boring weevil Anthypurinus
biimpressus Brisout. Preliminary host range no-choice tests and life cycle observations showed that
this species is restricted for feeding, oviposition and development to Russian thistle.
Keywords: Salsola, Russian thistle, foreign exploration, tumbleweed, natural enemies.
Introduction
Russian thistle or tumbleweed, Salsola tragus L.,
(Chenopodiaceae) is an invasive alien weed originating
from Central Asia and is the target of classical biological control in the United States (Goeden and Pemberton, 1995; Pitcairn, 2004). This plant has commonly
also been called Salsola australis R. Br., Salsola iberica
(Sennen and Pav) Botsch ex Czereparov, Salsola kali L.
and Salsola pestifer A. Nelson, and many synonyms occur in the literature (Mosyakin, 1996; Rilke, 1999). The
taxonomy of Russian thistle is complicated because of
high morphological variability of species in the genus
and the occurrence of hybrids and polyploids; however,
Biotechnology and Biological Control Agency, Via del Bosco 10, 00060
Sacrofano, Rome, Italy.
2
Atatrk University, Faculty of Agriculture, Plant Protection Department, 25240 TR Erzurum, Turkey.
3
Zoological Institute, Russian Academy of Science, 199034, St. Petersburg, Russia.
4
Via delle Giunchiglie 56, Rome, Italy.
5
ENEA C.R. Casaccia, s.p. 25, Via Anguillarese 301, 00123 S. Maria di
Galeria (RM), Italy.
6
USDA-ARS, 800 Buchanan Street, Albany, CA 94710, USA.
Corresponding author: F. Lecce <[email protected]>.
CAB International 2008
1
biochemical and molecular analyses are beginning to
clarify the taxonomy. Similar species that are also invasive in North America include Salsola paulsenii Litv.,
Salsola collina Pallas, Salsola type B and various hybrids (Ryan and Ayres, 2000; Akers et al., 2003; Gaskin
et al., unpublished data). S. kali L also occurs in North
America but is limited to seashores, primarily on the
Atlantic coast (Mosyakin, 1996). In Eurasia, this species occurs primarily on seaside beaches of the Mediterranean and Atlantic coasts (Rilke, 1999), whereas S.
tragus occurs primarily inland. All of these species are
taxonomically closely related and have been placed in
the Salsola section kali, sub-section kali, which is historically distributed across Eurasia (Rilke, 1999).
During a survey of genetic variability of S. tragus in
California, a new variant, called type B was discovered (Ryan and Ayres, 2000). Subsequent searches for
the geographic origin of this population have revealed
the same species only in South Africa and Australia.
Until now, we have not been able to locate any specimens of Eurasian origin. Morphological and biochemical analyses support the restoration of this species to S.
australis R. Brown (F. Hrusa, California Department
of Food and Agriculture, personal communication).
The origin of type B is important to biological control
because it differs significantly from S. tragus with re-
173
XII International Symposium on Biological Control of Weeds
spect to susceptibility to two prospective agents: a gallinducing midge, Desertovellum stackelbergi Mamaev
(Diptera: Cecidomyiidae; Sobhian et al., 2003a,b) and
a fungus, Colletotrichum gloeosporoides (Penz.) Penz.
and Sacc. in Penz (Bruckart et al., 2004). However,
apart from type B, the other species all appear to
originate from Eurasia and can be attacked by some of
the same prospective biological control agents (Smith,
2005).
Foreign explorations to look for prospective biological control agents were previously conducted in
Afghanistan, Pakistan, Egypt, Turkey, Uzbekistan,
China and on the Mediterranean coast of France (Baloch and Mushtaque, 1973; Goeden, 1973; Hafez et al.,
1978; Sobhian et al., 1999; Hasan et al., 2001). As a
consequence, two species of Coleophorid moths, Coleophora klimeschiella Toll, 1952, and Coleophora
parthenica Meyr., 1891, were introduced in the 1970s
to California and nearby states (Hawkes and Mayfield,
1976, 1978). These became widespread, but predators
and parasites prevent them from becoming abundant
enough to control the weed (Goeden and Pemberton,
1995). Other species that have been evaluated but
were rejected for lack of specificity include Lixus incanescens Boheman 1836 [= L. salsobe Becker, 1867]
(Coleoptera: Curculionidae), Piesma salsolae (Becker,
1867) (Hemiptera: Piesmatidae) and the fungus Colletotrichum gloeosporioides (Penzig) Penzig et Saccardo (Sobhian et al., 2003a,b; Bruckart et al., 2004).
The blister mite Aceria salsolae DeLillo and Sobhian,
1996 (Acari: Eriophyidae) has been petitioned for permit to release (Smith, 2005), and Gymnancyla canella
[Denis and Schiffermller], 1775 (Lepidoptera: PyraliTable 1.
dae) is undergoing the final stage of evaluation before
petitioning for release (Table 1). No other prospective agents worth evaluating were known. In the latest years, explorations were mainly targeted to Central
Asia, the closest region to the probable centre of origin
of S. tragus, and North Africa, which has the closest
climatic match to the San Joaquin Valley of California,
the locality where S. tragus is most problematic.
Methods
Field surveys
In 2004, the Biotechnology and Biological Control
Agency (BBCA) began involvement in a program of
explorations for Russian thistle natural enemies in the
Mediterranean Basin and in Western and Central Asia.
During the past 3 years, several surveys were carried
out in Italy, North African countries, Turkey and Kazakhstan during late spring until late summer.
Italy
Bibliography and herbarium surveys reported mainly
S. kali in Italy: this Russian thistle species is found
in this country only on undisturbed sandy sea shores.
The weed is very common, mainly along the Central,
Southern and island sandy beaches.
North Africa
Similar habitat and weed species have been recorded
in the three North African countries that we have inspected: Tunisia, Morocco and Egypt. All of them have
Status of prospective biological control agents of Russian thistle.
Taxonomic name
Evaluated species
Aceria salsolae (Acari: Eriophyidae)
Gymnancyla canella (Lepidoptera:
Pyralidae)
Lixus incanescens [=salsolae]
(Coleoptera: Curculionidae)
Piesma salsolae (Hemiptera:
Piesmatidae)
Colletotrichum gloeosporioides
(Phyllachorales: Phyllacoraceae)
Uromyces salsolae (Uredinales:
Pucciniaceae)
Newly tested species
Anthypurinus biimpressus
(Coleoptera: Curculionidae)
Baris przewalskyi (Coleoptera:
Curculionidae)
Philernus sp. (Coleoptera:
Curculionidae)
Common name
Current information
Blister mite
The mite attacks developing tips. Petition approved by
TAG, permit submitted to Animal and Plant Health
Inspection Service (APHIS)
Caterpillar feeds on seeds and young branch tips. Host
specificity testing almost completed
Adults feed on many plants in choice test at Montpellier,
France (Sobhian et al., 2003a,b). Rejected
Develops on beets in no-choice lab test at Montpellier,
France (R. Sobhian, personal communication). Rejected
More damaging to Russian thistle type A than to type B
(Bruckart et al., 2004). Being evaluated by W. Bruckart,
US Department of Agriculture-Agricultural Research
Service (USDA-ARS), Maryland
Damages Russian thistle type A (Hasan et al., 2001). Being
evaluated by W. Bruckart, USDA-ARS, Maryland
Seed and stem moth
Stem weevil
Plant bug
Fungus
Rust fungus
Jumping weevil
Found in Tunisia in 2004. Larvae and adults feed on leaves.
Biology is unknown
Abundant on Salsola sp. in Kazakhstan in 2004. Biology is
unknown
Found in Kazakhstan in 2004. Probably monophagous
Weevil
Weevil
174
Explorations in Central Asia and Mediterranean basin to select biological control agents for Salsola tragus
a peculiar combination of sandy beaches and sandy
deserts. Among them, only in Tunisia was the target
weed recorded in oases and in sandy areas at the beginning of the desert, while Russian thistle occurs in the
other two countries only on the sandy dunes along the
sea shore.
Turkey
S. kali and S. tragus are both present in Turkey, but
they occur in different areas and habitats. The first is
always associated with sandy beaches along the Mediterranean coast, while the latter occurs in more rocky
habitats, especially in the Central and Eastern Anatolia. Furthermore, climatic conditions are very different,
with a classic Mediterranean climate along the coast
and a continental climate in the interior regions.
Kazakhstan
Explorations were carried out mainly in the southeastern part of the country, in the region north of
Almaty. Russian thistle was relatively common in
disturbed areas, mainly along roads, in dense populations. Scattered populations were also found in wild
fields. In both cases, the weed was associated only
with sandy soils.
During the field explorations, the field sites where
Russian thistle was present were recorded, adult insects were sampled, plants were dissected and eggs
and immature larvae transferred into artificial diet for
rearing. In addition, plant parts (leaves, shoots, roots)
with immature stages or signs of mite or pathogen attack were collected and taken back to the laboratory,
where immature stages were reared to adult and sent,
together with field-collected adults, to taxonomists for
identification.
Results and Discussion
Potential biological control agents
Tunisia: Anthypurinus biimpressus Brisout 1869 (Coleoptera, Cu culionidae): This stem- and leaf-mining weevil is recorded from one area in Tunisia (near Gabes)
and was identified by one of us (E. Colonnelli). It is a
univoltine species, and larvae attack the plant in May
and June during the early phenological stages. From
field observations, the insect has gregarious behaviour:
the adults were found in 2004 and 2006 at two different
sites near Gabes in large numbers on single individual
plants. Almost all the Russian thistle plants at the sites
were heavily attacked.
Broconius biscrensis, Capiomont, 1874 (Coleoptera,
Curculionidae): A weevil has been reported on Russian
thistle in Central Tunisia according to a record in E.
Colonnelli collection. Additional surveys are planned
to collect and rear this weevil.
Kazakhstan: Baris przewalskyi Zaslavskii 1956 (Coleoptera, Curculionidae): This root-mining weevil was
collected at one site in Kazakhstan (near Ushtobe). It
has been identified by B. Korotyaev. The species was
found as large number of adults resting and feeding
on young shoots and around the crown of young Russian thistle plants. Unfortunately, the species has been
found at one site only, together with three other oligophagous Baris spp. (Baris sulcata Boheman, 1836,
Baris convexicollis Boheman, 1836 and Baris memnonia Boheman 1844), all very similar in appearance. In
addition, Cosmobaris scolopocea (Germar, 1826) and
Elasmobaris signetera (Faust, 1821) were found at the
same spot.
D. stackelbergi Mamaev (Diptera, Cecidomyiidae):
This stem-galling midge was found for the first time in
Uzbekistan by R. Sobhian (EBCL). During 2004, two
populations were recorded at two sites in south-eastern
Kazakhstan. This family is reported to be extremely
specific. Preliminary tests carried out by Sobhian in
Uzbekistan confirmed the narrow host range of the species. Additional tests carried out at the USDA, Albany,
CA, showed that gall formation may depend on the
presence of an unknown symbiotic fungus (Biscet and
Borkent, 1988).
Other biological candidate agents
Kazakhstan: A weevil, Philernus sp., was collected
in south-eastern Kazakhstan during 2004. Its biology
is unknown, but preliminary bibliography surveys do
not report any weevil of this genus associated with any
crop plant.
During 2006, two populations of a root-boring moth
were found in south-eastern Kazakhstan. The larvae
were found in silken tunnels on the side of the roots,
and feeding was observed on the external part of the
root, near the place where the tunnel was affixed with
the root system.
Italy: Several moths are reported to be closely related
to Russian thistle (S. kali) in the Central Mediterranean
Basin (A. Zilli, Museo Civico di Zoologia, Rome, Italy, personal communication). In particular, Discestra
sodae Rambur, 1829 and Discestra stigmosa Christoph,
1887 (Noctuidae) are considered having a narrow host
range within the genus Salsola and occur in Sardinia,
Central Italian coasts, and Sicily. Similar distribution
was also reported for several Cardepia spp., such as
Cardepia sociabilis Graslin, 1860, Cardepia deserticola Rotschild, 1920, and Cardepia hartigi Parenzon,
1988. Moreover, a new species of Gymnancila sp.
prope canella was found near Cefal, northern Sicily.
A population of G. canella from southern France is currently being evaluated for specificity (Sobhian, 2000;
Smith et al., 2007). Finally, noctuid moths of the genera Lacanobia Billberg, 1820 and Pseudohadena Alpheralay, 1889, even if not very common, are reported
as oligophagous for the genus Salsola.
175
XII International Symposium on Biological Control of Weeds
Preliminary screening of new potential
biological control agents
Tunisia: Lixus sp.: This weevil from eastern Tunisia is
probably L. incanescens, already tested by R. Sobhian
(EBCL) and shown to be oligophagous. We carried out
a quick screening on newly collected live specimen in
no-choice conditions. Feeding and in some cases oviposition have been recorded on all Chenopodiaceae
species tested (common beet, sugar beet, Russian thistle). The insect was rejected.
A. biimpressus Brisout: This weevil from eastern Tunisia was found for the first time on Salsola kali near
Gabs during late May 2004 (spring) and identified as
a probable monophagous species. Larvae and/or adults
feed on young leaves caused severe damage. Oviposition and larval development were obtained in the
laboratory from adults collected in the field in spring,
although larvae could not complete pupation. Preliminary laboratory observations showed that larvae appear
to be external feeders; this is uncommon for weevils.
The insect is likely univoltine because in 2006 at the
Gabs site, no larvae or adults were seen in late June on
plants that had been attacked in May. The generation
collected in September (autumn) behaved differently
from the one collected in spring. The autumn collected
insects fed on host plant and mated but did not oviposit.
Preliminary host range tests, carried out during 2006,
tested four crop plants within the family Chenopodiaceae and indicate that the host range does not include
species in the subfamily Chenopodioideae (e.g. Chenopodium album L., beets and spinach).
Eastern Kazakhstan: A wide range of arthropod
fauna has been found associated with the target weed
in Kazakhstan. The following species are likely
monophagous: the weevil Philernus sp., the gall midge
D. stackelbergii Mamaev and the weevil B. przewalskyi Zaslavskii. The following weevil species are
likely oligophagous: Lixus rubicundus Zoubkoff 1833,
Lixus polylineatus Petri 1900, Lixus scabricollis Boheman 1843, C. scolopacea, B. sulcata, E. signifera,
B. convexicollis, B. memnonia Temnorhinus elongatus
(Gebler 1845).
Acknowledgements
We are grateful to Alberto Zilli, Museo Civico di Zoologia, Rome, Italy.
References
Akers, P., Pitcairn, M.J., Hrusa, F. and Ryan, F. (2003) Identification and mapping of Russian thistle (Salsola tragus)
and its types. In: Woods, D.M. (ed.) Biological Control
Program Annual Summary, 2002. California Department
of Food and Agriculture, Plant Health and Pest Prevention
Services, Sacramento, California, USA, pp. 5257.
Baloch, G.M. and Mushtaque, M. (1973) Insects associated
with Halogeton and Salsola in Pakistan with notes on
the biology, ecology and host specificity of the important
enemies. In: Dunn, P.H. (ed.) Proceedings of the 2nd International Symposium on Biological Control of Weeds,
Rome. Commonwealth Agricultural Bureaux., Slough,
UK, pp. 103113.
Bisset, J. and Borkent, A. (1988) Ambrosia galls: the significance of fungal nutrition in the evolution of the Cecidomyiidae (Diptera). In: Pirozynski, K.A. and Hawksworth,
D.C. Coevolution of Fungi with Plants and Animals. Academic Press, London, pp. 203225.
Bruckart, W., Cavin, C., Vajna, L., Schwarczinger, I. and
Ryan, F.J. (2004) Differential susceptibility of Russian
thistle accessions to Colletotrichum gloeosporoides. Biological Control 30, 306311.
Goeden, R.D. (1973) Phytophagous insects found on Salsola
in Turkey during exploration for biological weed control
agents for California. Entomophaga 18, 439448.
Goeden, R.D. and Pemberton, R.W. (1995) Russian thistle.
In: Nechols, J.R., Andres, L.A., Beardsley, J.W., Goeden, R.D. and Jackson, C.G. (eds) Biological Control in the Western United States: Accomplishments
and benefits of regional research project W-84, 1964
1989. University of California, Division of Agriculture
and Natural Resources, Oakland. Publ. 3361. USA,
pp. 276280.
Hafez, M., Fayad, Y.H. and Sarhan, A.A. (1978) Coleophora parthenica Meyrick (Lepidoptera, Coleophoridae) in
Egypt, a potential agent for the biological control of the
noxious thistle, Salsola kali L. (Chenopodiaceae). Protection Ecology 1, 3344.
Hasan, S., Sobhian, R. and Herard, F. (2001) Biology, impact
and preliminary host-specificity testing of the rust fungus,
Uromyces salsolae, a potential biological control agent
for Salsola kali in the USA. Biocontrol Science and Technology 11, 677689.
Hawkes, R.B. and Mayfield, A. (1976) Host specificity and
biological studies of Coleophora parthenica Meyrick, an
insect for the biological control of Russian thistle. A Commemorative Volume in Entomology, Department of Entomology, University of Idaho, pp. 3743.
Hawkes, R.B. and Mayfield, A. (1978) Coleophora klimeschiella, biological control agent for Russian thistle:
host specificity testing. Environmental Entomology 7,
257261.
Mosyakin, S.L. (1996) A taxonomic synopsis of the genus
Salsola L. (Chenopodiaceae) in North America. Annals of
the Missouri Botanical Garden 83, 387395.
Pitcairn, M.J. (2004) Russian thistle. In: Coombs, E.M.,
Clark, J.K., Piper, G.L. and Cofrancesco, A.F., Jr. (eds)
Biological Control of Invasive Plants in the United States.
Oregon State University Press, pp. 304310.
Rilke, S. (1999) Revision der Sektion Salsola S.L. der Gattung Salsola (Chenopodiaceae). Bibliotheca Botanica
149, 1190.
Ryan, F.J. and Ayres, D.R. (2000) Molecular markers indicate
two cryptic, genetically divergent populations of Russian
thistle (Salsola tragus) in California. Journal of Botany
78, 5967.
176
Explorations in Central Asia and Mediterranean basin to select biological control agents for Salsola tragus
Sobhian R. (2000) Biological control of Russian thistle. In:
Spencer, N.R. (ed) Proceedings of the X International
Symposium on Biological Control of Weeds. United States
Department of Agriculture, Agricultural Research Service, Sidney, MT, and Montana State University - Bozeman, Bozeman, MT, USA, p. 247.
Sobhian, R., Tun, I. and Erler, F. (1999) Preliminary studies
on the biology and host specificity of Aceria salsolae DeLillo and Sobhian (Acari, Eriophyidae) and Lixus salsolae
Becker (Col., Curculionidae), two candidates for biological control of Salsola kali. Journal of Applied Entomology
123, 205209.
Sobhian, R., Fumanal, B. and Pitcairn, M. (2003a) Observations on the host specificity and biology of Lixus salsolae
(Coleoptera: Curculionidae), a potential biological control
agent of Russian thistle, Salsola tragus (Chenopodiaceae)
in North America. Journal of Applied Entomology 127,
322324.
Sobhian, R., Ryan, F.J., Khamraev, A., Pitcairn, M.J. and Bell,
D.E. (2003b) DNA phenotyping to find a natural enemy in
Uzbekistan for California biotypes of Salsola tragus L.
Biological Control 28, 222228.
Smith, L. (2005) Host plant specificity and potential impact of
Aceria salsolae (Acari: Eriophyidae), an agent proposed
for biological control of Russian thistle (Salsola tragus).
Biological Control 34, 8392.
Smith, L., Sobhian, R. and Cristofaro, M. (2007) Prospects
for biological control of Russian thistle (tumbleweed).
In: Proceedings of the California Invasive Plant Council
Symposium, Vol. 10, 2006, pp. 7476.
177
Eriophyoid mites on Centaurea solstitialis
in the Mediterranean area
R. Monfreda,1 E. de Lillo1 and M. Cristofaro2
Summary
Yellow star thistle (YST), Centaurea solstitialis L. (Asteraceae), is a weedy annual plant, native to
the Mediterranean area that infests rangelands, pastures and grasslands in the northern United States.
In the search for biological control agents, an eriophyoid mite capable of significant impact on weed
density and reproduction was studied. This research was focussed on: (a) geographical distribution of
Aceria solstitialis de Lillo et al. (Acari: Eriophyidae) in the YST native area; (b) dynamic study of the
eriophyoid population on YST in Apulia (Italy). In the first case, symptomatic plants were collected
in Bulgaria, Greece, Italy and Turkey. In the second case, four sites were selected in Apulia, and periodical samplings were undertaken on 20 plants from February to August 2004 to 2006. Four different
phenological categories of samples were collected separately from each plant: new rosette leaves,
mature rosette leaves, stems and flower heads. This study ascertained the presence of A. solstitialis in
the sampled localities of Mediterranean area, together with other mite species. In Apulia, ten species
were detected on YST, and some of them could be related to neighbouring plants. A. solstitialis was
detected only from May to August, and its density was highest at stem level, but it seems to be always
low and insufficient to provide a significant reduction of weed seed production in the studied area.
Keywords: weeds, Eriophyoidea, population dynamics, geographical distribution.
Introduction
Yellow star thistle (YST), Centaurea solstitialis L., is
an annual species belonging to the Asteraceae family
which has become a weed of extreme importance in
several states of the USA, with the most serious infestations in Arizona, California, Idaho, Oregon and
Washington (Maddox et al., 1985). Native to the Mediterranean region of Europe and Southern Eurasia, this
herbaceous plant probably arrived in the USA by way
of contaminated alfalfa seeds (Medicago sativa L.)
during the mid-1800s (Roch and Talbott, 1986; Roch and Roch, 1991) and actually represents a serious
weed of pastures, rangelands, abandoned croplands,
natural areas and roadsides (Lass et al., 1996).
YST is an herbaceous winter annual plant propagating only by seeds that usually germinate in autumn or
winter, depending on rainfall. In California, the transition from seedling to rosette occurs in late winter to
Di.B.C.A.Sezione Entomologia e Zoologia, Universit degli Studi di
Bari, via Amendola 165/a-70126 Bari, Italy.
2
ENEA BIOTEC-BBCA, Rome, Italy.
Corresponding author: R. Monfreda <[email protected]>.
CAB International 2008
1
late spring. The plant bolts during May and June to produce an erect and branched stem 20 to 75 cm in height,
develops bud in mid-June to early July and flowers in
mid-July to September. The spread and survival of this
weed depend on seed production, and several thousand
achenes may be produced per plant under optimal conditions (Maddox, 1981).
Several attempts have been made to find an efficient
biological control agent (Piper, 2001). Recently, attention has been paid to mites (Acari: Eriophyoidea) as
candidates for YST control (de Lillo et al., 2003). Eriophyoid mites are considered extremely important for
biological control of weeds because of their ability to
damage the target plant sufficiently to reduce its impact, their high degree of host-specificity, their capacity
for rapid population increase and their efficient dis
persal (Briese and Cullen, 2001).
Two eriophyoids have been reported on C. solstitialis: Aceria solcentaureae de Lillo et al. and A. solstitialis de Lillo et al. (de Lillo et al., 2003) with apparently
similar effects on stem and flower head development.
In particular, infested plants show reduced growth,
stem apex and flower heads remain green and fresh
during the hot-dry season, flower heads are less spiny
and seedhead appear to be small in size.
178
Eriophyoid mites on Centaurea solstitialis in the Mediterranean area
The aim of this research was to give a first contribution to the geographical distribution of A. solstitialis in
the Mediterranean basin and to investigate dynamics of
the eriophyoid population on YST in Apulia (Italy).
Methods and materials
To study A. solstitialis geographical distribution, parts
of symptomatic plants were collected in Greece (Alexandroupolis, Kilkis), Bulgaria (Plovdiv), Turkey
(Goreme) and Italy (Apulia and Sicily) in 2001 and
2003 to 2006. Eriophyoids were extracted from fresh
and dried materials, slide mounted and identified.
To investigate dynamics of eriophyoid populations,
YST samples were randomly collected from four sites
in Andria countryside (Italy). In 2004, 2005 and 2006,
20 plants were sampled from each site periodically at
30- and 15-day intervals from February to June and
July to August, respectively. Four categories of samples, when available, were separately collected from
each plant: new rosette leaf (with length 2 cm); mature rosette leaf (with length 10 cm); stem (without
flower head); flower head. Mites were extracted by a
washing and sieving method (de Lillo, 2001; Monfreda et al., 2007), identified, counted and recorded
per part of plant. The collected samples were carefully
inspected for detecting plant alterations related to eriophyoid activity.
Table 1.
Results
A. solstitialis was recorded on YST in Bulgaria, Greece,
Italy and Turkey, together with other species of eriophyoid mites (Table 1), some of which are related to
YST. Few specimens were found for other species for
which the identification was only at genus level. None
of them fitted with the descriptions of species known to
be associated with Centaurea spp.
In Apulia, ten species were detected on symptomatic
yellow star thistle during 20042006. No eriophyoids
were collected in February and March samplings, during the years of study. A. solstitialis was identified from
May to August in sites 2 and 3 only, where other species were also present.
In 2004 samplings, A. solstitialis occurred from May
to August. It appeared with very low population density in the first half of June on only new rosette leaves
in sampling site 2 but both on new and mature leaves
in site 3. It disappeared at the beginning of August on
stems at site 2 and at the end of August on stem and
flower heads at site 3. The highest population density
was found at the beginning of July at level of stem at
site 2 and 3, and in the early August on flower head
before flower senescence (site 3).
In 2005 samplings, A. solstitilis was present on
leaves at the end of May only in site 3, and on stems at
the end of May in site 2 and longer in site 3 up to the
ist of the eriophyoids collected on Centaurea solstitialis in some countries of the Mediterranean
L
area.
Collection date
Collection locality
Species
June 2001
July 2001
November 2001
June 2003
June 2003
June 2003
GoremeCappadocia (Turkey)
GoremeCappadocia (Turkey)
GoremeCappadocia (Turkey)
Kilkis (Greece)
Alexandroupolis (Greece)
GoremeCappadocia (Turkey)
June 2003
June 2003
June 2003
PalermoSicily (Italy)
Piano TorreSicily (Italy)
MongerratiSicily (Italy)
August 2004
AprilAugust 2004
Plovdiv (Bulgaria)
AndriaApulia (Italy)
AprilAugust 2005
AndriaApulia (Italy)
AprilAugust 2006
AndriaApulia (Italy)
Aceria solstitialis
Aceria solcentaureae
Aceria solstitialis
Aceria solstitialis
Aceria solstitialis
Aceria solstitialis
Aceria sp.
Aceria sp.
Aceria centaureae
Aceria centaureae
Aceria solstitialis
Aceria solstitialis
Aceria solstitialis
Aceria solcentaureae
Aceria centaureae
Aceria sp.
Aculodes sp.
Aculus sp.
Calepitrimerus sp.
Epitrimerus sp.
Eriophyes sp.
Trisetacus sp.
Aceria solstitialis
Aculodes sp.
Aceria solstitialis
Aceria sp.
Aculodes sp.
179
XII International Symposium on Biological Control of Weeds
180
Figure 1.
Population dynamics of Aceria solstitialis on Centaurea solstitialis during 2004, 2005 and 2006 in two sites in Apulia.
Eriophyoid mites on Centaurea solstitialis in the Mediterranean area
beginning of August. The highest density occurred in
the second part of June in site 2 and in the first part of
July in site 3, both at stem level.
Finally, in the 2006 samplings, this species was
found only at site 3 from June to August, and its population density was higher at level of the stems at the
beginning of August (Fig. 1).
University, Plovdiv, Bulgaria), and Franca Di Cristina,
Francesca Lecce, Alessandra Paolini and Carlo Tronci
(BBCA, Rome, Italy) for eriophyoid mite collection.
We would also like to thank Lincoln Smith, Ray Carruthers (both USDA, ARS, Albany, CA), and Michael
Pitcairn (CDFA) for their scientific and logistic support
to the project.
Discussion
References
A wide distribution of A. solstitialis in the Mediterranean basin was ascertained, together with other mite
species. In Apulia, ten species were detected on YST,
in the period of study (20042006). Aculus sp., Calepitrimerus sp., Epitrimerus sp., Eriophyes sp. and Trisetacus sp. could be related to neighbouring plants or
they could have been dispersed by wind from other localities and other hosts because of their very low population density on C. solstitialis in the sampling sites.
Specimens with similar morphometric features have
not been found in the other countries of our study. Aculodes sp. and Aceria sp. could represent new species,
but their host range needs to be verified.
Population density of A. solstitialis in 3 years of
observations was highest at level of the stems in July
and August samplings but always with very low values
(at most 30 mites per sample). Moreover, the mite appeared in April and disappeared in August; no information on overwintering location was obtained. The
hypothesis of mite dispersal by means of wind from
other areas of the Mediterranean basin (i.e. Balcanian
area), in coincidence of the highest density levels of
the mite on the host, is highly reasonable (Sabelis and
Bruin, 1996). The causes of their scarce density population on Apulian YST, while a large population was
observed on YST from other areas, are still unknown.
Possible explanations should be found in the genetic
background of YST populations collected in different
Mediterranean countries in case they could belong to
different strains.
Even though A. solstitialis seems a promising agent
for YST biological control, its Italian population seems
to be insufficient to provide a significant reduction of
seed production and requires further study.
Briese, D.T. and Cullen, J.M. (2001) The use and usefulness
of mites in biological control of weeds. In: Halliday, R.B.,
Walter, D.E., Proctor, H.C., Norton, R.A., Colloff, M.J.
(eds) Acarology: Proceedings of the 10th International
Congress. CSIRO Publishing, Melbourne, Australia, pp.
453463.
de Lillo, E. (2001) A modified method for eriophyoid mite
extraction. International Journal of Acarology 27 (1),
6770.
de Lillo, E., Cristofaro, M. and Kashefi, J. (2003) Three new
Aceria species (Acari: Eriophyoidea) on Centaurea spp.
(Asteraceae) from Turkey. Entomologica 36, 121137.
Lass, L.W., McCaffrey, J.P. and Callihan, R.H. (1996) Detection
of yellow starthistle (Centaurea solstitialis) and common
St. Johnswort (Hypericum perforatum) with multispectral
digital imagery. Weed Technology 10, 466474.
Maddox, D.M., Mayfield, A. and Poritz, N.H. (1985) Distibution of yellow starthistle (Centaurea solstitialis) and
Russian knapweed (Centaurea repens). Weed Science 33,
315327.
Maddox, D.M. (1981) Introduction, phenology, and density
of yellow starthistle in costal, intercostal, and central valley situations in California. Agricultural Research Results
ARR-W-20/July, pp. 133.
Monfreda, R., Nuzzaci, G. and de Lillo, E. (2007) Detection,
extraction and collection of eriophyoid mites. Zootaxa
1662, 3543.
Piper, G.L. (2001) The biological control of yellow starthistle
in the western U.S.: Four decades of progress. In: Smith,
L. (ed.) The First International Knapweed Symposium of
the Twenty-First Century. Coeur dAlene, Idaho, USDAARS, Albany, pp. 4855.
Roch, B.F.Jr. and Roch, C.T. (1991) Identification, introduction, distribution, ecology and economics of Centaurea species. In: James, L.F., Evans, J.O., Ralph, M.H. and
Child, R.D. (eds) Noxius Range Weeds. Westview Press,
Boulder, Colorado, pp. 274291.
Roch, B.F.Jr. and Talbott, C.T. (1986) The collection history
of Centaureas found in Washington State. Research Bulletin XB 0978. Agricultural Research Centre, Washington
State University, Pullman, WA, 36 pp.
Sabelis, M.W. & Bruin, J. (1996) Evolutionary ecology:
life history patterns, food plant choice and dispersal.
In: Lindquist, E.E., Sabelis, M.W. and Bruin, J. (eds)
Eriophyoid Mites. Their Biology, Natural Enemies and
Control. Elsevier, Amsterdam, Netherlands, pp. 329366.
Acknowledgements
This research was partly supported by the University
of Bari and the Biotechnology and Biological Control
Agency. We specifically thank Javid Kashefi (USDAARS EBCL, Thessaloniki Substation, Greece), Prof
Vili Harizanova and Dr Atanaska Stoeva (Agricultural
181
Diclidophlebia smithi (Hemiptera: Psyllidae)
a potential biological agent for
Miconia calvescens
E.G.F. Morais,1 M.C. Picano,1 R.W. Barreto,2 G.A. Silva,1
M.R. Campos1 and R.B. Queiroz1
Summary
Diclidophlebia smithi Burckhardt, Morais and Picano (Hemiptera: Psyllidae) is a monophagous
species which was selected as possible agent of biological control of miconia, Miconia calvescens
DC. (Melastomataceae), a native plant of Central and South America that has become an aggressive invader of forest ecosystems in French Polynesia, Hawaii and Australia. The objective of this
work was to study the biology and population dynamics of D. smithi in Viosa and Dionsio (state of
Minas Gerais, Brazil) from June 2001 to June 2002 and from February 2004 to February 2005 and
evaluate injuries caused to the host plant and occurrence of natural enemies of this psyllid. Frequency
distribution of the distance between the antennae indicated the existence of five nymphal instars.
Colonies of the psyllid were observed throughout the year in Viosa and Dionsio. Population peaks
occurred from April to July (winter: a period of low temperatures, drought and short photoperiod in
this region). Nymphs and adults were observed attacking buds, inflorescences and infrutescences of
M. calvescens and causing damage by sucking the plant sap and injecting toxins. Desirable traits such
as high population growth rate, easy mass rearing, occurrence throughout the year, host specificity,
attack to reproductive organs and potential capacity to adapt to different climatic conditions including
those similar to where M. calvescens invasions are occurring, all indicate that D. smithi is a promising
biological control agent of M. calvescens.
Keywords: classical biological control, weed, Hawaii, population dynamics.
Introduction
Miconia, Miconia calvescens DC. (Melastomataceae)
is a plant native from Central and South America that
became an aggressive forest ecosystem invader in
French Polynesia and Hawaii (Meyer, 1996; Csurhes,
1997; Medeiros et al., 1997; PIER, 2002) and is also
becoming a cause of concern in Australia (Csurhes
and Edwards, 1998). It was introduced as an ornamental into Tahiti in 1937 and slowly invaded the native
forests and now dominates 65% of the island (Meyer,
1996; Meyer and Florence, 1996; Meyer and Malet,
1997). Introduction into Hawaii occurred in the 1960s,
and since 1992 it has been included in the list of nox-
Departamento de Biologia Animal, Universidade Federal de Viosa,
Viosa, MG 36570-000, Brazil.
2
Departamento de Fitopatologia, Universidade Federal de Viosa, Viosa,
MG 36570-000, Brazil.
Corresponding author: E.G.F. Morais <[email protected].>
CAB International 2008
1
ious invasive weeds (Medeiros et al., 1997). It is now
regarded as one of the hundred worst invasive weeds in
the world (IUCN, 2007) and a serious threat to several
large tropical forest ecosystems in the world, particularly in oceanic islands (Csurhes, 1997).
Classical biological control probably represents the
sole sustainable alternative of control for invasive miconia populations. Studies were started in 1995 in Brazil aimed at finding pathogens to be used as classical
biocontrol agents of miconia. Brazil is part of the centre of origin of M. calvescens and the country where the
type material of the plant was first collected. Surveys
were extended to Costa Rica, Dominican Republic and
Ecuador. Results of the survey in Brazil were recently
published (Seixas et al., 2007). Surveys for arthropod
natural enemies of M. calvescens in Brazil started later
in 2001 and concentrated in Viosa, Dionsio and
Guaraciaba, which are municipalities in the state of
Minas Gerais. Over 70 species of arthropods were collected and identified, of which around 10 were recognized as having potential as biological control agents
182
Diclidophlebia smithi (Hemiptera: Psyllidae) a potential biological agent for Miconia calvescens
(Picano et al., 2005). This included a newly described
species of psyllid - Diclidophlebia smithi Burckhardt,
Morais and Picano (Hemiptera, Psylloidea) (Burckhardt et al., 2006).
D. smithi is a monophagous species that appears to
be widely distributed in south-eastern Brazil. It was
found in Dionsio, Guaraciaba and Viosa in Minas
Gerais State and Mangaratiba, Angra dos Reis (Ilha
Grande) in Rio de Janeiro State. Psyllids are small sapsucking insects that are typically highly specialized and
either monophagous or oligophagous, often having the
necessary attributes of a good biological control agent
(Hollis, 2006). Nymphs and adults cause damage to
their host plants either by sap sucking or toxin injection. This often leads to deformation of leaves and buds,
necrosis, premature senescence of leaves and, in some
species, to the formation of galls on infested plants.
Some species of psyllids are also known to function as
vectors of pathogenic phytoplasms and bacteria (Phyllis and Leann, 2006).
Before a biological control candidate can be further
considered for use in a classical biological control program, a series of studies on its biology are regarded as
necessary. These involve its precise identification, development of methods of raising it under controlled conditions, host-range evaluation, elucidation of life cycle,
investigating its natural enemies and potential impact
on host plant among others (Julien, 1997). This report
presents the biology of D. smithi including life cycle,
population dynamics, a description of damage caused to
host plants and the occurrence of natural enemies.
Materials and methods
Life cycle
This study was performed at the Laboratrio de Manejo Integrado de Pragas of the Departamento de Biologia Animal-Universidade Federal de Viosa. Nymphs of
D. smithi were brought from the field and transferred to
young potted M. calvescens plants kept in a greenhouse.
To determine the number of instars during the life
cycle of D. smithi, 20 new adult couples were placed on
five M. calvescens plants (four couples per plant) and
left there for egg-laying. Thirty nymphs obtained from
the eggs were placed separately on young field-collected
miconia leaves plus stem and maintained with the base
of the stem immersed in water.
A daily evaluation was made of the following parameters: distance between the antennae, length of antennae and length and breadth of thorax at the level of
the second pair of legs. Such measurements were made
with the help of a dissecting microscope Leica MZ75
to which a digital camera was attached. Sizes were obtained through the analysis of photos taken with the
camera with the software Leica Qwin. Additionally,
body sizes of 40 adults (20 females and 20 males) and
of 20 eggs were also measured.
Biometric data of the distance between the antennae
served as the basis for the preparation of multimodal
curves of frequency distribution aimed at determining
the number of instars for D. smithi. Values of distance
between the antennae were plotted on the abscissa axis,
whereas the frequency of occurrence of each value was
plotted on the ordinate axis. Each peak in the curve
would then correspond to an instar in the insects cycle
(Parra and Haddad, 1989). Means and standard deviation for each biometric parameter, as well as for the duration of each stage, were calculated after the number
of instars was determined.
These experiments were conducted at 20.4 0.3C,
relative humidity of 77.3 1.2% and daily light regime
of 12.1 0.04 h.
Population fluctuation, injuries
and natural enemies
Population density of D. smithi was evaluated in M.
calvescens plants in Viosa and Dionsio (state of
Minas Gerais, Brazil). Viosa is at 649 m of altitude,
204514 S, 425253 W. Mean temperature was
19.4oC and precipitation 1221.4 mm/year. Dionsio
is at 344 m of altitude and is located at 195034 S,
424636 W. Mean temperature was 23.2oC and precipitation 1003 mm/year.
Evaluations were made during two different periods
of time, first, from June 2001 to June 2002, and second, from February 2004 to February 2005. During the
first period, evaluations were made at 3-week intervals
at the two locations with a total of 15 evaluations at
Viosa and 16 at Dionsio. During the second period,
evaluations were made at 15-day intervals in Viosa
and monthly at Dionsio, resulting in 27 evaluations in
Viosa and 13 in Dionsio.
Ten M. calvescens plants were selected randomly in
each location, and the following plant organs (leaves,
branches, buds, flowers and fruits) were evaluated per
plant during each visit. Number of nymphs and adults
of D. smithi and possible predators associated with the
psyllid were recorded. The occurrence of D. smithi
was also investigated on other plants surrounding each
M. calvescens plant. Comparisons were made between
attacked and healthy organs of miconia, particularly for
the production of new growth, flowers, fruits, as well as
the development of plant organs.
Climatic data for Viosa were obtained from the
Estao Climatolgica Principal (INEMET/5o DISME/
UFV), whereas those for Dionsio were obtained from
the Estao Climatolgica de Ponte Alta belonging to
the private company CAF Santa Brbara Ltda.
Occurrence of parasites on D. smithi was investigated on branches colonized by the psyllid that were
collected at each visit to the sites and brought to the
lab. These branches were left in plastic boxes with thin
meshed screened openings. The base of those branches
183
XII International Symposium on Biological Control of Weeds
was inserted in a layer of humid vermiculite over
which a cardboard disk was placed separating the
vermiculite from the upper chamber to avoid direct
contact of the insects with the vermiculite. Branches
were maintained for 3 weeks in the lab under room
conditions. Appearance of parasitoids was evaluated
at 2-day intervals.
Population fluctuation curves for nymphs and adults
of D. smithi were prepared for each of the localities
(Viosa and Dionsio) for the two evaluation periods.
Information about the phenology of M. calvescens over
time was also recorded. Curves of recorded rainfall,
average air temperature, relative humidity and photoperiod were also prepared.
Results
Life cycle
Eggs of D. smithi are pale yellow when laid and rapidly darken to black and shiny. They are elliptic with
Table 1.
one pointed end and a peduncule attaching it to the
plant. Its length is 0.24 0.009 mm, and its diameter is
0.13 0.003 mm (Table 1).
The frequency curve of the length between the antenae indicated that D. smithi has five instars (Figure 1).
After egg-hatching, nymphs of the first instar are pale
yellow and almost transparent and become black with
time. Wing pads become visible during the third instar
and third and fourth instars and are entirely yellow. In
the fifth instar, the wing pads, as well as the last segment of the abdomen and the head become brown.
Body and antennae length are given in Table 1.
Recently emerged adults are yellow with transparent
wings that become brown a few minutes later. Females
are bigger (1.99 0.04 mm long) than males (1.71 0.03
mm long; Table 1). A female can lay from 25 to 45 eggs
during its lifetime in lab conditions (19 2C). Duration
of each phase in the life cycle is presented in Table 1.
Nymphs of D. smithi secrete large quantities of
whitish filamentous wax that can completely recover
the colony, possibly protecting them against natural enemies, dehydration and rain.
Biometric data and duration of stages in the life cycle of Diclidophlebia smithi (Hemiptera: Psyllidae).
Life stage
Body
Antenna
Length (mm)
Width (mm)
Length (mm)
0.24 0.009
0.26 0.003
0.32 0.007
0.42 0.006
0.63 0.025
1.00 0.021
1.71 0.03
1.99 0.09
0.13 0.003
0.16 0.003
0.20 0.004
0.25 0.003
0.36 0.010
0.48 0.009
0.60 0.008
0.62 0.014
0.12 0.005
0.14 0.006
0.19 0.010
0.35 0.016
0.50 0.011
0.60 0.011
0.59 0.014
Egg
First instar
Second instar
Third instar
Fourth instar
Fifth instar
Adult (male)
Adult (female)
Distance between (mm)
9.33 0.09
3.09 0.06
3.76 0.25
5.13 0.44
4.50 0.42
7.09 0.34
12.70 0.45
12.00 0.36
0.12 0.003
0.15 0.002
0.20 0.002
0.26 0.003
0.34 0.005
-
3rd instar
0.16
4th instar
2nd instar
0.12
Frequency
Duration (days)
5th instar
1st instar
0.08
0.04
0.00
0.10
0.15
0.20
0.25
0.30
0.35
Distance between antennae (mm)
Figure 1.
Frequency of distance between antennae in nymphs of Diclidophlebia smithi (Hemiptera: Psyllidae).
184
Diclidophlebia smithi (Hemiptera: Psyllidae) a potential biological agent for Miconia calvescens
Population fluctuation, injuries
and natural enemies
The highest degree of damage caused by D. smithi to
M. calvescens happened at high colony densities on
buds, flowers and fruits. In such situations, the buds
and young leaves became distorted and yellow, and
their growth was hampered.
Nymph populations peaked from April to June of
2002 in Viosa, coinciding with the period of the year
when there is abundant new growth. In Dionsio, the
M. calvescens phenology
Dionsio
Viosa
50
Nymphs/plant
population peak occurred in July of 2004 during a
period of abundant fruiting. The highest values for population densities for adults occurred in the first year of
evaluation, in July 2002, during a period of abundant
fruiting and flowering. In the second year of evaluation,
the peak occurred in April coinciding with the end of a
fruiting period when new growth was produced. Those
population peaks coincided with periods of cool weather,
short photoperiod and drought (Figures 2and 3).
The only predator found attacking D. smithi nymphs
was an unidentified Syrphidae larva. It was seldom
Viosa
(a)
Dionsio
40
30
20
10
0
(b)
Adults/plant
40
30
20
10
Figure 2.
02/05
12/04
10/04
08/04
06/04
Months
04/04
02/04
06/02
04/02
02/02
12/01
10/01
08/01
06/01
Phenology of Miconia calvescens: vertical stripes bud emission; diagonal stripes flowering;
horizontal stripes fruiting, and population density of nymphs (a) and adults (b) of Diclidophlebia smithi (Hemiptera: Psyllidae) in Dionsio and Viosa (state of Minas Gerais-Brazil),
20012002 and 20042005. Vertical lines on the population density graphs represent the standard error for the means.
185
XII International Symposium on Biological Control of Weeds
120
120
80
80
40
Figure 3.
Months
Relative humidity (%)
15
40
02/05
12/04
10/04
08/04
06/04
10
04/04
12/04
02/05
10/04
06/04
08/04
02/04
04/04
06/02
04/02
02/02
12/01
10/01
08/01
06/01
10
60
20
02/04
40
100
80
06/02
15
Air temperature
25
04/02
60
20
30
02/02
25
Relative humidity
Photoperiod
35
12/01
80
10/01
100
08/01
30
Air temperature
Relative humidity (%)
Relative umidity
Photoperiod
06/01
0
35
Air temperature (C) e Photoperiod (hours)
40
Air temperature (C) e Photoperiod (hours)
160
Rainfall (mm)
Rainfall (mm)
160
Months
Climatic data: precipitation, relative humidity and photoperiod in Viosa (a) and Dionsio (b), state of Minas
Gerais, Brazil, 20012002 and 20042005.
found in the field, but it was observed to be a very
voracious predator. No parasitoid was obtained during
the investigation of colonies brought from the field.
Discussion
Psyllid population peaked from April to June indicating that the density of this insect was related to a high
abundance of food such as buds, flowers and fruits that
formed during that period. Soft tissues are of critical
importance for population growth of this sap-feeding
insect. Similar connection of phenological states in
plants and insect population density was also observed
for other species in the Psyllidae such as Euphalerus
clitoriae (Burckhardt and Guajar, 2000) Diaphorina
citri Kuway, Diclidophlebia lucens (Burckhardt et al.,
2005) and Boreioglycaspis malaleucae Moore (Burckhardt and Guajar, 2000; Tsai and Liu, 2000; Burckhardt et al., 2005; Center et al., 2006).
The highest population densities of D. smithi also coincided with periods of cooler air temperatures, shorter
photoperiods and lower rainfall levels. Air temperature emerged as the climatic factor that had the highest influence on population density. High temperatures
may harm the development of D. smithi, particularly
when relative humidity is low. This was also observed
for other Psyllidae such as Euphalerus vittatus Crawford on Cassia fistula L. (Fabaceae) and E. clitoriae
(Burckhardt and Guajar, 2000) on Clitoria fairchildiana Howard. (Fabaceae) (Sahu and Mandal, 1999;
Gondim Junior et al., 2005). Air temperature is one of
the most important factors interfering with establishment, proliferation, dispersal and impact of organisms
in new areas of distribution (Baker, 2002). High air
temperatures cause a high mortality of nymphs and a
reduction in egg-laying for female psyllids because of
abnormal development of ovaries (Mehrnejad and Cop
land, 2005; Stratopoulou and Kapatos, 1995; Liu and
Tsai, 2000). Adult Agonoscena pistaciae Burckhardt
and Lauterer (Hemiptera: Psyllidae) had maximum
rates of egg-laying in lab conditions at 20C and a reduction in ovary development when temperatures were
above 30C (Mehrnejad and Copland, 2005). High air
temperatures can also reduce survival and longevity of
psyllids as observed for Trioza hirsuta Crawford (Hemiptera: Psyllidae) which had a longer period of survival for adults at 25C (Dhiman and Singh, 2003) and
D. citri, for which the ideal temperature is around 20C
(Liu and Tsai, 2000).
Rainfall also influences population density of D.
smithi. Rain can cause mortality through mechanical
impact on individuals and can remove individuals from
their host plant or, in the case of nymphs, remove the
wax cover that protects them (Center et al., 2006), exposing them to adverse environmental conditions. Rain
may harm psyllid mating and movement, leading to a
reduction in population growth. Photoperiod may have
an indirect effect on this insects populations. Periods
of adverse environmental conditions do not appear to
represent a significant obstacle for this psyllid species
since colonies of D. smithi recover readily after unfavourable periods, and colonies are seen in the field
throughout the year.
Knowledge about effects of climate on a potential
biocontrol agent is of critical importance in order to
estimate its chance of success and to choose the correct
186
Diclidophlebia smithi (Hemiptera: Psyllidae) a potential biological agent for Miconia calvescens
time of the year for its introduction into the field. In the
case of D. smithi as a biocontrol for M. calvescens, the
ideal time of the year would be a period of lower rainfall levels and cooler air temperatures and when there
was plenty of new growth.
The attack of D. smithi to buds of M. calvescens is
likely to effect plant development through starvation of
such new organs of photoassimilates brought in the sap.
The mass of wax filaments appear to have a deleterious
effect on the formation of normal flowers and fruits. If
seed production is affected, the impact might be important since M. calvescens relies entirely on seeds for its
dispersal (Meyer, 1998). In the case of citrus attacked
by D. citri, a reduction on seed setting has been demonstrated (Michaud, 2004). More detailed impact studies are presently being performed for D. smithi on M.
calvescens.
Features of D. smithi such as it being relatively easily amenable to mass rearing under controlled conditions, its occurrence in the field throughout the year, the
fact that it attacks the reproductive organs of the plant,
its high host specificity (discussed elsewhere) and its
ability to adapt to a range of climatic conditions indicates that it has good potential as a classical biological
control agent to be used against M. calvescens.
Acknowledgements
This work forms part of a research project submitted
as a M.Sc. dissertation to the Departamento de Biologia Animal/Universidade Federal de Viosa by E.G.F.
Morais. This study was supporded by USGS BRD Pacific Island Ecosystem Research Centre, National Park
Service, the Research Corporation of Hawaii and the
Conselho Nacional de Desenvolvimento Cientfico e
Tecnolgico (CNPq). The authors acknowledge CAF
Santa Brbara Ltda which allowed part of the field
work to be performed in its property and provided the
team with local support and climatic data.
References
Baker, R.H.A. (2002) Predicting the limits to the potential
distribution of alien crop pests. In: Hallman, G.J. and
Schwalbe, C.P. (eds) Invasive Arthropods in Agriculture
Problems and Solutions. Science Publishers, Enfield, NH,
pp. 207224.
Burckhardt, D. and Guajar, M. (2000) Euphalerus clitoriae
sp. n., a new psyllid species from Clitoriae fairchildiana
(Fabaceae: Papilionoideae), and notes on other Euphalerus spp. (Hemiptera, Psylloidea). Revue Suisse de Zoologie 107, 325334.
Burckhardt, D., Hanson, P. and Madrigal, L. (2005) Diclidophlebia lucens, n. sp. (Hemiptera: Psyllidae) from Costa
Rica, a potential control agent of Miconia calvescens
(Melastomataceae) in Hawaii. Proceedings of the Entomological Society of Washington 107, 741749.
Burckhardt, D., Morais, E.G.F. and Picano, M.C.(2006)
Diclidophlebia smithi sp. n., a new species of jumping
plant-lice (Hemiptera, Psylloidea) from Brazil associated
with Miconia calvescens (Melastomataceae). Mitteilungen der Schweizerischen Entomologischen Gesellschaft
79, 241250.
Center, T.D., Pratt, P.D., Tipping, P.W., Rayamajhi, M.B.,
Van, T.K., Wineriter, S.A., Dray Jr., F.A. and Purcell, M.
(2006) Field colonization, population growth, and dis
persal of Boreioglycaspis melaleucae Moore, a biological
control agent of the invasive tree Melaleuca quinquenervia
(Cav.) Blake. Biological Control 39, 363374.
Csurhes, S.M. (1997) Miconia calvescens, a potentially invasive plant in Australias tropical and sub-tropical rainforests. In: Meyer, J.Y. and Smith, C.W. (eds) Proceedings
of the First Regional Conference on Miconia Control. Pa
peete, Taiti, pp. 7277.
Csurhes, S.M. and Edwards, R. (1998) Potential environmental weeds in Australia. Queensland Department of Natural
Resources, Coorparoo, Australia, 208 pp.
Dhiman, S.C. and Singh, S. (2003) Some ecological aspects
of Trioza hirsuta Crawford (Homoptera: Psyllidae): A
pest of Terminalia tomentosa. Journal of Experimental
Zoology India 6, 373376.
Gondim, M.G.C., Jr., Barros, R., Silva, F.R. and Vasconcelos, G.J.N. (2005) Occurrence and biological aspects of
the clitoria tree psyllid in Brazil. Science Agriculture 62,
281285.
Hollis, D. (2006) Australian Psylloidea: jumping plantlice
and lerp insects. Systematic Entomology 31, 199200.
IUCN. International Union for Conservation of Nature. 100
of the Worlds Worst Invasive Alien Species. Auckland
Invasive species specialist group. Available at: www.iucn.
org/places/medoffice/invasive_species/docs/invasive_
species_booklet.pdf, accessed 29 January 2007.
Julien, M.H. (1997) Success, and failure, in biological control
of weeds. In: Julien, M.H. and White, G. (eds) Biological Control of Weeds: Theory and Practical Application.
ACIAR Monograph, 9. Canberra: Australian Centre for
International Agricultural Research, pp. 915.
Liu, Y.H. and Tsai, J.H. (2000) Effects of temperature on biology and life table parameters of the Asian citrus psyllid,
Diaphorina citri Kuwayama (Homoptera: Psyllidae). Annals of Applied Biology 137, 2016.
Medeiros, A.C., Loope, L.L., Conant, P. and Mcelvaney, S.
(1997) Status, ecology, and management of the invasive
plant, Miconia calvescens DC (Melastomataceae) in the
Hawaiian Islands. Bishop Museum Occasional Papers 48,
2336.
Mehrnejad, M.R.and Copland, M.J.W. (2005) The seasonal
forms and reproductive potential of the common pistachio
psylla, Agonoscena pistaciae (Hem., Psylloidea). Journal
of Applied Entomology 129, 342346.
Meyer, J.Y. (1996) Status of Miconia calvescens (Melastomataceae), a dominant invasive tree in the Society Islands
(French Polynesia). Pacific Science 50, 6676.
Meyer, J.Y. (1998) Observations on the reproductive biology
of Miconia calvescences DC (Melastomataceae), an alien
invasive tree on the island of Tahiti (South Pacific Ocean).
Biotropica 30, 609624.
Meyer, J.Y. and Florence, J. (1996) Tahitis native flora endangered by the invasion of Miconia calvescens DC. (Melastomataceae). Journal of Biogeography 23, 775781.
Meyer J.Y. and Malet, J.P. (1997) Study and management of the alien invasive tree Miconia calvescens DC.
187
XII International Symposium on Biological Control of Weeds
(Melastomataceae) in the islands of Raiatea and Tahaa (Society Islands, French Polynesia): 19921996. Technical
Report, 111. Manoa: Cooperative National Park Resources
Studies Unit, University of Hawaii at Manoa, 56 pp.
Michaud, J.P. (2004) Natural mortality of Asian citrus psyllid (Homoptera: Psyllidae) in central Florida. Biological
Control 29, 260269.
Parra, R.P.P. and Haddad, M.L. (1989) Determinao do Nmero de nstares de Insetos. Piracicaba: FEALQ, 49 pp.
Phyllis, G.W. and Leann, B. (2006) Insect vectors of phytoplasmas. Annual Review of Entomology 51, 91111.
Picano, M.C., Barreto, R.W., Fidelis, E.G., Semeo, A.A.,
Rosado, J.F., Moreno, S.C., Barros, E.C., Silva, G.A. and
Johnson, T. (2005) Biological control of Miconia calvescens by phytophagous arthropods. Technical Report 134.
Manoa: Pacific Cooperative Studies Unit, University of
Hawaii at Manoa, Manoa, HI, 30 pp.
PIER. Pacific Island Ecosystems at Risk, 2002. Miconia calvescens. Available at: http://www.hear.org/pier/species/
miconia_calvescens.htm, accessed 29 January 2007.
Sahu, S.R. and Mandal, S.K. (1999) Seasonal activity of
amaltas psyllid Euphalerus vittatus Crawford (Psyllidae:
Hemiptera). Environmental and Ecology 17 509510.
Seixas, C.D., Barreto, R.W. and Killgore, E. (2007). Fungal pathogens of Miconia calvescens (Melastomataceae)
from Brazil, with reference to classical biological control.
Mycologia 99, 99111.
Stratopoulou, E.T. and Kapatos, E.T. (1995) The dynamics
of the adult population of pear psylla, Cacopsylla pyri
L. (Hom., Psyllidae) in the region of Magnesia (Greece).
Journal of Applied Entomology 119, 97101.
Tsai, J.H. and Liu, Y.H. (2000) Biology of Diaphorina citri
(Homoptera: Psyllidae) on four host plan. Journal of Economic Entomology 93, 17211725.
188
A lace bug as biological control agent of
yellow starthistle, Centaurea solstitialis
L. (Asteraceae): an unusual choice
A. Paolini,1 C. Tronci,1 F. Lecce,1 R. Hayat,2 F. Di Cristina,1
M. Cristofaro3 and L. Smith4
Summary
The lace bug Tingis grisea Germ. (Hemiptera: Tingidae) is a univoltine sap-feeder associated with
the genus Centaurea L. and distributed throughout Central and Southern Europe and the Middle East.
In 2002, one Turkish population of T. grisea was selected as a potential biological control agent for
yellow starthistle, Centaurea solstitialis L., (Asteraceae: Cardueae), a weed of primary concern in
the USA. Field observations showed that significant damage was caused to the host plant especially
when many individuals were feeding on the same plant. Life-cycle and biology observations were
made to assess the duration of the five nymphal instars of T. grisea under laboratory conditions, as
well as female fecundity and longevity. Starvation and oviposition no-choice tests were carried out in
order to determine the host specificity of the insect. Results showed a clear oligophagous behaviour
closely restricted to the genus Centaurea. In addition, among the three Centaurea spp. on which full
larval development was ascertained (C. solstitialis, Centaurea sulphurea, Centaurea cyanus), yellow
starthistle was clearly most suitable regarding number of eggs laid and number of adults obtained.
Keywords: YST, host range, Tingis grisea.
Introduction
Yellow starthistle, Centaurea solstitialis L., (Asteraceae: Cardueae) is an important invasive alien weed
of rangeland in the western United States and is the
target of a USDA classical biological control program
(Turner et al., 1995; Sheley et al., 1999; Smith, 2004;
Di Tomaso et al., 2006). Six insect agents that attack
C. solstitialis flowerheads have already been introduced
(Cristofaro et al., 2002; Pitcairn et al., 2004), but they do
not appear to be reducing the weed population sufficiently (Pitcairn et al., 2000, 2006). Therefore, it is desirable
to find new agents that attack other organs of the plant
or earlier phenological stages. A rust, Puccinia jaceae
Biotechnology and Biological Control Agency, Via del Bosco 10, 00060
Sacrofano, Rome, Italy.
2
Atatrk University, Faculty of Agriculture, Plant Protection Department, 25240 TR Erzurum, Turkey.
3
ENEA C.R. Casaccia BIOTEC, Via Anguillarese 301, 00123 S. Maria
di Galeria, Rome, Italy.
4
USDA-ARS, 800 Buchanan Street, Albany, CA 94710, USA.
Corresponding author: A. Paolini <[email protected]>.
CAB International 2008
1
var. solstitialis Otth., has been released (Woods, 2004;
Fisher et al., 2006), and two beetles are being evaluated
(Cristofaro et al., 2004; Smith, 2004; Smith, 2007); all
of these attack immature plants. However, it would also
be useful to have an agent that stresses the plant later in
the growing season, during the critical period when it
is flowering and producing seed (Smith, 2004). During
foreign exploration for new biological control agents in
eastern Turkey in 2002, we discovered a large population of the lace bug Tingis grisea Germ. 1835 (Hemiptera: Tingidae) feeding on mature C. solstitialis rosettes
and on bolting plants.
In the literature, this lace bug has been reported
from 11 species of Centaurea, including C. solstitialis, as well as from Crupina vulgaris L. (Stusak, 1959),
a very closely related plant species (Susanna et al.,
1995). Its geographical distribution is very wide: it
occurs from central Spain across southern Europe to
southern Russia; generally overlapping the distribution
of C. solstitialis and some close relatives (Komarov,
1934; Klokov et al., 1963; Wagenitz, 1975; Dostl,
1976). Little is known about its biology (Stusak, 1959;
Pricart, 1983). T. grisea is reported as univoltine and
overwinters as adult. Oviposition begins in May, and
189
XII International Symposium on Biological Control of Weeds
eggs are inserted into the young tissue of stems and axils. Eggs are very small; the operculum measures 0.16
0.05 mm (Stusak, 1957). In Ukraine, nymphs appear
at the beginning of June and adults at the beginning of
July. Five nymphal instars have been described.
This paper reports results of studies on the life cycle, rearing and host-plant specificity of this insect to
determine whether it warrants further evaluation as a
candidate for biological control of yellow starthistle.
Methods and materials
Collection of insects
Insects emerging from winter diapause were col
lected in late March of 2004, 2005 and 2006 in the
vicinity of Horasan (Erzurum Region, 1600 m ASL),
Eastern Turkey. In the laboratory, lace bugs were kept
in a 3-l glass beaker at low temperature (8C) and
12:12 h L/D photoperiod. Insects were allowed to feed
on freshly cut leaves of C. solstitialis (US biotype) held
in water vials; crumpled tissue paper was also provided
as shelter for insects to rest.
Laboratory rearing
Insect rearing was carried out on natural substrate,
using potted plants of C. solstitialis (US biotype) at
23C to 26C and 16:8 h L/D. Single pairs of T. grisea were confined to a portion of yellow starthistle
stem anchored in a foam disk on the bottom of a 17
5 cm transparent plastic cylinder, capped with fine nylon mesh. A hole in the side, closed by a foam plug,
was used for insect manipulation. After 7 days, insects
were removed and transferred to another stem under
the same conditions. Beginning 10 days after insect
exposure, stems were cut off and daily examined under a stereo microscope to search for neonate nymphs.
The same procedure was repeated several times for
each pair of insects. Emerged nymphs were used for
host-range, larval transfer experiments and life-cycle
observations.
Host specificity
The host specificity of the insect was assessed by
means of no-choice tests on plant species related to
yellow starthistle, including US native and US commercial crops.
ings and richness of trichomes), it was impossible to
count eggs. Thus, we indirectly evaluated oviposition
success by retrieving emerged nymphs from stems exposed to insects. In 2006, insects were tested on leaves
instead of stems. In fact, preliminary trials carried out
at the beginning of the season showed that T. grisea is
able to lay eggs both on leaves and on stems. In this
way, we were able to see eggs by observing leaves under the stereo microscope with backlighting. Leaves
with eggs were stored in a plastic box on tissue paper
until nymph emergence. Each pair was kept on yellow
starthistle before and after being tested on any other
plant species in order to give the insects the possibility
to feed on the host plant and to be sure that females
were actually ovipositing. Replicates on test plants
were considered invalid if the continuity of a females
oviposition ability could not be demonstrated on C. solstitialis after a replicate with zero eggs on any non-host
plant. Two to 11 specimens for each plant species were
tested in 2004 and 2006; the plants tested are listed in
Table 1.
Larval transfer experiment
Nymphs of T. grisea (two to six per replicate) were
transferred to intact leaves of potted C. solstitialis and
other test-plant species, confined in transparent plastic
cylinders at 23C to 26C and 16:8 h L/D. The first
observation occurred after 7 days in order to assess
nymphal development and mortality. Afterwards, observations were carried out every 3 to 4 days until all the
nymphs either reached the adulthood or died. For each
plant species, we tested an average of five specimens.
Plants tested from 2004 to 2006 are listed in Table 2.
Results and discussion
Life cycle
Life-cycle observations, carried out during laboratory rearing and oviposition and larval transfer trials,
show that under laboratory conditions (23C to 26C,
16:8 h L/D), first-instar nymphs emerged 10 to 12 days
after oviposition. The duration of the first and second
larval stages was approximately 3 to 4 days, while the
development of each stage, from third to fifth instars,
took 7 to 8 days. Total development was approximately
31 days (Figure 1).
No-choice oviposition experiment
No-choice oviposition experiment
In 2004, stems of C. solstitialis (US biotype) and
other test plants were exposed to a pair of insects in
transparent cylinders at 23C to 26C and 16:8 h L/D.
After 3 to 4 days, insects were removed and stems observed under the stereo microscope in order to count
eggs. Because of the extremely small size of egg opercula and the complexity of stem features (tissue fold-
Results of no-choice oviposition tests, performed in
2004 and in 2006, clearly showed that T. grisea oviposits most on C. solstitialis and, with limited success, on
closely-related species (Table 1), including Centaurea
stoebe, Centaurea cyanus, Centaurea diffusa, Centaurea sulphurea and Acroptilon repens. In 2004, oviposition occurred on only three of the eight plant species
tested. The number of larvae that emerged per replicate
190
Table 1.
Summary of no-choice oviposition tests carried out in 2004 and 2006.
2004 No-choice oviposition
Number of Number of
plants tested emerged
larvae
Subtribe: Centaureinae
Acroptilon repens
Carthamus tinctorius Linoleic
Carthamus tinctorius Oleic
Crupina vulgaris
Genus: Centaurea
Centaurea cyanus
Centaurea diffusa
Centaurea stoebe
Centaurea melitensis
Centaurea solstitialis
Centaurea sulphurea
Subtribe: Carduinae
Carduus
pycnocephalus
Cirsium brevistylum
Cirsium hydrophilum
Cirsium loncholepis
Cirsium occidentale
Cynara scolymus
Tribe: Heliantheae
Helianthus annuus
Valid
replicates
Non-valid
replicates
2006 No-choice oviposition
Number of
larvae/valid
replicates
Number of
plants tested
Number of
eggs laid
Number of
emerged
larvae
% emerged
larvae
Valid
replicates
Non-valid
replicates
Number of
eggs/valid
replicates
Number of
larvae/valid
replicates
3
6
5
0
0
0
0
0
3
2
0
5
1.7
0
0
0
1.3
2.5
5
3
4
6
1
1
25
17
2
3
4
0
2
2
0.5
0.3
16
22
25
19
0.9
3
19
2
0
266
3
0
106
0
0
40
0
2
59
2
1
19
0
0
4.5
1.5
0
1.8
0
2
2
3
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
3
2
4
0
0
0
0
0
0
0
0
0
0
Table 2.
Summary of the larval transfer results conducted during 2004 to 2006.
Stage of larvae transferred a
L3 or L4
L1 or L2
L1
Total larvae
transferred
12
12
4
59
64
28
20
55
10
20
Number of
plants tested
Number of
larvae transferred
% developed
L3
% developed
adult
Number of
plants tested
Number of
larvae transferred
% developed
adult
5
6
3
30
33
18
0
6
0
0
0
0
7
6
1
29
31
10
0
3
0
2
5
2
2
20
27
10
10
0
56
30
30
0
41
0
0
28
29
10
2
12
2
4
32
81
63
2
3
15
20
73
25
8
3
47
20
1
2
2
2
1
2
4
10
10
10
10
10
10
26
0
10
0
0
20
0
0
0
0
0
0
0
0
0
10
10
26
26
10
10
36
10
27
First-instar nymphs, L2 second-instar nymphs, L3 third-instar nymphs, L4 fourth-instar nymphs.
4
4
16
16
0
0
10
10
1
2
6
6
1
2
5
17
XII International Symposium on Biological Control of Weeds
192
Subtribe: Centaureinae
Acroptilon repens
Carthamus tinctorius - Linoleic
Carthamus tinctorius - Oleic
Crupina vulgaris
Genus: Centaurea
Centaurea americana
Centaurea cyanus
Centaurea diffusa
Centaurea stoebe
Centaurea melitensis
Centaurea rothrockii
Centaurea solstitialis-CA
Centaurea sulphurea
Subtribe: Carduinae
Carduus pycnocephalus
Cirsium brevistylum
Cirsium cymosum
Cirsium hydrophilum
Cirsium loncholepis
Cirsium occidentale
Cirsium vinaceum
Cynara scolymus
Tribe: Heliantheae
Helianthus annuus
Total plants
tested
A lace bug as biological control agent of yellow starthistle, Centaurea solstitialis L. (Asteraceae)
Figure 1.
Diagram of estimated development time of immature stages of Tingis grisea under laboratory conditions (23 + 3C, 16:8 h L/D photoperiod).
was 0.9 on C. solstitialis, 1.3 on C. cyanus and 2.5 on
C. stoebe. The total number of nymphs emerged was
considerably higher on yellow starthistle, although it
was very low compared to the number of replicates carried out (Table 1). Lack of knowledge about development times of the insect in this preliminary phase, in
addition to a relatively low survival rate of the eggs on
cut stems, was probably responsible for this low number
of nymphs recorded.
In 2006, females laid nearly 94% of their eggs on
yellow starthistle, and 40% of them produced nymphs.
The relatively low emergence rate can be attributed
to low egg survival on cut leaves due to rapid withering and occurrence of mould. Although we found eggs
on C. cyanus, C. diffusa, C. sulphurea and A. repens,
nymphs emerged only from eggs laid on C. cyanus and
C. diffusa. In both C. sulphurea and A. repens, leaves
with eggs became mouldy several days after oviposition, thus further experiments are needed to improve
the measure of nymphal emergence on these plants.
Larval transfer experiment
In larval development no-choice tests carried out
from 2004 to 2006, nymphal survivorship was greatest
on yellow starthistle (Table 2): 63% of first and second instar nymphs and 73% of third and fourth instar
nymphs reached the adult stage on C. solstitialis, while
a smaller percentage was able to complete the development on a small number of closely related species
(C. cyanus and C. sulphurea). In general, development
and survival was greater for old nymphs than for young
ones, on both the target and nontarget plants. We observed development of one adult from 3rd instar on an
oleic variety of Carthamus tinctorius and one on Cynara scolymus, but none of the younger (1st and 2nd
instar nymphs) became adults. This insect is not a pest
of either of these crops, which suggests that transfer of
1st and 2nd instars is more valid than transfer of 3rd
and 4th instars.
Conclusions
Preliminary host-specificity results, obtained from
2004 to 2006, showed clear oligophagy by T. grisea.
Among the species on which oviposition occurred (C.
cyanus, C. stoebe, C. diffusa, C. sulphurea and A. repens), the target weed C. solstitialis was clearly preferred in terms of number of eggs laid and number of
nymphs emerged. In addition, only a few nymphs that
were transferred to non-target plants completed development on species closely related to yellow starthistle
(C. cyanus and C. sulphurea). 3rd instar nymphs could
sometimes develop on critical nontarget plants, such as
C. tinctorius and C. scolymus, but transfer of 3rd and
4th instar nymphs represents an extreme situation that
is not likely to occur in the field because the nymphs
are not highly mobile. Failure of young larvae to develop on C. americana is very promising because this
is the closest native North American relative to the target weed.
Further feeding and oviposition trials under nochoice and choice conditions are required to better define the host range of this insect and understand if it
represents a good candidate for the biological control
of C. solstitialis. Moreover, laboratory tests and openfield trials are needed to evaluate its impact on the target species.
Acknowledgements
We are grateful to Levent Gltekin and Gksel Tozlu,
Atatrk University (Erzurum, Turkey), for their support in field collections.
References
Cristofaro, M., Hayat, R., Gultekin, L., Tozlu, G., Zengin,
H., Tronci, C., Lecce, F., Sahin, F. and Smith, L. (2002)
Preliminary screening of new natural enemies of yellow
193
XII International Symposium on Biological Control of Weeds
starthistle, Centaurea solstitialis L. (Asteraceae) in Eastern Anatolia. In: Proceeding of the Fifth Turkish National
Congress of Biological Control. 47 September 2002, Erzurum, Turkey, pp. 287295.
Cristofaro, M., Dolgovskaya, M.Yu, Konstantinov, A., Lecce,
F., Reznik, S.Ya., Smith, L., Tronci, C. and Volkovitsh,
M.G. (2004) Psylliodes chalcomerus Illiger (Coleoptera:
Chrysomelidae: Alticinae), a flea beetle candidate for
biological control of yellow starthistle Centaurea solstitialis. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings of
the XI International Symposium on Biological Control of
Weeds. CSIRO Entomology, Canberra, Australia, pp. 75
80.
Di Tomaso, J., Kyser, G.B. Pitcairn, M.J. (2006) Yellow
starthistle Management Guide. California Invasive Plant
Council. Publ. no. 2006-03. 74p.
Dostl, J. (1976) Centaurea L. In: Tutin, G., Heywood,
V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M. and Webb, D.A. (eds) Flora Europaea, Vol. 4.
Cambridge University Press, Cambridge, UK, pp. 254
301.
Fisher, A.J., Bruckart, W.L., McMahon, M.B., Luster, D.G.
and Smith, L. (2006) First report of Puccinia jaceae var.
solstitialis pycnia on yellow starthistle in the United
States. Plant Disease 90, 1362.
Klokov, M.B., Sonsovskii, D.I., Tsvelev, N.N. and Cherepanov C.K. (1963) Centaurea. Flora URSS XXVIII,
370579. Institutum Botanicum nomine V. Komarovii
Academiae Scientiarum URSS. Editio Academiae Scientiarum Moscow, URSS.
Komarov, V.L. (ed.) (1934) Flora of the U.S.S.R. Akademiya
Nauk SSSR. Botanicheskii institut 28, 571573.
Pricart, J. (1983) Faune de France: France et Rgions Limitrophes. Ch. 69. Hmiptres Tingidae Euro-Mditerranens. Paris: Fderation Franaise des Socits de Sciences
Naturelles. p 618.
Pitcairn, M.J., Woods, D.M., Joley, D.B., Turner, C.E. and
Balciunas, J.K. (2000) Population buildup and combined
impact of introduced insects on yellow starthistle (Centaurea solstitialis L.) in California. In: Spencer, N.R.
(ed.) Proceedings of the X International Symposium on
Biological Control of Weeds. Montana State University,
Bozeman, MT, pp. 747751.
Pitcairn, M.J., Piper, G.L. and Coombs, E.M. (2004) Yellow
starthistle. In: Coomb, E.M., Clark, J.K., Pipe, G.L. and
Confrancesco, A.F., Jr (eds) Biological Control of Invasive Plants in the United States. Oregon State University
Press, Corvallis, pp. 421435.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, J.
(2006) Yellow starthistle continues its spread in California. California Agriculture 60(2), 8390.
Sheley, R.L., Larson, L.L. and Jacobs, J.J. (1999) Yellow
starthistle. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State Univ. Press., Corvallis, OR, pp. 408416.
Smith, L. (2004) Prospective new agents for biological control of yellow starthistle. In: Proceedings 56th Annual
California Weed Science Society. 1214 January, 2004,
California Weed Science Society, Salinas, CA, pp. 136
138.
Smith, L. (2007) Physiological host range of Ceratapion basicorne, a prospective biological control agent of Centaurea
solstitialis (Asteraceae). Biological Control 41, 120133.
Stusak, J.M. (1957) Beitrag zur Kenntuis der Eier der Tingiden (Heteroptera, Tingidae). Beitrge zur Entomologie
7, 2028.
Stusak J.M. (1959) Early stages of the lace bug Tingis grisea
Germar (Hemiptera-Heteroptera, Tingidae). Casopis Ceskoslovenske Spolecnosti Entomologiske, Acta Societatis
Entomologicae Cechosloveniae 56(2), 181200.
Susanna, A., Garcia-Jacas, N., Soltis, D.E. and Soltis, P.S.
(1995) Phylogenetic relationships in tribe Cardueae (Asteraceae) based on ITS sequences. American Journal of
Botany 82, 10561068.
Turner, C.E., Johnson, J.B. and McCaffrey, J.P. (1995) Yellow
starthistle. In: Nechols, J.R., Andres, L.A., Beardsley, J.W.,
Goeden, R.D. and Jackson, C.G. (eds) Biological Control
in the Western United States: Accomplishments and benefits of regional research project W-84, 19641989. University of California, Division of Agriculture and Natural
Resources, Oakland Publ. 3361, pp. 270275.
Wagenitz, G. (1975) 79. Centaurea L. In: Davis, P.H. (ed.)
Flora of Turkey and the East Aegean Islands, Vol. 5. Edinburgh University Press, Edinburgh, pp. 465585.
Woods, D.M. (2004) Development and release of a plant
pathogen as a new biological control of yellow starthistle.
Cal-IPC News 12(2), 67.
194
Pathogens from Brazil for classical
biocontrol of Tradescantia fluminensis
O.L. Pereira,1 R.W. Barreto1 and N. Waipara2
Summary
Tradescantia fluminensis Vell., also known as wandering Jew, is an herbaceous monocot native to
South America. It is an invasive plant in New Zealand and the south-eastern United States where it
is considered highly invasive by the Florida Exotic Pest Plant Council. The pathobiota of T. flumi
nensis in Brazil is almost unknown and could include phytopathogenic microorganisms that could be
used in classical biological control programs. A survey for specialized, coevolved phytopathogenic
microorganisms of T. fluminensis was initiated in 2003. Five fungal species have been collected including three basidiomycetesa rust fungus (Uredo sp.), Kordyana tradescantiae (Pat.) Racib. and
Ceratobasidium sp.; a hyphomyceteCercospora apii Fresen. and an ascomyceteMycosphaerella
sp. A bacterial disease was also observed and the bacterium identified as Burkholderia andropogonis
(Smith, 1911), based on morphological, biochemical and molecular methods. Its pathogenicity to
T. fluminensis was confirmed, and a host-range test was performed. Unfortunately, results indicated
that the bacterium is not sufficiently host-specific for classical introductions. Observations of the
damage caused by fungal pathogens in the field suggest that those with the best potential as biological
control agents are Uredo sp., K. tradescantiae and Mycosphaerella sp.
Keywords: classical biological control, invasive weed, plant disease.
Introduction
Tradescantia fluminensis Vell. (wandering Jew; local
name in Braziltrapoeraba) is one among a series of
weed species of world importance belonging to the
Commelinaceae. It is native to South America and is
particularly abundant along the coast in Southeastern
and Southern Brazil where it forms small patches on
humid rocky habitats such as along creek margins. It
never forms dense extensive populations, and it is not
regarded as a weed of importance in Brazil. Conversely,
in situations where it was introduced into exotic tropical and subtropical regions of the world, it became a
very serious invader of native ecosystems. It is ranked
among the most invasive species of Florida (FLEPPC,
2003) and is particularly harmful to forest ecosystems
in New Zealand, affecting invertebrate communities
Departamento de Fitopatologia, Universidade Federal de Viosa, Vi
osa, MG, 36570-000, Brazil.
2
Manaaki Whenua Landcare Research, 261 Morrin Road, Tamaki
Campus, University of Auckland, Private Bag 92170, Auckland, New
Zealand.
Corresponding author: O.L. Pereira <[email protected]>.
CAB International 2008
1
(Toft et al., 2001; Standish, 2004), hampering natural
processes of forest regeneration and nutrient cycling
(Standish et al., 2001, 2004; Standish, 2002). It has no
significant natural enemies (arthropods or pathogens)
in New Zealand (Winks et al., 2003). Surveys of fungal pathogens of native weeds in Brazil have yielded a
plethora of potential biological control agents over the
years (Barreto et al., 1995. 1999a,b; Barreto and Evans,
1994; Barreto and Torres, 1999; Monteiro et al., 2003;
Pereira and Barreto, 2000, 2005; Soares and Barreto,
2006; Seixas et al., 2007). Two of the fungal pathogens
found during these surveys have been introduced from
Brazil into other parts of the world, namely: Prospo
dium tuberculatum (Speg.) Arthur for the biological
control of Lantana camara L. (Barreto et al., 2001a;
Ellison et al., 2006) in Australia and Colletotrichum
gloeosporioides f.sp. miconiae for the biological control of Miconia calvescens Schrank and Mart. ex DC.
in Hawaii (Barreto et al., 2001b). Although Brazil is
considered to be the centre of origin of T. fluminensis,
there is not a single pathogen recorded to be associated
with this plant species in this country in the world literature (Table 1). A cooperative research project recently
initiated between the Universidade Federal de Viosa
(Brazil) and Manaaki Whenua Landcare Research New
195
XII International Symposium on Biological Control of Weeds
Table 1.
ungal pathogens recorded on Tradescantia fluminensis in the literaF
ture (Petrak, 1950; Gmez and Kisimova-Horovitz, 1997; Waipara,
2006; Farr et al., 2007).
Fungal species
Country
Alternaria sp.
Botrytis cinerea
Cercospora sp.
Cladochytrium replicatum
Colletotrichum sp.
Phakopsora tecta
Pythium sp.
Rhizoctonia sp.
Sclerotinia sclerotiorum
Kordyana tradescantiae
USA
USA
USA
USA
USA
Argentina
Hawaii
USA
New Zealand
Ecuador, Costa Rica
Zealand Ltd. is aimed at surveying and evaluating the
native pathobiota associated with T. fluminensis in Brazil for potential classical biological control agents. This
paper gives a preliminary account of the pathogens
found during these surveys and their potential as classical biological control agents for T. fluminensis.
Materials and methods
The field survey was systematic and involved all states
of Southern Brazil. Records of T. fluminensis were
compiled from eight Brazilian herbaria. The following
Southern and Southeastern Brazilian states were visited
during January 2003 and December 2005: Minas Gerais,
Rio de Janeiro, Esprito Santo, So Paulo, Paran, Santa
Catarina and Rio Grande do Sul. Further details of
the procedure adopted for the systematic survey can
be found in Barreto and Evans (1994). The diseased
parts of the plants suspected to be damaged by fungal
or bacterial pathogens were collected, dried in a plant
press and taken to the laboratory. Seedlings infected by
biotrophic fungi were also brought to the laboratory in
Viosa (MG). Fungal structures were removed from
specimens and mounted in lactophenol or lactofucsin.
Observations of morphology, measurements and illustrations were carried out with an OLYMPUS BX 50
light microscope fitted with a drawing tube. Isolations
were conducted by collecting spores from sporulating
lesions with a fine pointed needle and plating them on
Vegetal Broth Agar medium (Pereira et al., 2003). The
isolates of non-biotrophic fungi were stored on silica
gel according to Dhingra and Sinclair (1995). The materials examined were deposited in the herbarium at
the Universidade Federal de Viosa (Herbarium VIC).
Additional materials previously deposited at VIC were
also examined.
Preliminary pathogenicity experiments were conducted for all the basidiomycetous fungi found, i.e.
Ceratobasidium sp., K. tradescantiae (Pat.) Racib. and
Uredo sp. T. fluminensis plants originating from Brazil
or imported from New Zealand (NZ) were used in the
pathogenicity experiments. To verify the pathogenicity
of K. tradescantiae, the fungus was cultivated in MelinNorkrans modified medium (MNM) and incubated in
the dark at 25C. After 10 days, sporidia were collected
by pouring 30 ml of sterile water on the culture surface and scraping it with a rubber spatula. The resulting
suspension was filtered through four layers of cheese
cloth, and the final concentration of the suspension was
adjusted to 1 107 sporidia/ml for inoculation. The cell
suspension was sprayed on the leaf surface (abaxially
and adaxially) without wounding. After inoculation, 10
plants were covered for 48 h with plastic bags wetted
inside and having water-soaked cotton internally and
left at room temperature (approximately 25C). After
that period, the plastic bags were removed, and plants
were maintained in a greenhouse (26 2C) and watered daily. Ten non-inoculated healthy plants, kept
under the same conditions, served as controls. For the
biotrophic fungi Ceratobasidium sp. and Uredo sp., ten
healthy potted T. fluminensis plants imported from NZ
were cultivated side-by-side (pots kept 5 cm apart) with
diseased plants collected during field surveys. Plants
were kept for 1 year on a shaded bench outdoors and
watered regularly.
Results
Five fungal species and a bacterium, collected in four
different states, were found associated with diseased
T. fluminensis: three basidiomycetesa rust fungus
(Uredo sp.), K. tradescantiaecausing white smutlike symptoms and the blight-causing-fungus Cerato
basidium sp.; a leaf-spot- and stem necrosis-causing
hyphomyceteCercospora apii Fresen. and an ascomycete that is associated with leaf-spotsMycosphae
rella sp. (Table 2). The phytopathogenic bacterium
was identified as Burkholderia andropogonis (Smith,
1911). Three of the fungal pathogens were isolated in
culture: K. tradescantiae, C. apii and Mycosphaerella
sp. Repeated attempts to isolate Ceratobasidium sp. associated with leaf blight were unsuccessful. We believe
that this fungus is in fact a biotroph, since often even a
complete colonization of the abaxial surface of leaves
196
Pathogens from Brazil for classical biocontrol of Tradescantia fluminensis
Table 2.
ungal pathogens found on Tradescantia fluminensis during field
F
surveys in Brazil.
Fungal species
Distribution in Brazilian statesa
Ceratobasidium sp.
Cercospora apii
Mycosphaerella sp.
Uredo sp.
Kordyana tradescantiae
SC (1); RS (3)
MG (2); SC (1); PR (2)
RS (1)
PR (1); SC (1); RS (2)
PR (3); SC (3); RS (9)
The numbers in parentheses represent how many times each fungus was collected in a
state (MG Minas Gerais; PR Paran; RS Rio Grande do Sul; SC Santa Catarina).
(easily observed by an extensive external coverage of
the tissues by a mycelial mat) was not accompanied by
any sign of necrosis. Attempts to isolate K. tradescan
tiae on vegetable broth-agar (VBA) were unsuccessful. Several other culture media were also tried such as
potato-dextrose agar (PDA), corn-meal-agar (CMA),
potato-carrot agar (PCA; Dhingra and Sinclair, 1995)
but also failed to promote any fungal growth. K. trad
escantiae was finally successfully isolated on MNM,
a culture medium commonly used for ectomycorrhizal
basidiomycetous fungi (Marx, 1969).
No disease symptoms were observed on plants from
Brazil or NZ inoculated with K. tradescantiae, and no
symptoms were observed to spread when placing Brazilian plants infected by Uredo sp. and Ceratobasidium
sp. beside healthy NZ plants after over 1 year of observation.
Discussion
Kordyana tradescantiae and Uredo sp. are reported for the first time on T. fluminenis in Brazil. K.
tradescantiae has been reported on T. fluminensis only
from Ecuador (Petrak, 1950) and Costa Rica (Gmez
and Kisimova-Horovitz, 1997), and no rusts had been
reported on T. fluminensis in Brazil (Hennen et al.
2005). Although in some field situations these biotrophic basidiomycetes caused no severe disease symptoms on T. fluminensis, on other occasions (particularly
under heavily shaded areas), damage was significant.
Diseased plants appeared weakened and defoliated as
compared to healthy T. fluminensis plants. It seems
that both fungalspecies are promising candidates for
the classical biological control of T. fluminensis (Table
3). Although K. tradescantiae is known to attack plants
Table 3.
belonging to several genera in the Commelinaceae,
this is not a limitation for its use as a biological control
agent in New Zealand where there is not any native
plant nor any crop plant of relevance belonging to this
family. The other basidiomycete, Ceratobasidium sp.,
caused no significant damage in the field. It still remains
unclear whether the slight blight symptoms appearing
on colonized leaves only represent naturally senescent
leaves or become necrotic because of the fungus infection. This species does not appear to deserve further
consideration as a possible candidate agent (Table 3).
It is nevertheless interesting to note that no Cerato
basidium species has previously been reported on the
genus Tradescantia, and no other species in this genus
was ever reported as a foliar biotroph (Roberts, 1999).
Based on the morphological characteristics observed
during this study, Ceratobasidium sp. was recognized
as a new species that will be described separately.
C. apii and Mycosphaerella sp. caused severe necrotic disease symptoms on leaves and stems of T.
fluminensis in the field. Crous and Braun (2003) listed
numerous hosts belonging to many distinct plant fam
ilies for C. apii; however, this is the first report of C.
apii on T. fluminensis. Despite its supposedly wide
host range, we are planning to conduct host range tests
based on the centrifugal phylogenetic method (Wapshere, 1974) to evaluate the specificity of this isolate of
C. apii. It is known that specificity can exist in populations within a fungal species known to have a wide
host range (Barreto et al., 2001b; Pereira et al., 2003).
In case this isolate prove to be host-specific, C. apii
may deserve further consideration for use in classical
biological control.
No Mycosphaerella spp. was ever reported attacking
members of Tradescantia. A study of the morphology
Characteristics of fungal pathogens found on Tradescantia fluminensis during field surveys in Brazil.
Fungal species
Disease
Damage to host
Likely specificity
Cultured
Biological control potential
Ceratobasidium sp.
Cercospora apii
Mycosphaerella sp.
Uredo sp.
Kordyana tradescantiae
Faint leaf-blight
Leaf-spot
Leaf-spot
Rust
White smut
insignificant
Significant
Significant
Significant
Significant
High
Non-specific
High
High
Low (on Com
melinaceae)
No
Yes
Yes
No
Yes
Low
Uncertain
High
High
High
197
XII International Symposium on Biological Control of Weeds
of Mycosphaerella sp. from T. fluminensis indicates
that this is a new species which will be described elsewhere. As it causes severe necrotic disease symptoms
on T. fluminensis and members of Mycosphaerella are
often host-specific, this fungus is being considered as a
promising candidate for the classical biological control
of T. fluminensis (Table 3).
The preliminary pathogenicity tests with K. trades
cantiae, Uredo sp. and Ceratobasidium sp. were unsuccessful. No disease symptoms were reproduced through
artificial inoculation (with K. tradescantiae). Perhaps
sporidia of this fungus produced in culture are noninfective, and the fungus relies on basidiospores as an
infective stage or, perhaps, the fungus loses pathogenicity when cultivated. Nevertheless, it was observed
that this disease was naturally transmitted from diseased to healthy plants in Viosa. This did not happen
with either Uredo sp. or Ceratobasidium sp.
It was observed that plants infected with Uredo
sp. and Ceratobasidium sp., brought from the field
in Southern Brazil, were gradually cured of infection
along the months after being cultivated under Viosa
conditions at a lower latitude. This suggests that those
two fungal species depend on a cooler climate to produce new infection cycles and, therefore, to preserve
their populations.
The bacterium collected in the state of Rio de Janeiro
was identified as B. andropogonis, and its pathogenicity was demonstrated. Inoculated plants of the biotype
brought from New Zealand were highly susceptible to
this pathogen, and plant death commonly resulted from
inoculations. Unfortunately, host-range tests indicated
that, although restricted to monocots, this bacterial isolate was capable of infecting all eight species of the
Commelinaceae included in the test plus species in
five additional families, i.e. pinnaple (Bromeliaceae),
Paepalanthus macrocephalus (Eriocaulaceae), maize
and sorghum (Poaceae), cattail (Typhaceae) and ginger (Zingiberaceae). Although B. andropogonis is not
known to be a pathogen of maize or sorghum in Brazil,
further studies are needed to fully clarify the risk represented by B. andropogonis to crop and non-crop plants.
Until then, its potential for introduction into New Zealand or other regions of the world or its use as a bioherbicide cannot be considered any further.
Until now, only a limited area of the Neotropics was
surveyed for T. fluminensis pathogens. It is expected
that the continuation of the surveys with the expansion
to new areas of natural occurrence of T. fluminensis
will result in new additions to this still-limited list of
pathogens. In addition, the potential for biological control of the pathogens already found is being evaluated.
One ongoing study aims at checking the susceptibility
of the New Zealand biotype of T. fluminensis to the rust
and to K. tradescantiae. Stations with sentinels (potted
plants of the New Zealand biotype) were established at
selected places where these pathogens occur on native
populations of T. fluminensis in Southern Brazil. These
will be visited at 3-month intervals for inspection of
natural attack by pathogens or arthropods.
Genetic studies are also underway to both clarify
the clonal status of T. fluminensis in New Zealand as
well as compare some DNA regions of the NZ biotype
with those found in the native Brazilian range. Chloroplast DNA from over 40 plant specimens from different
regions of Brazil are being sequenced to obtain information on the plants lineages that may help to locate
where in Brazil the NZ biotype has originated. These
results may then be used to source suitable biotypes of
each pathogen agent.
Acknowledgements
This work forms part of a research project submitted
as a DSc dissertation to the Departamento de Fitopatologia/Universidade Federal de Viosa by O.L. Pereira.
The authors thank the Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq), Brazil
and the National Biocontrol of Weeds Collective in
New Zealand for financial support.
References
Barreto, R.W. and Evans, H.C. (1994) The mycobiota of the
weed Chromolaena odorata in southern Brazil with particular reference to fungal pathogens for biological control. Mycological Research 98, 11071116.
Barreto, R.W., Evans, H.C. and Ellison, C.A. (1995) The
mycobiota of the weed Lantana camara in Brazil, with
particular reference to biological control. Mycological Re
search 99, 769782.
Barreto, R.W., Evans, H.C. and Hanada, R.E. (1999a) First
record of Cercospora pistiae causing leaf spot of water
lettuce (Pistia stratiotes) in Brazil, with particular reference to weed biocontrol. Mycopathologia 144, 8185.
Barreto, R.W., Evans, H.C. and Pomella, A.W.V. (1999b)
Fungal pathogens of Calotropis procera (rubber bush),
with two new records from Brazil. Australasian Plant Pa
thology 28, 126130.
Barreto, R.W., Pereira, J.M., Ellison, C. and Thomas, S.
(2001a) Controvrsia e fundamentao cientfica para a
introduo da ferrugem P. tuberculatum como agente de
controle biolgico para Lantana camara na Austrlia. In:
7th Simpsio de Controle Biolgico. Poos de Caldas,
Brazil, p. 148.
Barreto, R.W., Seixas, C.D.S. and Killgore, E. (2001b) Col
letotrichum gloeosporioides f.sp. miconiae: o primeiro
fungo fitopatognico brasileiro a ser introduzido no exterior
para o controle biolgico clssico de uma planta invasoras
(Miconia calvescens). In: 7th Simpsio de Controle Biol
gico. Poos de Caldas, Brazil, p. 109.
Barreto, R.W. and Torres, A.N.L. (1999) Nimbya alternan
therae and Cercospora alternantherae: two new records
of fungal pathogens on Alternanthera philoxeroides (allegatorweed) in Brazil. Australasian Plant Pathology 28,
103107.
198
Pathogens from Brazil for classical biocontrol of Tradescantia fluminensis
Crous, P.W. and Braun, U. (2003) Mycosphaerella and its ana
morphs: 1. Names published in Cercospora and Passalora.
CBS Biodiversity Series, CBS, Utrecht, the Netherlands,
571 pp.
Dhingra, O.D. and Sinclair, J.B. (1995) Basic Plant Pathol
ogy Methods. CRC Press. Boca Raton, FL, 434 pp.
Ellison, C.A., Pereira, J.M.; Thomas, S., Barreto, R.W. and
Evans, H.C. (2006) Studies on the rust Prospodium tu
berculatum, a new classical biological control released
against the invasive weed Lantana camara in Australia.
1. Life-cycle and infection parameters. Australasian plant
pathology 35, 309319.
Farr, D.F., Rossman, A.Y., Palm, M.E. and McCray, E.B.
(nd) Fungal Databases, Systematic Botany & Mycology
Laboratory, ARS, USDA. Available at: http://nt.ars-grin.
gov/fungaldatabases/ (accessed March 21, 2007).
FLEPPC (2003) List of Floridas Invasive Species. Florida
Exotic Pest Plant Council. Available at: http://www.fleppc.
org/03list.htm.
Gmez, L.D. & Kisimova-Horovitz, L. (1997) Basidiomicetos de Costa Rica. Exobasidiales, Cryptobasidiales. Notas
histricas, taxonmicas y filogeogrficas. Revista de Bio
logia Tropical 45, 12931310.
Hennen, J.F., Figueiredo, M.B., Carvalho, A.A. and Hennen, P.G. (2005) Catalogue of species of plant rust fungi
(Uredinales) of Brazil. Available at: http://www.jbrj.gov.
br/publica/uredinales/Brazil_Catalogue1drevisado.pdf
(verified 12th April 2007).
Marx, D.H. (1969) The influence of ectotrophic mycorrhizal fungi on the resistence pine roots to pathogenic infections I. Antagonism of ectomycorrhizal fungi to roots
pathogenic fungi and soil bacteria. Phytopathology 59,
153163.
Monteiro, F.T., Vieira, B.S. and Barreto, R.W. (2003) Curvu
laria lunata and Phyllachora sp.: two fungal pathogens of
the grassy weed Hymenachne amplexicaulis from Brazil.
Australasian Plant Pathology 32, 449453.
Pereira, J.M. and Barreto, R.W. (2000) Additions to the mycobiota of the weed Lantana camara (Verbenaceae) in
southeastern Brazil. Mycopathologia 151, 7180.
Pereira, J.M., Barreto, R.W., Ellison, A.C. and Mafia, L.A.
(2003) Corynespora casiicola f. sp. lantanae: a potential
biocontrol agent from Brazil for Lantana camara. Biolog
ical Control 26, 2131.
Pereira, O.L. and Barreto, R.W. (2005) The mycobiota of the
weed Mitracarpus hirtus in Minas Gerais (Brazil) with
particular reference to fungal pathogens for biological
control. Australasian Plant Pathology 34, 4150.
Petrak, F. (1950) Beitrge zur pilzflora von Ekuador. Sydowia
4, 450587.
Roberts, P. (1999) Rhizoctonia-forming fungi. Royal Botanic
Gardens, Kew, UK, 239 pp.
Seixas, C.D.S., Barreto, R.W. and Killgore, E. (2007) Fungal pathogens of Miconia calvescens (Melastomataceae)
from Brazil, with reference to classical biological control.
Mycologia 99, 99111.
Soares, D.J. and Barreto, R.W. (2006) Additions to the Brazilian mycobiota of the grassy weed, Hymenachne am
plexicaulis, with a discussion on the taxonomic status of
Paraphaeosphaeria recurvifoliae. Australasian Plant Pa
thology 35, 347353.
Standish, R.J. (2002) Experimenting with methods to control
Tradescantia fluminensis, an invasive weed of native forest remnants in New Zealand. New Zealand Journal of
Ecology 26, 16170.
Standish, R.J. (2004) Impact of an invasive clonal herb on
epigaeic invertebrates in forest remnants in New Zealand.
Biological Conservation 116, 4958.
Standish, R.J., Robertson, A.W. and Williams, P.A. (2001)
The impact of an invasive weed Tradescantia fluminensis
on native forest regeneration. Journal of Applied Ecology
38, 12531263.
Standish, R.J., Williams, P.A., Robertson, A.W., Scott, N.A.
and Hedderley, D.I. (2004) Invasion by a perennial herb
increases decomposition rate and alters nutrient availability in warm temperate lowland forest remnants. Biologi
cal Invasions 6, 7181.
Toft, R.J., Harris, R.J. and Williams, P.A. (2001) Impacts of
the weed Tradescantia fluminensis on insect communities
in fragmented forests in New Zealand. Biological Conser
vation 102, 3146.
Waipara, N. (2006) Isolation of white rot, Sclerotinia sclero
tiorum, causing leaf necrosis on Tradescantia fluminen
sis in New Zealand. Australasian Plant Disease Notes 1,
2728.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 20111.
Winks, C.J., Waipara, N.W., Gianotti, A. and Fowler, S.V.
(2003) Invertebrates and Fungi Associated with Trades
cantia fluminensis (Commelinaceae) in New Zealand. De
partment Of Conservation Report, 122.
199
Field and laboratory observations of the life
history of the Swiss biotype of Longitarsus
jacobaeae (Coleoptera: Chrysomelidae)
K.P. Puliafico,1 J.L. Littlefield,2 G.P. Markin3 and U. Schaffner4
Summary
The Italian biotype of ragwort flea beetle, Longitarsus jacobaeae (Waterhouse) (Coleoptera: Chrysomelidae) is considered the most important biological control agent for the suppression of tansy ragwort, Senecio jacobaea L. (Asteraceae), in Mediterranean climates. Repeated attempts to introduce
this beetle into colder climates have failed to establish populations capable of weed control. The
spread of tansy ragwort into the northern Rocky Mountains of Montana prompted a reexamination of
the cold-adapted biotype of L. jacobaeae reported from Switzerland. A detailed field and laboratory
examination of L. jacobaeae life history from naturally occurring populations in central Europe was
conducted. Adult flea beetles, first collected in late June, started oviposition 2 weeks after emergence
from pupae and reached peak oviposition rate after 4 weeks. Over-wintering occurs in the egg stage,
as diapause delayed hatch until spring. Larvae initially fed within the leaf tissues in early spring
and moved into the root crowns later in the season. Over 70% of second-year plants dissected were
infested with larvae in naturally occurring field populations. L. jacobaeae from Switzerland were
found to be phenologically adapted to continental climates and were released in Montana starting in
autumn 2002.
Keywords: Senecio jacobaea, Montana, ragwort flea beetle.
Introduction
The ragwort flea beetle, Longitarsus jacobaeae (Water
house) (Coleoptera: Chrysomelidae), is a specialist
herbivore that feeds on the invasive weed tansy ragwort, Senecio jacobaea L. (Asteraceae), a biennial
herb native to Eurasia (Frick, 1970). L. jacobaeae
populations originating in Italy are considered the cornerstone of the successful biological control of tansy
ragwort in the Pacific Northwest west of the Cascade
Mountains (Hawkes and Johnson, 1978; McEvoy and
Coombs, 1999). S. jacobaea population densities were
reduced by 99% in the first 4 years near Fort Bragg,
Plant, Soils, and Entomological Science, University of Idaho, Moscow,
ID 83844-2339, USA.
2
Land, Resources and Environmental Sciences, Montana State University, Bozeman, MT 59717, USA.
3
USDA Forest Service, Rocky Mountain Research Station, Forestry Sciences, Montana State University, Bozeman, MT 59717, USA.
4
CABI Europe-Switzerland, 1 Rue des Grillions, Delmont CH-2800,
Switzerland.
Corresponding author: K.P. Puliafico <[email protected]>.
CAB International 2008
1
California as a result of root feeding by L. jacobaeae
larvae (Hawkes and Johnson, 1978). The discovery of
a major infestation of tansy ragwort in northwest Montana rekindled interest in the biological control of this
noxious weed. The previous success of L. jacobaeae,
along with its compatibility with other biological control agents released in Montana (Hawkes and Johnson,
1978; Markin, 2003), made introduction of this agent
a high priority. However, repeated attempts to establish populations from Oregon into Montana from 1998
to 2002 met with little success (Markin, 2003). Italian
L. jacobaeae adults collected from Oregon emerge
from pupation in the spring, aestivate throughout the
summer and await autumn rains to start reproduction
and larval development. Adaptations for Mediterranean habitats may limit the ability of the Italian strain
to increase populations in colder climates.
This study was initiated to explore the suitability
of populations of L. jacobaeae from north-western
Switzerland for introduction in Montana. Comparisons of the autumn-breeding Italian and summerbreeding Swiss flea beetle strains were conducted under
laboratory and greenhouse conditions in Rome, Italy
200
Field and laboratory observations of the life history of the Swiss biotype of Longitarsus jacobaeae
(Frick, 1971; Frick and Johnson, 1972, 1973; Windig,
1991). However, over-wintering traits were never fieldvalidated, and the mechanisms of winter survival of
the Swiss strain were merely speculated (Windig and
Vrieling, 1996). Further, descriptions of egg diapause
may have contained experimental artifacts because of
interbreeding of strains in the laboratory (Frick and
Johnson, 1972). Our first objective was to determine
the over-wintering life stage of the Swiss flea beetles
in their native habitat and to clarify the patterns of
egg diapause under controlled conditions. We then examined the distribution and development of Swiss L.
jacobaeae under field and greenhouse conditions to
determine if their life history traits may better suit
this strain to the short summer/cold winters found in
Montanas continental climate.
Materials and methods
Studies on the ecology and seasonal phenology of
L. jacobaeae were conducted during the summers of
2000 through 2002 at four field sites in Switzerland:
LHimelette, St. Imier, Mervilier, and Mettembert
(Table 1). Sites were located on gentle to moderately
steep, south to south-western facing slopes in seasonally grazed pastureland. The plant community was
dominated by grasses, with mixed forbs occupying
less than 35% cover. Tansy ragwort occurred in scattered patches at densities of up to 30 plants per square
metre. Snow cover varied between years but extended
into mid-March only at the high-elevation LHimelette
site when spring phenologies were observed in 2001.
Laboratory experiments were conducted at the Biological Control Containment Facility at Montana State
University, Bozeman, MT. After approval for release
was granted for the Swiss populations in 2002, further
garden and greenhouse experiments were conducted at
the USDA Forest Service-Rocky Mountain Research
Station in Bozeman, Montana.
Field observations
Adult emergence from pupation was determined
through weekly vacuum samples carried out in Switzerland prior to and during the adult emergence from
pupation in JuneJuly 2001. The first field-collected
adult ragwort flea beetles were used to determine the
starting date of oviposition from different populations.
Table 1.
Adults were sexed, and paired beetles were placed in
1-l clear plastic cylinders with fresh-cut S. jacobaea
leaves inserted in 2-cm cubes of moistened floral foam
placed on top of a 90-mm filter paper. Five replicates
were used for each population. Every 2 to 3 days,
beetles were transferred to clean containers with fresh
food, and cylinders, leaves, paper, and foam were inspected for eggs.
The spatial and temporal aspects of L. jacobaeae
larval biology of Swiss populations were investigated
in 2001. Plant samples were initially collected March
15 and then, starting April 15, every 2 weeks from the
four sites. The last sampling period was July 27 at the
lower elevation sites and August 11 at LHimelette. For
each site, ten samples were collected along a wandering
transect starting at a randomly selected point in the pasture. Samples included a tansy ragwort plant and all of
the soil and other plants within a 10-cm radius. Samples
were removed using a spade and hand trowel and stored
individually in a plastic bag at 2C until dissection.
S. jacobaeae plants from each sample were categorized into five demographic groups: seedling (<5 leaves),
rosette, multiple rosettes (attached to a single root crown), bolting, and flowering. Plants were dissected, and
the locations of larvae and determination of the larval
instars were recorded. Larval instars were determined
according to the morphology and color of the anal
sclerites and the color and width of the head capsule
(Newton, 1933; Windig, 1991).
Laboratory experiments
Two experiments were conducted during the winter
of 2002 to 2003 to investigate adult emergence from
pupation. The first experiment utilized neonate larvae
from St. Imier inoculated on plants maintained in the
greenhouse, while the second experiment was conducted at ambient outdoor temperatures with eggs from
Mettembert.
Tansy ragwort plants were grown from seeds for
3 months in 3-l plastic pots in the greenhouse with a 14-h
photoperiod, and average temperatures of 25C (day)
and 20C (night). Medium-sized rosettes (average of
17.7 leaves) were randomly assigned L. jacobaeae density treatments (0, 5, 10, 15, 20 or 30 eggs or larvae per
plant) in a complete block design with five replicates.
An additional treatment of five eggs or larvae was initiated as part of each block to be destructively sampled
irst 2001 vacuum collections of adult Longitarsus jacobaeae and onset of oviposition for field-collected adults
F
from four field sites in Switzerland.
Site
Mettembert
Mervilier
St. Imier
LHimelette
Coordinates
4724 N, 720 E
4724 N, 718 E
4709 N, 659 E
4708 N, 701 E
Elevation (m)
640
660
800
1160
201
Adult emergence
23 June
29 June
7 July
13 July
Oviposition
2 July
4 July
10 July
17 July
XII International Symposium on Biological Control of Weeds
to determine larval developmental stage. Larvae were
transferred directly to plants; while cold-treated eggs
(minimum 60 days at 2C) were transferred to the base
of the rosettes on blotter paper and covered with moist
peat moss.
Plants for the second experiment were acclimatized for 3 weeks (average 10C) until the seasons first
persistent snowfall on December 2, 2002. Plants were
moved to prepared garden plots immediately after application of egg treatments, covered with snow, and
maintained outdoors until July when larval development was nearing completion.
After the larvae from destructive samples from each
experiment reached the late third instar and pre-pupal
stages, all test plants were placed in opaque emergence
cages. As adult beetles moved into clear collection
vials, the date of emergence was recorded. At the end
of the experiment, the bags were inspected for adult
beetles that had not moved into the vials.
The time required for eclosion of eggs of the Swiss
biotype from the St. Imier population was investigated
to determine the period of egg diapause. Five replicates
of ten apparently viable eggs from a single collection
date were placed on filter paper in 90 mm diameter
Petri dishes loosely sealed with paper towels. Dishes
were kept moist by applying water on the outer portion
of the paper towel and stored in plastic bags to maintain
humidity.
Treatments varied in length of exposure to cold.
The zero day treatment was immediately placed in the
growth chamber held at a constant temperature (20.0
0.5C) with a 14-h photophase. All other treatments
were cooled to 2 1C and stored for 20 to 180 days
without light. At 20-day intervals, dishes were removed
from cold storage and placed in the growth chamber.
Eggs were observed daily for hatch or signs of mortality, and all hatched and dead eggs were removed.
Remaining viable eggs were returned to the growth
chamber.
Statistical analysis was conducted using Minitab
v. 12 (Minitab Inc., 1998) with = 0.05. One-way
analysis of variance (ANOVA) was used to compare
plant demographic groups for larval feeding damage
and infestation rates. Tukeys pairwise comparisons of
demographic groups were analyzed using a family =
0.05 (individual = 0.0066). For both the garden and
greenhouse adult emergence experiments, data were
pooled across all inoculation levels and tested for normality with the AndersonDarling normality test. The
non-parametric KruskalWallis test was then used to
compare egg and larval inoculation treatments for rates
of adult emergence. In the egg diapause experiment,
the mean egg hatch from each replicate represented
samples of the larger population and were normally
distributed (A2 < 0.603, df > 4, P > 0.05). We com
pared cold treatments using the calculated parameters
of each replicate in the one-way ANOVA model. Simple
linear regression analysis was used to describe change
in eclosion rates. Comparison of regression lines were
accomplished using paired t tests.
Results
Field observations
Adult flea beetles emerged in early summer in the
2001 field surveys in Switzerland (Table 1). The first
adult L. jacobaeae flea beetles were collected during
the third week of June at 600 m sites. Adult emergence
from higher elevations was delayed approximately 1
week for every 200 m of increased elevation (Table 1).
Newly emerged adults continued to be collected for a
period of 2 to 3 weeks at low and medium elevations.
Beetles from the highest elevation site emerged over
a longer period, possibly due to variation in egg hatch
caused by delayed snowmelt in microhabitats shaded
by nearby conifer trees. First adult emergence from the
St. Imier and LHimelette sites occurred approximately
120 days after the collections of neonate larvae in
March 2001. Our results are at least a month later than
those reported by Frick (1971), but beetles transported
from Switzerland to Rome, Italy, and Albany, CA, may
have started development sooner due to exposure to
lower elevations and earlier warm weather.
Adult flea beetles collected in Switzerland began
oviposition within 2 weeks of the first adult collections
(Table 1). Populations from higher elevations were first
collected later in the season and had a corresponding
later onset of oviposition. After the initial onset of oviposition for each population, additional collections of
adults commenced oviposition immediately following
transport to the laboratory.
Larval feeding damage was apparent in 72% of
all plants dissected from Swiss field sites in 2001, although larvae were recovered from only 37% of these
plants (Figure 1). Larval presence varied significantly
among plant demographic groups (F4,395 = 5.40, P <
0.001), with seedlings and flowers having fewer larvae than rosettes, multiple-rosettes and bolting plants
(Figure 1). Seedlings comprised 15% of the plants examined but were damaged significantly less than other
plants despite their availability in all sampling periods
(Figure 1). Only first-instar larvae were recovered from
seedlings, possibly because these plants were too small
to sustain larval development beyond this stage. Bolting plants occurred later in the season (May 1 at lower
elevations) and were predominantly infested with third
larval instars. Flowering plants were first collected on
June 28 at lower elevations, coinciding with the beginning of adult emergence at these sites (Table 1).
Larvae of the Swiss L. jacobaeae had a distinct spatial distribution pattern within the plant that changed
over time as the larvae and plants developed (Figure 2).
Almost all early-season larvae (94.7%) were found in
the basal rosette leaves of S. jacobaea. Freshly hatched
larvae entered the plant through the leaf blade and fed
202
Field and laboratory observations of the life history of the Swiss biotype of Longitarsus jacobaeae
Figure 1.
Larval utilization of available tansy ragwort plants by plant demographic group by Swiss
2001 field populations of Longitarsus jacobaeae (n = total). Shaded portions of bar graphs
indicate feeding damage. Means with the same letter (capital letter, damage; lower case, larval
infestation) are statistically indistinguishable in Tukeys pairwise comparisons ( = 0.05).
Mean number larvae
3.5
3.0
2.5
2.0
1.5
1.0
0.5
Stems
Crown
11
-A
u
-J
ul
28
ul
-J
14
un
30
-J
un
16
-J
2-
Leaves
Figure 2.
Ju
r
21
-A
pr
5M
ay
19
-M
ay
Ap
7-
ar
24
-M
10
-M
ar
0.0
Roots
Mean number of Longitarsus jacobaeae larvae per tansy ragwort plant, separated by plant region, throughout the growing season in the 2001 LHimelette, Switzerland, population (1200 m).
between the epidermal layers as they moved toward the
leaf veins and petiole. Once the first instars reached a
leaf vein, they continued to feed downward to the base
of the petiole. It was common to find more than one
larva in a single leaf, with their feeding tunnels intertwined in the petiole. Later-season larvae moved into
the upper root crown (Figure 2). A small percentage
(6.1%) of the larvae of all instars was found within the
base of stems of bolting plants; however, there were
no larvae or signs of larval feeding damage in stems
above the lowest stem leaf or approximately 2.5 cm
above the basal rosette. The majority of third-instar lar-
vae (83.5%) were found within the root crown. Larval
feeding on the root crown occurred in the root cortex.
Larvae were rarely collected from the roots (0.43% of
total), and most damage to roots occurred only 1 cm
away from the root crown as third-instar larvae exited
the plant for pupation in late spring.
Laboratory experiments
L. jacobaeae larvae raised under ambient conditions
in the garden in Montana completed their development
130.7 4.8 days (n = 41) after snowmelt, with mean
203
XII International Symposium on Biological Control of Weeds
adult emergence on July 20, 2003. Adult emergence
occurred over a period of 3 weeks (Figure 3), and the
time required for emergence did not differ between eggdensity treatments (KruskalWallis test, H = 1.71, df = 4,
P = 0.788). There were also no significant differences
in percent of adult emergence from different-density
egg inoculation treatments (KruskalWallis test, H =
3.90, df = 4, P = 0.450). Development rates observed
in Montana were slightly longer than the average development rates observed in Switzerland but were well
within the range observed at higher elevation sites.
Second-generation St. Imier beetle larvae placed
directly on greenhouse plants completed their develop
ment in an average of 80.6 0.1 days (n = 61) after
inoculation. These beetles emerged over a period of
21 days (Figure 3). There was no significant differ
ences in percent of adult emergence among treatments
(KruskalWallis test, H = 1.91, df = 4, P = 0.752). The
time required for complete larval development corre
sponded with observations by Frick (1971), who found
that laboratory-reared larvae required an average of 80
days to complete development.
The Swiss ragwort flea beetles have been reported
to have a facultative egg diapause (Frick, 1971). In
our study, eggs of the Swiss biotype hatched after an
average of 65.1 2.9 days without exposure to cold
temperature when kept at a constant 20C (Figure 4),
but the eclosion period was reduced after exposure
to cold temperatures. The time required for eggs to
hatch was negatively correlated with time of exposure
to cold over the first 60 days (y = 0.733x + 67.068,
R2 = 0.938, P 0.001). Time to eclosion continued to
decrease for cold treatments of 80 days or more but at a
more gradual rate (y = 0.0516x + 19.995, R2 = 0.532,
P < 0.001). We estimated that diapause was completed after 69 days of cold exposure at the point that the
Percent Emergence
20%
Greenhouse Plants
Outdoor Plants
15%
10%
5%
0%
0
10
15
20
Days after first emergence
Adult Longitarsus jacobaeae emergence from the populations raised in the greenhouse (from St.
Imier, x = 11.6 0.8 days, n = 61 adults) and under open-field conditions at Bozeman, MT, in
2003 (from Mettembert, x = 6.7 0.1 days, n = 41 adults).
Mean incubation (days)
Figure 3.
90
80
70
60
50
y = -0.733x + 67.068
R2 = 0.9379
Estimated completion of diapause
40
30
20
10
0
y = -0.0516x + 19.995
R2 = 0.5316
0
Figure 4.
20
40
60
80 100 120 140
Cold treatment (days at 2C)
160
180
Mean incubation period (per replicate) for Swiss Longitarsus jacobaeae eggs
at 20C with 14 h photophase after removal from cold treatments (2 2C).
The regression lines are significantly different (t = 17.31, df = 46, P < 0.0001).
204
Field and laboratory observations of the life history of the Swiss biotype of Longitarsus jacobaeae
regression lines intersect (Figure 4; significant change
in regression lines: t = 17.31, df = 46, P < 0.0001).
Discussion
We found that Swiss populations of L. jacobaeae have
several life history traits that make them suitable candidates for biological control in cold continental climates.
Adult beetles are reproductively active throughout the
summer. They lay diapause eggs that persist through
the winter and are ready to hatch in early spring. The
larvae attack the rosette plants just as they are recovering from winter stress. Larvae of Swiss L. jacobaeae inhabit fresh leaves in early development and then move
into the root crowns during the second and third instars.
Damage caused by larval feeding affects plant growth
and may reduce reproductive output of their host plant.
Large numbers of larvae have been recorded from tansy
ragwort in open-field host tests and can cause mortality
at high densities (Hawkes, 1968; Puliafico, 2003).
The ragwort flea beetles collected in Switzerland
differed in several important life history traits from
those originally studied by Frick (1971). No signs of
over-wintering larvae or pupae were found during extensive field collections and observations during the
3 years of this study. Larvae of the Swiss populations
demonstrated distinct spatial partitioning of their host
plants as they develop, with first and early secondinstar larvae almost exclusively inhabiting the foliage
and above-ground portions of the plant and most thirdinstar larvae found only in the root crowns. Finally,
adult emergence was recorded at the end of June and
first 2 weeks of July in their native habitats in Switzerland. Almost all of the differences between our results
and those originally published by Frick (1971) can
be attributed to laboratory artifacts caused by raising
cold-adapted Swiss flea beetles under conditions better
suited for their Mediterranean counterparts from Italy.
We also found strong evidence of egg diapause broken
by extended cold treatment. Crossbreeding of Swiss
and Italian strains (Frick and Johnson, 1972) may have
contributed to some of the inconsistencies between our
egg eclosion data and those previously reported. Much
of the confusion caused by these laboratory results has
been repeated throughout the subsequent literature (e.g.,
Hawkes and Johnson, 1978; Windig, 1991, Coombs et
al., 1999). Clarification of Swiss L. jacobaeae life history traits will improve the ability to establish these
beetles for biological control, while increasing our understanding of their potential impact on tansy ragwort
infestations in colder continental climates.
Acknowledgements
The authors would like to acknowledge the technical assistance of A. de Meij, Y. Wang, E. Reneau, M.
Statsney, S.Teyssiere and A. Ordoniz. Thanks to K.
Marske, M. Julien, and P. Hatcher for comments on an
earlier draft. This study was funded by the Montana
Noxious Weed Trust Fund and the USDA Forest Service-Rocky Mountain Research Station.
References
Coombs, E.M., McEvoy, P.B. and Turner, C.E. (1999) Tansy
ragwort. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State University Press, Corvallis, OR, pp. 389400.
Frick, K.E. (1970) Longitarsus jacobaeae (Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy
ragwort. 1. Host plant specificity studies. Annals of the
Entomological Society of America 63, 284296.
Frick, K.E. (1971) Longitarsus jacobaeae (Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy
ragwort. II. Life history of a Swiss biotype. Annals of the
Entomological Society of America 64, 834840.
Frick, K.E. and Johnson, G.R. (1972) Longitarsus jacobaeae
(Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy ragwort. 3. Comparison of the biologies of the egg stage of Swiss and Italian biotypes. Annals
of the Entomological Society of America 65, 406410.
Frick, K.E. and Johnson, G.R. (1973) Longitarsus jacobaeae
(Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy ragwort. 4. Life history and adult aestivation of an Italian biotype. Annals of the Entomological
Society of America 66, 358366.
Hawkes, R.B. (1968) The cinnabar moth, Tyria jacobaeae,
for control of tansy ragwort. Journal of Economic Ento
mology 61, 499501.
Hawkes, R.B. and Johnson, G.R. (1978) Longitarsus jacoba
eae aids moth in the biological control of tansy ragwort.
In: Freeman, T.E. (ed.) Proceedings of the 4th Interna
tional Symposium on the Biological Control of Weeds.
University of Florida, Gainesville, FL, pp. 193196.
Markin, G.P. (2003) Biological control of tansy ragwort in
Montana: status of work as of December 2002 (unpub
lished report). USFS Rocky Mountain Research Station,
Bozeman, MT.
McEvoy, P.B. and Coombs, E.M. (1999) Biological control
of plant invaders: regional patterns, field experiments, and
structured population models. Ecological Applications 9,
387401.
Newton, H.C.F. (1933) On the biology of some species of
Longitarsus (Col., Chrysom.) living on ragwort. Bulletin
of Entomological Research 24, 511520.
Puliafico, K.P. (2003) Molecular taxonomy, bionomics and
host specificity of Longitarsus jacobaeae (Waterhouse)
(Coleoptera: Chrysomelidae): The Swiss population re
visited. Masters of Science, Montana State University,
Bozeman, MT.
Windig, J.J. (1991) Life cycle and abundance of Longitarsus
jacobaeae (Col.: Chrysomelidae), biocontrol agent of Se
necio jacobaea. Entomophaga 36, 605618.
Windig, J.J. and Vrieling, K. (1996) Biology and ecology
of Longitarsus jacobaeae and other Longitarsus species
feeding on Senecio jacobaea. In: Jolivet, P.H.A. and
Cox, M.L. (eds) Chrysomelidae Biology, vol. 3 General
Studies. Amsterdam, SPB Academic Publishing, pp.
315326.
205
Fungal survey for biocontrol agents of
Ipomoea carnea from Brazil
D.J. Soares and R.W. Barreto
Summary
Ipomoea carnea Jacq., also known as morning glory, is native of tropical America, and its purported
centre of origin is the Paraguay Basin. This plant is feared by ranchers because of its well-documented
toxicity to cattle. Because of its showy flowers, it became a popular ornamental in Brazil and was
introduced into others countries, becoming an aggressive wetland ecosystem invader. Little is known
about its mycobiota in Brazil which may include fungal pathogens that could be used in classical biocontrol programmes. Coleosporium ipomoeae (Schwein.) Burril and Puccinia puta H.S. Jacks. and
Holw. ex F. Kern, Thurst. and Whetzel are the only fungi recorded in the literature attacking this plant
in Brazil. An intensive search for specialized, coevolved fungal pathogens of I. carnea was initiated
in 2003 in Brazil. Twenty-one fungal species were collected. Among these were the two previously
known rusts, C. ipomoeae and P. puta, and Aecidium sp., Albugo sp., an unidentified ascomycete,
Mycosphaerella sp., five coelomycetes (Colletotrichum sp., Phoma sp. Phomopsis sp., and two Phyllosticta spp.) and ten hyphomycetes (Alternaria sp., Cercospora sp., Cladosporium sp., Curvularia
sp., Dactylaria-like, Fusarium-like, Nigrospora sp. Passalora sp. and two Pseudocercospora spp.).
Observations of the damage caused by such fungal diseases in the field indicate that the fungi with
the best potential as biological agents are C. ipomoeae, P. puta, Albugo sp., the Phyllostica sp. that
colonizes stems, and Phomopsis sp.
Keywords:aquatic weeds, biological control, coevolved pathogens, Ipomoea fistulosa,
Ipomoea carnea subsp. fistulosa.
Introduction
Morning glory, Ipomoea carnea Jacq., (local name in
Brazil is algodo-bravo) is a shrubby perennial amphibious plant belonging to the Convolvulaceae. It is
considered to be native to South America and particularly common in the basins of the rivers Paraguay and
So Francisco (Lorenzi, 2000). It is also widely distributed in Brazil as an ornamental species for its showy
violet flowers (Kissmann and Groth, 1995). This plant
is also one of the most feared poisonous weeds to Brazilian ranchers since it is able to cause severe nervous
disorder when ingested by bovines, sheep or goats (Tokarnia et al., 2000).
Ipomoea carnea was introduced into areas outside
the Neotropics, and it now causes serious invasions
of wetland habitats in Southern India and Pakistan
where streams, mangroves and other ecosystems may
Departamento de Fitopatologia, Universidade Federal de Viosa, Viosa,
MG, 36570-000, Brazil.
Corresponding author: D.J. Soares <[email protected]>.
CAB International 2008
1
be blocked, hampering irrigation and access (H.C. Evans personal communication, 2006). It is also included
in the Florida Exotic Pest Councils List of Floridas
Most Invasive Species as a weed category II (FLEPPC,
2003).
Surveys of fungal pathogens of plants native to Brazil that are weeds elsewhere have yielded a plethora of
potential biocontrol agents (Barreto and Evans, 1994,
1995a,b, 1998; Barreto and Torres, 1999; Barreto et al.,
1995, 1999a,b, 2000; Pereira and Barreto, 2000, 2005;
Monteiro et al., 2003; Soares and Barreto, 2006; Soares
et al., 2006), and two of the fungi highlighted as promising classical biocontrol agents during such surveys
have already been introduced from Brazil into other regions of the world: Prospodium tuberculatum (Speg.)
Arthur for the biological control of Lantana camara
L. (Ellison et al., 2006) in Australia and Colletotrichum gloeosporioides (Penz.) Penz. and Sacc. for the
biological control of Miconia calvescens DC in Hawaii
(Barreto et al., 2001).
Recently, a survey for fungal pathogens of I. carnea
was started, aimed at finding fungi to be used in the
future as biocontrol agents for this weed. Puccinia puta
H. S. Jacks. and Holw. ex F. Kern, Thurst. and Whetzel
206
Fungal survey for biocontrol agents of Ipomoea carnea from Brazil
was the only fungus previously recorded on I. carnea
(referred to as Ipomoea crassicaulis) in Brazil. Recently, a second rust fungus, Coleosporium ipomoeae (Schwein.) Burril, was recorded by Vieira et al., (2004).
Materials and methods
The collecting procedure adopted during the survey
was as described in Barreto (1991). The collecting trips
occurred between July 2004 and February 2006. Information on some ad hoc collections that were made before the main survey work is also included. The survey
covered a wide geographic area of central and southern
Brazil including the states of Minas Gerais, So Paulo,
Rio de Janeiro, Esprito Santo, Paran, Santa Catarina,
Rio Grande do Sul, Mato Grosso, Mato Grosso do Sul,
Gois and Rondnia.
The diseased parts of the plants suspected to be
damaged by fungal pathogens were collected, dried
in a plant press and taken to the lab. The isolation of
Table 1.
the potential agents was performed by direct transfer
of fungal structures to Petri dishes containing 15 ml
of VBA medium (Pereira et al., 2003), with the help
of a dissecting microscope and a sterilized fine point
needle. The fungi obtained were preserved in silica-gel
according to Dhingra and Sinclair (1996).
Selected specimens were deposited in the local
herbarium (Herbarium VIC). Fungal structures were
removed from diseased tissues and mounted in lactophenol. Observations of morphology were carried out
with an OLYMPUS BX 50 light microscope.
In order to confirm the pathogenicity of two selected
fungi (Passalora sp. and Alternaria sp.), isolates were
cultivated in VBA and incubated in the dark for 48 h
at 25C and later submitted to 12 h near-ultraviolet irradiation and 12 h dark. Four disks taken from 10-dayold cultures were placed abaxially and adaxially on
three leaves of two healthy potted I. carnea plants.
After inoculation, plants were left for 48 h in a humid
chamber prepared by covering the plants with plastic
Fungi recorded on Ipomoea carnea from Brazil by Soares (2007).
Fungus
Aecidium cf.
distinguendum
Albugo sp.
Alternaria alternata
Cercospora sp.
Cladosporium sp.
Coleosporium
ipomoeae
Colletotrichum sp.
Curvularia sp.
Dactylaria-like
Fusarium-like
Mycosphaerella
sp.
Nigrospora sp.
Passalora sp.
Phoma sp.
Phomopsis sp.
Phyllosticta sp. 1
Phyllosticta sp. 2
Pseudocercospora
sp. 1
Pseudocercospora
sp. 2
Puccinia puta
Unidentified
Ascomycete
Disease
Rust
Damage to host
Significant
Purported specificity
To the genus Ipomoea
Culturability
Not cultivable
Biocontrol potential
High
White rust
Leaf-spot
Significant
Significant
High
Non-specific
Not cultivable
Cultivable
High
Uncertain
Leaf-spot
Associated to
leaf-spots
Rust
Insignificant
Insignificant
Not investigated
Low
Cultivable
Cultivable
Low
Low
Significant
To the genus Ipomoea
Not cultivable
High
Anthracnose
(stems)
Associated to
leaf-spots
Associated to
leaf-spots
Associated to
leaf-spots
Leaf-spot
Moderate
Uncertain
Cultivable
Moderate
Insignificant
Low
Cultivable
None
Insignificant
Uncertain
Cultivable
None
Insignificant
Uncertain
Cultivable
None
Moderate
High
Cultivable
Moderate
Associated to
leaf-spots
Leaf-spot
Associated to
leaf-spots
Stem necrosis
Stem and
petiole blight
Associated to
leaf-spots
Leaf-spot
Insignificant
Low
Cultivable
None
Significant
Insignificant
High
Low
Cultivable
Cultivable
High
Low
Significant
Severe
Uncertain
High
High
Very high
Insignificant
Uncertain
Cultivable
Apparently
not cultivable
Cultivable
Moderate
High
Cultivable
Moderate
Leaf-spot
Moderate
High
Cultivable
Moderate
Rust
Stem canker
Significant
Significant
To the genus Ipomoea
Uncertain
Not cultivable
Attempts
unsuccessful
High
High
207
Low
XII International Symposium on Biological Control of Weeds
bags wetted inside and having water-soaked cotton internally and left at room temperature (approximately
25C). After that period, the plastic bags were removed, and plants were left on a bench under room
conditions and observed daily for the appearance of
symptoms. Three non-inoculated leaves of each of two
healthy plants, kept under the same conditions, served
as controls.
Results
Twenty-one fungal species were found in association
with I. carnea during the survey (Table 1). Among
these, at least two taxonomic novelties were promptly
recognized and will be dealt with separately in a taxonomic publication, namely: Passalora sp. and Phyllosticta sp.1. All the other fungi that were found represented new host or geographic records.
Inoculation of I. carnea with Passalora sp. yielded
symptoms equivalent to those observed in the field on
all inoculated leaves after 20 days. Non-inoculated
leaves remained healthy. Typical structures of the Passalora sp. were present on the diseased tissues, and the
fungus was re-isolated from newly infected tissues.
The species of Alternaria on I. carnea had the morphology and cultural characteristics typical of Alternaria
alternata (Fr.) Keissler. Its pathogenicity to I. carnea was
proven, and similar symptoms to those observed in the
field were observed within 15 days of inoculation. This
fungus has not been recorded on I. carnea until now.
Attempts to isolate the Phyllostica sp.1 associated
with stem and petiole lesions were unsuccessful. This
fungus appears to have a biotrophic habit. Plant tissues
surrounding the fungus colonies were observed to retain a healthy appearance until late stages of infection.
Necrosis, leaf drop and death of the apical buds only
occurred at the final stages of infection.
Discussion
At the present stage of this research, it would be too
early to dismiss any of the fungi as not promising for
use as biocontrol agents for I. carnea. Some of the fungi
collected in association with I. carnea are either evident
saprophytes or suspected to have such status, as is the
case of Nigrospora sp., Cladosporium sp., Curvularia
sp., the Dactylaria-like fungus, the Fusarium-like fungus
and Phyllosticta sp.2. Otherwise, the damage associated
with the other fungi, listed in Table 1, was significant
as observed in the field. In general, plants infected with
such fungi appeared weaker and defoliated as compared
with individuals in healthy I. carnea populations. The
fungi appearing to be the most promising candidates for
use in weed biocontrol, deserving further evaluations
are: the rusts C. ipomoeae and P. puta, Albugo sp., Passalora sp., Phyllostica sp.1 and Phomopsis sp.
Both rust fungi were frequently found throughout the
year associated with moderately high plant defoliation.
However, they appear to have a wide host range within
the Convolvulaceae since both have been recorded on
other species in this family, including sweet potato (in
the case of C. ipomoeae; Hennen et al., 2005). There
may be host-specific strains of C. ipomoeae and P. puta
that could safely be introduced into other regions of the
globe, but even if these species are proven to be polyphagous within the Convolvulaceae, their introduction into other areas of Brazil against noxious I. carnea
population might still be considered.
Albugo sp. was found only a few times, in the states
of Mato Grosso, Mato Grosso do Sul and So Paulo.
This fungus appears to have a more restricted geographic distribution compared with the two rusts. It
caused a complete leaf curling or leaf blight (when the
attack occurred on the petioles). However, its specificity and potential to be used as a biocontrol agent requires further investigation.
Passalora sp. could prove useful as a classical biological control or even as a mycoherbicide against I.
carnea. Although no sporulation was obtained for this
fungus in the conditions that were used, the potential
for mass production of spores, which is critical for its
viability as a mycoherbicide, was not properly investigated.
Phomopsis sp. was consistently found associated
with stem necrosis and easily sporuled in culture; however, its pathogenicty and specificity has not yet been
tested.
Phyllosticta sp.1 appears to be the most promising
candidate to be used as a classical biological control
agent. The damage inflicted naturally by this fungus
on I. carnea populations was evident. Infected plants
in advanced disease stages were weakened and almost
completely defoliated. On diseased plants, foliage on
each individual stem was often reduced to only six or
eight terminal leaves.
Although it used to be thought that A. alternata had
several pathotypes that produce host-specific toxins,
this was considered wrong by Simmons (1999). If further investigation on this fungus on I. carnea confirms
that it fits within the non-specific, cosmopolitan A.
alternata-group, this would restrict its potential as a
classical biocontrol agent but not necessarily result in
its rejection for use as a biocontrol agent of I. carnea.
This fungus grows well and sporulates abundantly in
culture and could be further evaluated for development
of a mycoherbicide to be used in Brazil similarly to
what is being done with an isolate of A. alternata obtained from Eichhornia crassipes in India (Babu et al.,
2002, 2003, 2004).
Half of the fungi previously recorded in the literature
in association with I. carnea were recorded only from
countries outside the native range of this plant species
in the Neotropics. Most of the records from countries
such as India, Pakistan and Malaysia probably represent
saprophytic, weakly pathogenicopportunistic or genera
listic pathogens of no relevance for biocontrol (Table 2).
208
Fungal survey for biocontrol agents of Ipomoea carnea from Brazil
Table 2.
Fungi recorded on Ipomoea carnea and their synonyms worldwide. Extracted from Farr et al. (no date).
Fungus name
Aecidium agnesiae (Syd.) Z. Urb.
Aecidium distinguendum P. Syd. and Syd.
Aecidium sp.
Albugo ipomoeae (as spelt by the author)
Albugo ipomoeae-panduratae (Schwein.) Swingle
Aplosporella ipomoeae S. Ahmad
Botryodiplodia theobromae Pat.
Capnodium sp.
Cercospora ipomoeae G. Winter
Coleosporium ipomoeae
Cytospora ipomoeae S. Ahmad and Arshad
Dischloridium cylindrospermum S.K. Srivast.
Dothiorella ipomoeae S. Ahmad
Guignardia cytisi (Fuckel) Arx and E. Mll.
Lasiodiplodia theobromae (Pat.) Griffon and Maubl.
Leptosphaeria macrospora (Fuckel) Thm.
Macrophoma ipomoeae Pass.;
Marasmiellus scandens (Massee) Dennis and D.A. Reid
Meliola malacotricha Speg.
Monilochaetes infuscans Harter
Munkovalsaria donacina (Niessl) Aptroot
Ophiobolus herpotrichus (Fr.) Sacc.
Periconia byssoides Pers.
Pestalotiopsis adusta (Ellis and Everh.) Steyaert
Phoma herbarum var. herbarum Westend.
Phomopsis ipomoeae Petr.
Pseudocercosporella ipomoeae Sawada ex Deighton
Puccinia achyroclines (Henn.) H.S. Jacks. and Holw.a
Puccinia distinguenda H.S. Jacks. and Holw.
Puccinia megalospora (Orton) Arthur and J.R. Johnst.
Puccinia nocticolor Holw.
Puccinia puta
Puccinia rubicunda Holw.
Tuberculina persicina (Ditmar) Sacc.
a
Country/Region
Cuba
Caribbean; Cuba; Venezuela
Venezuela
Cuba
Caribbean; Cuba
India; Pakistan
Pakistan
Caribbean; Cuba
India
Cuba; Colombia; Brazil
India
India
India
Pakistan
Venezuela
Pakistan
India; Pakistan
Malaysia
Malaysia
India
India
Pakistan
Venezuela
India
Pakistan
Venezuela
Venezuela
Brazil
Ecuador; Venezuela
Mexico
Guatemala
Colombia; Venezuela; Puerto Rico; Brazil
Mexico
Caribbean
his record is regarded here as dubious, since the original publication (Hennen et al., 1982) which was cited by Farr et al. (no date) makes no
T
mention of I. carnea or its synonyms as host for this fungus.
Although only a relatively limited area of the native
range of I. carnea was surveyed, several potential biocontrol agents were found. It is, therefore, expected that
the expansion of the survey into new areas in Brazil or
other parts of the Neotropics will reveal a much larger
list of potential fungal agents for biocontrol of I. carnea. Although Brazil is considered as part of the centre
of origin of this plant, only two fungi were previously
known on this host in Brazil. Results of the present
study added 19 new taxa to this list, most of which are
clearly pathogenic to I. carnea (Table 1). Pathogenicity and host-specific tests are now being conducted to
confirm the status of fungi selected as having possible
potential for use in the biocontrol of I. carnea.
Acknowledgements
This work forms part of a research project submitted
as a PhD dissertation to the Departamento de Fitopato-
logia/Universidade Federal de Viosa by D.J. Soares.
The authors thank the Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq) and the
Fundao de Amparo Pesquisa do Estado de Minas
Gerais (FAPEMIG) for financial support.
References
Babu, R.M., Sajeena, A., Seetharaman K., Vidhyasekaran, P.,
Rangasamy, P., Prakash, M.S., Raja, A.S. and Biji, K.R.
(2002) Host range of Alternaria alternata - a potential
fungal biocontrol agent for waterhyacinth in India. Crop
Protection 21, 10831085.
Babu, R.M., Sajeena, A. and Seetharaman K. (2003) Bioassay
of the potentiality of Alternaria alternata (Fr.) Keissler as
a bioherbicide to control waterhyacinth and other aquatic
weeds. Crop Protection 22, 10051013.
Babu, R.M., Sajeena, A. and Seetharaman K. (2004) Solid
substrate for production of Alternaria alternata conidia:
209
XII International Symposium on Biological Control of Weeds
a potential mycoherbicide for the control of Eichhornia
crassipes (waterhyacinth). Weed Research 44, 298304.
Barreto, R.W. (1991) Studies on the pathogenic mycoflora of
selected weeds from the State of Rio de Janeiro (Brazil),
PhD thesis, University of Reading, England, UK.
Barreto, R.W. and Evans, H.C. (1994) The mycobiota of
the weed Chromolaena odorata in southern Brazil with
particular reference to fungal pathogens for biological
control. Mycological Research 98, 11071116.
Barreto, R.W. and Evans, H.C. (1995a) Mycobiota of the
weed Cyperus rotundus in the state of Rio de Janeiro,
with an elucidation of its associated Puccinia complex.
Mycological Research 99, 407419.
Barreto, R.W. and Evans, H.C. (1995b) The mycobiota of
the weed Mikania micrantha in southern Brazil with
particular reference to fungal pathogens for biological
control. Mycological Research 99, 343352.
Barreto, R.W. and Evans, H.C. (1998) Fungal pathogens of
Euphorbia heterophylla and E. hirta in Brazil and their
potential as weed biocontrol agents. Mycopathologia 141,
2136.
Barreto, R.W. and Torres, A.N.L. (1999) Nimbya alternantherae and Cercospora alternantherae: two new records
of fungal pathogens on Alternanthera philoxeroides (alligatorweed) in Brazil. Australasian Plant Pathology 28,
103107.
Barreto, R.W., Evans, H.C. and Ellison, C.A. (1995) The
mycobiota of the weed Lantana camara in Brazil, with
particular reference to biological control. Mycological
Research 99, 769782.
Barreto, R.W., Evans, H.C. and Hanada, R.E. (1999a) First
record of Cercospora pistiae causing leaf spot of water
lettuce (Pistia stratiotes) in Brazil, with particular reference to weed biocontrol. Mycopathologia 144, 8185.
Barreto, R.W., Evans, H.C. and Pomella, A.W.V. (1999b)
Fungal pathogens of Calotropis procera (rubber bush),
with two new records from Brazil. Australasian Plant Pathology 28, 126130.
Barreto, R.W., Charudattan, R., Pomella, A. and Hanada, R.
(2000) Biological control of neotropical aquatic weeds
with fungi. Crop Protection 19, 697703.
Barreto, R.W., Seixas, C.D.S. and Killgore, E. (2001) Colletotrichum gloeosporioides f.sp. miconiae: o primeiro fungo fitopatognico brasileiro a ser introduzido no exterior
para o controle biolgico clssico de uma planta invasoras
(Miconia calvescens) In: 7th Simpsio de Controle Biolgico. 0307 June 2001, Poos de Caldas.
Dhingra, O.D. and Sinclair, J.B. (1996) Basic Plant Pathology Methods, 2nd ed. Lewis Publishers, Boca Raton, FL.
Ellison, C.A., Pereira, J.M., Thomas, S.E., Barreto, R.W. and
Evans, H.C. (2006) Studies on the rust Prospodium tuberculatum, a new classical biological control agent released
against the invasive weed Lantana camara in Australia. 1.
Life-cycle and infection parameters. Australasian Plant
Pathology 35, 309319.
Farr, D.F., Rossman, A.Y., Palm, M.E., and McCray, E.B.
(no date) Fungal Databases, Systematic Botany & Mycology Laboratory, ARS, USDA. Available at: http://nt.
ars-grin.gov/fungaldatabases/, accessed March 21, 2007.
FLEPPC (2003) List of Floridas Invasive Species. Florida
Exotic Pest Plant Council. Available at: http://www.fleppc.
org/03list.htm.
Hennen, J.F., Figueiredo, M.B., Caralho Jr, A.A. and Hennen,
P.G. (2005) Catalogue of the species of plant rust fungi
(Uredinales) of Brazil. Available at: http://www.jbrj.gov.
br/publica/uredinales/Brazil_Catalogue1drevisado.pdf.
Hennen, J.F., Hennen, M.M. and Figueiredo, M.B. (1982)
ndice das ferrugens (Uredinales) do Brasil. Arquivos do
Instituto Biolgico de So Paulo 49(Suppl. 1), 1201
Kissmann, K.G. and Groth, D. (1995) Plantas Infestantes e
Nocivas. Tomo II. BASF S.A., So Paulo, Brasil.
Lorenzi, R. (2000) Plantas daninhas do Brasil: terrestres,
aquticas, parasitas e txicas. Instituto Plantarum de Estudos da Flora Ltda. Nova Odessa, Brasil.
Monteiro, F.T., Vieira, B.S. and Barreto, R.W. (2003) Curvularia lunata and Phyllachora sp.: two fungal pathogens of
the grassy weed Hymenachne amplexicaulis from Brazil.
Australasian Plant Pathology 32, 449453.
Pereira, J.M. and Barreto, R.W. (2000) Additions to the
mycobiota of the weed Lantana camara (Verbenaceae) in
southeastern Brazil. Mycopathologia 151, 7180.
Pereira, J.M., Barreto, R.W., Ellison, A.C. and Mafia, L.A.
(2003) Corynespora casiicola f. sp. lantanae: a potential
biocontrol agent from Brazil for Lantana camara. Biological Control 26, 2131.
Pereira, O.L. and Barreto, R.W. (2005) The mycobiota of the
weed Mitracarpus hirtus in Minas Gerais (Brazil) with
particular reference to fungal pathogens for biological
control. Australasian Plant Pathology 34, 4150.
Simmons, E.G. (1999) Alternaria themes and variations
(236243) Host-specific toxin producers. Mycotaxon 70,
325369.
Soares, D.J. (2007) Fungos associados once plantas aquticas no Brasil e seu potencial para controle biolgico. PhD
thesis. Universidade Federal de Viosa, MG/Brasil.
Soares, D.J. and Barreto, R.W. (2006) Additions to the Brazilian mycobiota of the grassy weed, Hymenachne amplexicaulis, with a discussion on the taxonomic status of
Paraphaeosphaeria recurvifoliae. Australasian Plant Pathology 35, 347353.
Soares, D.J. Ferreira, F.A. and Barreto, R.W. (2006) First report of the aecial stage of Puccinia scirpi on Nymphoides
indica in Brazil, with comments on its worldwide distribution. Australasian Plant Pathology 35, 8184.
Tokarnia, C.H., Dbereiner, J. and Peixoto, P.V. (2000) Plantas txicas do Brasil. Editora Helianthus, Rio de Janeiro,
Brazil.
Vieira, F.M.C., Pereira, O.L. and Barreto, R.W. (2004) First
report of Coleosporium ipomoeae on Ipomoea fistulosa in
Brazil. Fitopatologia Brasileira 29, 693.
210
Biological control of lippia
(Phyla canescens): surveys for the plant
and its natural enemies in Argentina
A.J. Sosa,1 M.G. Traversa,2 R. Delhey,2 M. Kiehr,2
M.V. Cardo1 and M.H. Julien3
Summary
Lippia, Phyla canescens (Kunth) Greene (Verbenaceae) is a fast-growing, mat-forming plant native
to South America. It is a weed in Australia, where it was introduced as an ornamental during the
nineteenth century. The knowledge about the biology of lippia is currently limited to unconcluded
taxonomical studies; there is scarce information about the ecology and natural enemies in the native
range. Surveys for the plant and its natural enemies were initiated in Argentina in 2005 to determine
its distribution and to search for possible biological control agents, both insects and phytopathogens.
We have found Phyla sp. in 54 out of 102 sites sampled, mostly east of 66W, circumscribing the
weed to the Chaco Domain. In places where the plant was present, at least 20 arthropods and 16 fungi
were found. Among insects, the most promising candidates are three flea beetles (Chrysomelidae):
two species of Longitarsus and Kuschelina bergi Harold. Pathogens include the rust Puccinia cf.
lantanae Farl., Cercospora cf. lippiae Ellis and Everh. and three Colletotrichum spp., associated with
leaf spots and stem cankers. Additional information on their biology and host specificity is required
to propose any of these as biological control candidates.
Keywords: plant distribution, arthropods, pathogenic fungi.
Introduction
Lippia, Phyla canescens (Kunth) Greene (Verbenaceae)
is a fast-growing, mat-forming plant. It is widespread
in and thought to be native to South America (from
southern Ecuador, throughout Peru, Chile, Argentina,
Uruguay, Paraguay, Bolivia and Brazil) (Collantes,
et al., 1998; Mlgura de Romero et al., 2003). It was
also recorded from fossil pollen in Santa Fe and Buenos
Aires Provinces, in Argentina (Alzugaray et al., 2003;
Fontana, 2005), reinforcing this area as a centre of
origin.
P. canescens has been reported naturalized from
France, Spain, Italy, Algeria, Botswana, Senegal, Egypt,
USDA-ARS South American Biological Control Laboratory. Bolivar
1559 (B1686EFA), Hurlingham, Buenos Aires, Argentina.
2
Laboratorio de Patologa Vegetal, Departamento de Agronoma,
Universidad Nacional del Sur. (8000), Baha Blanca, Buenos Aires,
Argentina.
3
CSIRO Entomology European Laboratory. Campus International de
Baillarguet. 34980 Montferrier sur Lez, France.
Corresponding author: A.J. Sosa <[email protected]>.
CAB International 2008
1
South Africa, New Zealand and Australia (Kennedy,
1992). This plant is a weed in Australia, where it was
introduced as an ornamental plant during the second
half of the nineteenth century. It is a major threat to
biodiversity and riparian areas and has a significant
impact on conservation and grazing systems due to its
increasing density and distribution (Julien et al., 2004),
causing annual losses of cattle production of 38 million
Australian dollars and even greater estimated losses of
environmental service. Current short-term and unsustainable control methods include the use of herbicides,
cultivation and grazing management (Earl, 2003; Julien et al, 2004). Biological control is proposed as the
sustainable method for this weed and may be the only
option for conservation areas, woodlands, forests and
along stream banks. However, until this study, there
was little information about the plant and its natural
enemies in the native range.
Worldwide, Kennedy (1992) recognized nine species of Phyla including P. canescens. In Argentina, an
additional species, Phyla reptans (Kunth) Greene was
recognized together with Phyla betulaefolia Greene
and P. canescens (Mlgura de Romero et al., 2003).
211
XII International Symposium on Biological Control of Weeds
Australia has P. canescens and Phyla nodiflora Greene
(Munir, 1993); however, there is confusion between
these two species in the literature, and it remains uncertain whether P. nodiflora is native to Australia or not
(Leigh and Walton, 2004).
There are no records of arthropods associated with
Phyla spp. and few for fungal pathogens. In North America, two biotrophic fungi: Oidium sp. (powdery mildew) and Meliola lippiae Maubl. (black mildew) and
the following necrotrophs: Cercospora lippiae Ellis
and Everh. (leaf spot), Sphaceloma lippiae Baines and
Cummins (anthracnose) and Sclerotium rolfsii Sacc.
(southern blight) were identified on P. nodiflora and/or
Phyla lanceolata (Michaux) Greene (Farr et al., 1989).
C. lippiae (syn. Pseudocercospora lippiae) occurred on
Phyla strigulosa (Mart. and Gal.) Moldenke and Phyla
spp. in the Caribbean, Asia and Africa (Ellis, 1976).
On P. nodiflora (=Lippia nodiflora, including also
P. reptans?) in India, C. apii Fresen. emend. Crous and
U. Braun was recorded (Crous and Braun, 2003) and
in South America, Meliola lantanae Syd. and P. Syd.
and Phoma zappaniae Speg. (Vigas, 1961). In Argentina, the rust Puccinia lantanae Farl. was found on
P. canescens in Salta Province, on Phyla sp. in Tucumn Province and in Paraguay (Vigas, 1961; Lindquist, 1982). There are also records of Prospodium
spp. on Lippia (including Phyla) in South America
(Vigas, 1961).
To gain knowledge of the natural distribution, centre
of origin and natural enemies of P. canescens in South
America, surveys were initiated in December 2005.
The aim was to identify suitable arthropod and phytopathogen biological control agents for Australia.
Methods and materials
Surveys
P. canescens is recorded in 17 provinces of Argentina (Mlgura de Romero et al., 2003). Considering
these records and bibliographical information, surveys were conducted in four ecological regions of
Argentina from December 2005 to February 2007
(Figure 1): (1) Wetland Chaco, a humid area with a
rainy season in summer, characterized as wetlands
with patches of mesquite forests, which includes the
north-eastern provinces of Entre Ros, Santa Fe, Corrientes, Chaco and Formosa; (2) Dry Chaco (similar
to the wet Chaco but mostly grassland) and Yungas
(subtropical mountain rain forest), including the northwestern provinces of Jujuy, Salta, Santiago del Estero,
Tucumn, Catamarca, La Rioja, Crdoba, Mendoza
and San Luis; (3) Pampas, grassland of Buenos Aires
Province; and (4) Transition zone between Southern
Chaco, Pampas and Patagonia, a cold and dry area
including Ro Negro, Neuqun and Mendoza but
wet in the Ro Negro basin where P. canescens was
found.
P. canescens is a small, prostrate plant, often
growing as an understory plant and therefore difficult to see. Sampling was conducted along roadsides
every 100180 km. At each site, the presence of the
plant was checked, and if present, natural enemies
were collected. In 27 sites where plants were difficult to identify in the field (because of absence of
inflorescences), specimens were collected and cultivated in the greenhouse for identification. Repeated
sampling was conducted at two (region 4) and ten
sites (regions 13) in each region at different times
of the year. Soil was sampled at 14 sites with and
without lippia, and pH and relative humidity were
recorded. Samples were analysed for phosphates, nitrates, nitrites, ammonia, calcium, phosphorous, iron
and humus (LaMotte Combination Soil Model STH14). This information was examined with Reciprocal
Averaging (PC-ord 4).
Natural enemies
Arthropods: Arthropods were collected directly from
plants using aspirators or from plant parts attacked by
endophagous species (e.g. miners or gall formers) and
taken to the laboratory for rearing. Material was placed
in plastic containers (8 cm diameter, 5 cm high) with
moistened tissue paper and leaves of the plant as food,
and kept in growth chambers at 25C and 12 h light.
All adults that were collected or reared were sent to
taxonomists for identification. Relevant biological information was recorded.
Laboratory studies were conducted for the flea
beetle collected in two places: Tres Arroyos (38.52S,
60.51W) and Nueva Atlantis (36.85S, 56.69W) in
Buenos Aires Province. Eggs laid by field-collected
adults were kept in plastic containers (10 cm diameter,
2 cm high), and P. canescens leaves were added to the
resulting larvae. The last instars were reared in a bigger
container, and when the larvae decreased their activity (prepupal stage), they were transferred to another
container (8 cm diameter, 5 cm high) with soil as substrate.
Pathogens: Plants were inspected for symptoms of
disease, representative samples were collected, and
the presence of fungi was checked in the lab. When
necessary, samples were placed into humid chambers
to encourage sporulation. Isolations were made by
placing disinfected leaf and stem pieces onto different
agar culture media: potato dextrose agar (PDA), water agar (WA) and specific media (Phyla leaf-oat meal
agara modification of carrot leaf-oat meal agar;
Dhingra and Sinclair, 1985). The mycelial cultures
were exposed to an UV regime for sporulation. Fungi
were then identified. The isolated fungi are maintained on PDA. Desiccated specimens of infected
plants have been incorporated in the herbarium of the
Phytopathology Laboratory of the University of Baha
Blanca.
212
Biological control of lippia (Phyla canescens): surveys for the plant and its natural enemies in Argentina
Results
Surveys
Phyla sp. was recorded in 54 of 102 sites sampled,
mostly east of 66W and north of 40S, from sea level
(Buenos Aires Province) to 2100 m (Volcn, Jujuy
Province) suggesting a natural distribution in the Chaco
Region and in the Pampas (Chaco Domain) (Cabrera
and Willink, 1980; Figure 1). In the northern half of
region 2, pure and mixed stands of P. canescens and
20S
70W
P. reptans and some possibly intermediate forms were
found. High phenotypic variation and poor distinguishing characters (presence or absence of conspicuous secondary leaf venation) made it difficult to discriminate
these species in the field. Specimens that were taken
for identification turned out to be both P. reptans and
P. canescens. Elsewhere, populations found were easily identified as P. canescens.
Preliminary analyses of soil samples did not detect
any obvious differences between sites with or without
Phyla spp.
65W
60W
55W
20S
Bolivia
Paraguay
25S
Chile
25S
30S
Argentina
Brazil
30S
2
Uruguay
35S
35S
3
40S
4
70W
Figure 1.
40S
0
65W
60W
200
km
400
55W
Sites sampled for Phyla spp. and natural enemies associated. Black dots indicate plant presence and white dots, plant absence. 1 Wetland Chaco, 2 Dry Chaco and Yungas, 3 Pampas
and 4 Transition between Southern Chaco, Pampas and Patagonia.
213
XII International Symposium on Biological Control of Weeds
Natural enemies
Arthropods: So far, about 20 species of arthropods
have been found: four flea beetles [Kuschelina bergi
Harold, Longitarsus spp. and Disonycha glabrata (Fabricius) (Chrysomelidae)], a leaf mining fly (Agromyzidae?), two thrips (Thysanoptera), four species of
Lepidoptera (two micro moths and two hairy caterpillars), eriophyid mites, unidentified stem gallers, four
leafhoppers (Cicadellidae) and two Cercopidae. The
latter six sap feeders are generalists. Leafhoppers were
found throughout the range of both plant species (P.
canescens and P. reptans). Different beetle species had
different ranges, sometimes with overlapping distribution; however, Longitarsus spp. were found throughout
the ranges of both plants. In the field, Longitarsus spp.
were found only as adults feeding on leaves; attempts
to lab rear have not yet been successful.
The flea-beetle collected and studied in the laboratory, K. bergi, was only found on P. canescens in
the Pampas and in the transition zone (regions 3 and
4, Figure 1). It was only found on litter or the ground
amongst prostrate stems. In the lab, it has five larval
instars and takes about 2months to complete its life
cycle. Pupation occurs in the substrate and takes about
2 weeks. Further studies including host range tests are
planned. There is no biological information available
on the other natural enemies.
Pathogens: At this early stage of the research, at
least 16 species of fungi have been found associated
with the two Phyla spp. Several of them are secondary invaders (e.g. Nigrospora, Sordaria, Podospora),
others are clearly pathogenic (Puccinia cf. lantanae,
C. lippiae, Colletotrichum spp.), and others might
or might not be involved in the etiology of diseases
(Fusarium sp., Bipolaris sp., Alternaria sp., Phoma
sp., Phomopsis sp.).
P. cf. lantanae was found on Phyla cf. reptans in one
place in the north-eastern province of Jujuy. Only telia were present on leaves and stems. Teliospores were
mainly one-celled, but there were also some two-celled
spores. C. lippiae was isolated from P. canescens associated with stromata on necrotic leaves and circular
leaf spots. There are at least three Colletotrichum spp.
involved in leaf-spot symptoms: Colletotrichum dematium (Pers.: Fr.) Grove on P. reptans and Colletotrichum cf. orbiculare on P. canescens, while a third species, tentatively placed in Colletotrichum, was found
on P. cf. reptans.
Discussion
The relatively short period of surveys so far has provided a range of arthropod and phytopathogen species
that are being considered as potential agents. The fleabeetle K. bergi, a leaf-feeder, is the first insect to be
successfully reared in the lab and will soon undergo
preliminary host testing. Other beetles, thrips and Lep-
idoptera await further observations and studies. The
microcyclic rust (P. cf. lantanae) is also under consideration, though preliminary studies suggest its host is
Phyla reptans, the species we collected it from, rather
than P. canescens, the host recorded by Spegazzini
(1909). C. lippiae is the most widespread phytopathogen, causing necrotic leaf spots. Three Colletotrichum
spp. have been isolated and are of interest. We have not
yet identified any arthropodpathogen associations.
The native range of P. canescens is thought to be in
South America (Kennedy, 1992), although at present the
centre of diversity is not known. It is widely dispersed in
the Pampas and dry Chaco in Argentina, and this is where
surveys have been concentrated so far. Only two of the
three Phyla species recorded from Argentina have been
found. Phyla reptans, and possible intermediate forms
with P. canescens, occur in the wetter north-western
areas of Argentina, whereas P. canescens appears to be
the only Phyla species that occurs in the drier Chaco
and the vast grassland areas. Recent molecular studies indicate that P. canescens might be restricted to the
south and that the species in the north is P. reptans (M.
Fatemi, unpublished data, 2007). Therefore, we tentatively hypothesize that the centre of dispersion of the
genus in Argentina is the Chaco Domain.
Future surveys will seek Phyla species and their
natural enemies in countries neighbouring Argentina.
In addition, taxonomic (including morphometrics and
molecular analyses) and cytogenetic studies of Phyla
and repeated collections of natural enemies will help
us determine the centres of origin, delimit the native
ranges and select potential biological control agents for
P. canescens.
Acknowledgements
We thank M. Mlgura de Romero (Instituto de Botnica Darwinion, San Isidro, Buenos Aires) and N.
Cabrera (Museo de La Plata) for plant and flea beetle
identifications, respectively. We appreciate the field
assistance of Marcelo Valverde and lvaro Alsogaray
from El Rey and Copo National Parks. We also thank
H. Hinz and H. Evans (Scientific Committee) for their
comments and suggestions that improved the original
manuscript. This study is supported by the New South
Wales Department of Natural Resources Wetlands Recovery Project.
References
Alzugaray, C., Feldman, S. R. and Lewis, J. P. (2003) Dinmica
del banco de semillas de un espartillar de Spartina argentinensis. Ciencia e Investigacin Agraria 30, 197209.
Cabrera, A. L. and Willink A. (1980) Biogeografa de Amrica
Latina. Monografa 13, Serie Biologa. OEA,Washington
DC, USA.
Collantes, M. B., Stofella, S. L., Ginzo, H. D. and Kade, M.
(1998) Productividad y composicin botnica divergente
de dos variantes florsticas de un pastizal natural de la
214
Biological control of lippia (Phyla canescens): surveys for the plant and its natural enemies in Argentina
Pampa Deprimida fertilizadas con N y P. Revista de la
Facultad de Agronoma, La Plata 103, 4559.
Crous P. W. and Braun, U. (2003) Mycosphaerella and its anamorphs: names published in Cercospora and Passalora.
CBS Biodiversity Series No.1, 571 pp.
Dhingra, O. D. and Sinclair, J. B. (1985) Basic Plant Pathology Methods. CRC Press, Boca Raton, FL, 355 pp.
Earl, J. (2003) The distribution and impacts of Lippia (Phyla
canescens) in the Murray Darling System. Agricultural
Information and Monitoring Services. ABN: 73 918 506
894.
Ellis, M. B. (1976) More Dematiaceous Hyphomycetes.
Commonwealth Mycological Institute, Kew, Surrey, UK,
507 pp.
Farr, D. F., Bills, G. F., Chamuris, G. P. and Rossman, A. Y.
(1989) Fungi on Plants and Plant Products in the United
States. APS Press, St. Paul, MN, USA, VIII+, 1252 pp.
Fontana, S. L. (2005) Coastal dune vegetation and pollen representation in south Buenos Aires Province, Argentina.
Journal of Biogeography 32, 719735.
Julien, M. H., Storrie, A., and McCosker, R. (2004) Lippia,
Phyla canescens, an increasing threat to agriculture and
the environment. 476479. In: B. M. Sindel and S. B.
Johnson (eds) 14th Australian Weeds Conference. Charles
Sturt University, Wagga Wagga, Australia, 718 pp.
Kennedy, K. (1992) A systematic study of the genus Phyla
Lour. (Verbenaceae: Verbenoideae, Lantanae). Doctoral
thesis. The University of Texas at Austin, TX, USA.
Leigh, C. and Walton, C.S. (2004) Lippia (Phyla canescens)
in Queensland. Deparment of Natural Resources, Mines
and Energy, Brisbane, Queensland, Australia, 34 pp.
Lindquist, J.C. (1982) Royas de la Repblica Argentina y
zonas limtrofes. INTA, Buenos Aires, Argentina, 574 pp.
Mlgura de Romero, M.E., Rotman, A. D. and Atkins, S.
(2003) Verbenaceae, tribu Lantaneae, In: A. M. Anton
and F. O. Zuloaga (eds) Flora Fanerogmica Argentina
84, 146.
Munir, A. A. (1993) A taxonomic revision of the genus Phyla
Lour. (Verbenaceae) in Australia. Journal of the Adelaide
Botanical Gardens 15, 109128.
Spegazzini, C. (1909) Mycetes argentinenses. Anales del Museo Nacional de Buenos Aires, serie III 12, 257458.
Vigas, A. P. (1961) ndice de fungos da Amrica do Sul. Instituto Agronomico, Campinas, Brazil, 921 pp.
215
Potential biological control agents of
field bindweed, common teasel and
field dodder from Slovakia
P. Tth,1 M. Tthova2 and L. Cag1
Summary
Field explorations during 2001 to 2006 in Slovakia resulted in the discovery of several potential biological control agents of the three weeds, field bindweed, Convolvulus arvensis L.; common teasel,
Dipsacus fullonum L.; and field dodder, Cuscuta campestris Yuncker. The five top candidates are
described in the following. The larvae of the agromyzid fly Melanagromyza albocilia Hendel (Agromyzidae) mine in the stems and root crowns of field bindweed, causing the death of infested shoots.
The number of infested plants ranged from 46.7% to 99.2% and the number of infested stems from
4.1% to 37.2% in southwest Slovakia. The larvae and adults of the tortoise beetle, Hypocassida subferruginea (Schrank) (Chrysomelidae), almost completely destroyed leaves of field bindweed in some
uncultivated habitats in the warmest localities of Slovakia. Development of the species is rapid under
favourable conditions and takes only 22 to 27 days; females have a high fecundity, and it is easy to
rear. The most important natural enemy of H. subferruginea recorded in Slovakia was the egg parasitoid Brachista pungens (Mayr) (Trichogrammatidae). Adult moths of Endothenia gentianaeana (Hbner) and Cochylis roseana (Haworth) (Tortricidae) were reared in high numbers from flowerheads
of common teasel during the study. Of E. gentianaeana, only one larva was found per flowerhead,
feeding within the central cavity, while larvae of C. roseana were gregarious. Especially C. roseana
was destroying a large number of seeds within the flowerheads of teasel. Considerable parasitization
of E. gentianaeana by Glypta mensurator (Fabricius) (Ichneumonidae) was noted. Weevils from the
genus Smicronyx (Curculionidae) were found to be the principal natural enemies of dodders in Slovakia. Larvae of Smicronyx spp. induce stem galls, which prevents flowering and fruiting of field dodder
vines. Smicronyx jungermanniae (Reich) was the most abundant species, accounting for up to 96% of
the total number of weevils (n = 877) reared from field dodder galls.
Keywords: Convolvulus, Dipsacus, Cuscuta, insect, candidates.
Introduction
Field bindweed, Convolvulus arvensis L. (Convolvulaceae), has been described as the 12th worst weed in
the world, the seventh most important in Europe and
the most important weed in European orchards. Field
bindweed tolerates a great range of environmental
conditions and elevations; for more information on
bindweed, see Tth (2000). Although a relatively large
number of species has been recorded from the Convol-
Slovak Agricultural University, Department of Plant Protection,
A. Hlinku 2, 949 76 Nitra, Slovak Republic.
2
Slovak Agricultural University, Department of Sustainable Development, Marinska 10, 949 01 Nitra, Slovak Republic.
Corresponding author: P. Tth <[email protected]>.
CAB International 2008
1
vulaceae, only a small number of them seem to have
potential as biological control agents (Tth and Cag,
2005). Of these, only Tyta luctuosa (Denis and Schiffermuller) (Lepidoptera: Noctuidae) and Aceria malherbae Nuzzaci (Acari: Eriophyidae) have so far been
used in the classical biological control of bindweeds in
North America (Rees and Rosenthal, 1996).
Common teasel, Dipsacus fullonum L. (Dipsacaceae),
grows mostly in non crop areas in Slovakia. River
banks, roadsides and disturbed areas are the most common habitats of teasel throughout Slovakia. Teasel is an
invasive species in North America. Whereas the plant
is not a problem on Slovakian agricultural land, it is
considered a noxious weed locally in the USA. Teasels
are invading plains, waste grounds, old fields, pastures
and grow along edges of forests (Sforza, 2002). While
a large number of insects feed on teasels, no biological
control agent has up to now been introduced.
216
Potential biological control agents of field bindweed, common teasel and field dodder from Slovakia
Field dodder, Cuscuta campestris Yuncker (Cuscutaceae), is an annual stem parasite with leafless,
thread-like, orange or yellow stems that twine over
other plants. Field dodder is distributed worldwide and
has very low host specificity, attacking many different
host plants simultaneously. There are five dodder species known in Slovakia, but C. campestris is the only
introduced species; for more information on dodder,
see Tth and Cag (2001). C. campestris was introduced from North America to Europe in 1883 (Jehlk,
1998). It is not yet possible to provide an authoritative assessment of the biological control prospects for
dodder. The surveys that have been made give an indication of potential biological control candidates, but
knowledge of their host range and of the conditions
required for them to achieve effective suppression is
incomplete (CAB, 1987).
The main objective of this study was to determine
the most important insect guild feeding on the abovementioned weeds in Slovakia and to evaluate their potential use in classical or inundative biological control
programmes.
Materials and methods
Field bindweed
During the growing season of 1998 to 1999, 2002 to
2003 and 2005, three sites of field bindweed (Kamenica
nad Hronom, ajkov, Vrble) in southwest Slovakia
were checked weekly from mid-April until the beginning of October, following the natural phenology of the
plant. An additional seven locations were visited three
times at monthly intervals. The locations were chosen
in different geographic and climatic regions. Besides,
opportunistic sampling was conducted at numerous locations with different climates during different times
of the year. Collection sites were grassy or weedy
roadsides, fallow fields, C. arvensis-infested cropland
and vacant town lots. At each collection site, plants
were inspected for damage. Insects were collected by
sweeping (150 sweepings per site), or by aspirating or
hand-picking them from plants. The field surveys were
concentrated on the agromyzid fly, Melanagromyza
abocilia Hendel (Diptera: Agromyzidae) and tortoise
beetles of the genus Hypocassida (Coleoptera: Chrysomelidae). More detailed methods are described in Tth
et al. (2005) and Tth and Tthova (2006).
Preliminary host-specificity tests were conducted
with M. abocilia. Two economically important field
bindweed relatives in the Convolvulaceae, Ipomoea
batatas (L.) Lam. (sweet potato) and Ipomoea alba L.
(ornamental plant) were used in the experiments. Ten
potted field bindweed plants were exposed together
with ten plants of I. batatas and ten plants of I. alba at
each of three corn fields infested by field bindweed in
June 2002 and 2003. At the end of September, plants
exposed as well as 30 naturally growing bindweed
plants were evaluated for attack of Melanagromyza
albocilia.
Common teasel
During 2003 to 2004, the study was extended to
plants from the genus Dipsacus. Several surveys were
conducted from April to June in southwest Slovakia,
concentrating on flower-feeding insects. The plant was
mostly found in natural areas. To rear adult insects
and their parasitoids, flowerheads of D. fullonum were
collected and placed in glass boxes with a perforated
top under lab conditions (20C, 70% RH). A total of
200 flowerheads was collected. Emerged adults were
identified.
Field dodder
During the growing season 2001, 2003 and 2006,
the occurrence of insects feeding on field dodder was
observed irregularly in the agroecosystems of Slovakia following the natural phenology of dodders. A total of 82 localities were chosen in different geographic
and climatic regions throughout Slovakia. Collection
sites were field dodder-infested croplands planted with
various crops, fallow fields and roadsides. At each locality, Cuscuta species were identified. The localities
were inspected especially to record the presence of the
weevils from the genus Smicronyx (Coleoptera: Curculionidae). At each collection site, field dodder plants
were inspected for the presence of stem galls and galls
collected. To assess adult emergence, field-collected
stem galls were placed in plastic tubes (8 cm diameter,
4.5 cm high) with perforated tops for aeration and kept
in the laboratory at 20C 1C. Emerged adults were
identified.
Results
During the study, 108 organisms were collected from
field bindweed (see Tth and Cag, 2005 for details),
seven from teasel, and six from dodder.
Field bindweed
M. albocilia Hendel was the only species feeding
within the stems and roots of field bindweed in Slovakia. Between 10% and 100% of field bindweed plants
and up to 50% of the shoots were infested. The life history of M. albocilia and impact on the host were described in detail by Tth et al. (2005). M. albocilia was
found in 91 locations of 132, confirming it is a common insect in Slovakia and closely related to its host
plant C. arvensis. Seven species of Hymenoptera were
reared from pupae and larvae of M. albocilia as solitary,
larval and pupal parasitoids belonging to four families:
Aneuropria foersteri Kieffer (Diapriidae), Sphegigaster truncata Thomson Sphegigaster aculeata (Walker),
Cyrtogaster vulgaris Walker (Pteromalidae), Macro-
217
XII International Symposium on Biological Control of Weeds
neura (Eupelmus) vesicularis (Retzius) (Eupelmidae),
Chorebus cyparissa (Haliday) and Bracon picticornis
Wesmael (Braconidae). Parasitoids from the families
Braconidae and Pteromalidae predominated, making up 96.3% of parasitoids that emerged. S. truncata,
C. cyparissa and B. picticornis were the most abundant parasitoids reared from M. albocilia, accounting
for 36.3%, 35.6% and 21.8%, respectively, of the total
number reared. Preliminary studies conducted in Slovakia revealed that parasitoids suppressed 77% of the
stem miner population in the field.
During preliminary host-specificity tests, M. albocilia attack was found in both wild and artificially exposed field bindweed plants at the three field localities
but no attack on I. batatas and I. alba.
Tortoise beetles were important field bindweed defoliators. Seven species were found in association with
field bindweed in Slovakia. These were Cassida sanguinosa Suffrian, Cassida nebulosa Linnaeus, Cassida
stigmatica Suffrian, Cassida vibex Linnaeus, Cassida
murraea Linnaeus, Cassida viridis Linnaeus and Hypocassida subferruginea (Schrank) (Coleoptera: Chrysomelidae). The most abundant and widespread species was H. subferruginea. The species is recorded as
specific on field bindweed and Calystegia sepium (L.)
R. Br. (Convolvulaceae) in Slovakia (Tth and Tthov,
2006). Females deposited oval, light-red eggs in small
groups of two to five on the leaf surface of field bindweed (mostly on the underside) from mid-April to the
end of May. The larvae fed on leaves from mid-May
to the beginning of July. The highest number of adults
was found between 21 May and 11 June. For more details of the life history of H. subferruginea, see Tth
and Tthov, (2006). The larvae and adults were able
to almost completely destroy the leaves of field bindweed plants in some uncultivated habitats. The only
common natural enemy of H. subferruginea recorded
in Slovakia was the egg parasitoid Brachista pungens
(Mayr) (Hymenoptera: Trichogrammatidae).
Common teasel
Seven insect species were found associated with
common teasel in Slovakia. These were Macrosiphum
rosae (L.) (Sternorrhyncha: Aphididae), Metzneria
neuropterella Zeller (Lepidoptera: Gelechidae), Endothenia gentianaeana (Hbner), Diceratura ostrinana
(Guene), Cochylis roseana (Haworth) (Lepidoptera:
Tortricidae), Myelois circumvoluta (Fourcroy) (syn.
cribrumella, cribrella) and Homoeosoma nebulellum
(D. and Sch.) (Lepidoptera: Pyralidae). M. rosae and
D. ostrinana were feeding on leaves and rosettes. Immature stages of other moths occupied the flowerheads.
Although most of the species were rare and caused minor damage to the host plant, adult moths of E. gentianaeana and C. roseana were reared in high numbers
from D. fullonum flowerheads. Infestation of flowerheads by E. gentianaeana reached almost 100% ev-
erywhere in Slovakia. Larvae were feeding within the
central cavity of flowerheads and destroying the seeds.
On average, one larva damaged about ten seeds during
each attempt to cut an exit hole. Infestation of plants
infested by C. roseana ranged from 70% to 100%.
Larvae of C. roseana were found to feed gregariously
(730 larvae per flowerhead), and destroyed the largest number of seeds. During the study, several parasitoids were recovered from flowerheads. Considerable
parasitization was only recorded for E. gentianaeana,
where Glypta mensurator (Fabricius) (Hymenoptera:
Ichneumonidae) dominated.
Field dodder
Species from four orders were regularly found feeding on dodder plants, i.e. aphids (Sternorrhyncha), bugs
(Heteroptera), weevils (Coleoptera) and flies (Diptera).
Aphids mostly consisted of Aphis fabae Scopoli (Aphididae), bugs of Lygus rugulipennis Poppius (Miridae),
and the diptera were dominated by the stem-mining
fly, Melanagromyza cuscutae Hering (Agromyzidae)
(for details about M. cuscutae, see Tth et al., 2004a).
All three species were locally common but not harmful
for dodders. Weevils from the genus Smicronyx (Coleoptera: Curculionidae) were found to be the principal natural enemies of dodders in Slovakia. Larvae of
Smicronyx spp. caused stem galls on field dodder. Such
damage prevents flowering and fruiting of field dodder vines and destruction of flowers and seeds in other
Cuscuta. A total of 877 Smicronyx specimens were
reared from infested plants during the study. Smicronyx
jungermanniae (Reich) was the most abundant species,
accounting for up to 96% of the total number of weevils reared from field dodder galls. Smicronyx coecus
(Reich) and Smicronyx smreczynskii Solari were rare
and accounted for only 5.6% and 1.0%. Larvae of
S. jungermanniae and Smicronyx smreczinskii caused
stem galls on Cuscuta europaea L. and C. campestris as well as seed destruction on Cuscuta epithymum
(L.) L., and C. europaea. On the other hand, larvae of
S. coecus were found in flowers and seeds of C. epithymum and C. europaea only. Infestation of field dodder
by Smicronyx spp. ranged between 0% and 100%.
Discussion
Larvae of M. albocilia exclusively mined the stems
and root crowns of field bindweed in Slovakia. Spencer (1973) as well as Awadallah et al. (1976) stated the
same. Rosenthal and Buckingham (1982) listed C. arvensis and also Convolvulus althaeoides as its hosts.
M. albocilia therefore appears to be highly host specific on the target weed. Plants infested by M. albocilia looked healthy from the outside during its larval
stage. The stems started to become weak and dry after
pupation within the stems. In addition, exit holes may
facilitate infection by diseases (Tth, 2000). A complex
218
Potential biological control agents of field bindweed, common teasel and field dodder from Slovakia
of seven hymenopterous parasitoids was shown to have
a high impact on the populations of the stem-boring
fly (Tth et al., 2004b). Although infestation of field
bindweed was high, parasitism reduced the agromyzid
population by about 77%. In conclusion, M. albocilia
could be an important biological control agent of field
bindweed, especially in areas where the plant is invasive (e.g. North America), and parasitization of the fly
would be expected to be lower.
Seven species of tortoise beetles were collected during the surveys in Slovakia. The predominant and only
species attacking field bindweed was H. subferruginea.
This species occurs on field bindweed throughout the
Palearctic region (Kismali and Madanlar, 1990) and
is unable to complete development on sweet potato
(I. batatas; Rosenthal and Buckingham, 1982). In Slovakia, H. subferruginea was most frequent in warm
and dry regions, less widespread in temperate regions
and absent from cold regions (Tth, 2000). In natural
xerotherm ecosystems, the species often completely
defoliated plants, while in cultivated crops, the beetles
were not able to control their host plant. However,
their effect was clear in vineyards with living green
cover. Although predation and parasitization of tortoise
beetles is mentioned as a major factor lowering their
populations in the field, except for the egg parasitoid,
B. pungens, no parasitoids of larvae and adults were
recorded during our study. H. subferruginea was prioritized for the inundative biological control of field
bindweed in Slovakia.
For common teasel, the tortricid C. roseana was selected as a potential classical biological control agent
because of its ability to cause high seed reductions, its
common occurrence in Slovakia and its distinct preference for D. fullonum (Cheesman, 1996). While Sforza
(2002) concluded that C. roseana and E. gentianaeana
should be similar in their potential as biological control
agents for teasel, our results show that E. gentianaeana
is not very promising because one larva damaged only
about ten seeds.
Parasitic weeds (Cuscuta spp.) only reproduce by
seeds. Thus, complete biological control of these weeds
should be achievable by using organisms which damage
the seeds. In the absence of species that kill the weeds
at the seedling stage, the suppression of seed production is thus believed to be more important than damage to individual plants. Research efforts should also
be directed to continue investigations on phytophagous
arthropods, which can effectively be combined to provide maximum stress on parasitic weeds. Smicronyx
spp. prevent flowering and fruiting of field dodder, either directly, through their feeding activity in the seed
capsules, or indirectly, through weakening the shoots.
If the stem of the species is not attached to its host plant
beyond the attacked (galled) part, the entire section is
killed (Baloch et al., 1967). Thus, Smicronyx is able to
cause 100% seed reduction. In addition, they attacked
field dodder over the whole growing season. We expect
that Smicronyx species, above all Smicronyx jungermaniae, are very promising biological control agents of
field dodder.
Acknowledgements
The authors wish to thank Dr J. Luk for his help
in parasitoid and J. Cunev for weevil identifications.
Part of this work was supported by the Grant Agency
VEGA, project No. 1/3451/06.
References
Awadallah, K.T., Tawfik, M.F.S. and Shalaby, F.F. (1976) Insect fauna of bind-weed, Convolvulus arvensis L., in Giza,
Egypt. Bulletin de la Socit Entomologique dEgypte 60,
1524.
Baloch, G.M., Mohyuddin, A.I. and Ghani, M.A. (1967) Biological control of Cuscuta sp. II. Biology and host-plant
range of Melanagromyza cuscutae Hering (Dipt. Agromyzidae). Entomophaga 12, 481489.
CAB (1987) Digest: Potential for biological control of Cuscuta spp. and Orobanche spp. Biocontrol News and Information 8, 193199.
Cheesman, O.D. (1996) Life histories of Cochylis roseana
(Haworth) and Endothenia gentianaeana (Lepidoptera: Tortricidae) on wild teasel. The Entomologist 115,
6580.
Jehlk, V. (1998) Alien expansive weeds of the Czech Republic and Slovak Republic. Academia Praha, Czech Republic, 506 pp.
Kismali, S. and Madanlar, N. (1990) The role of Chrysomelidae (Coleoptera) species for the biological control of
weeds and the status of the species in Izmir. In: Proceedings, 2nd Turkish National Congress of Biological Control. September 2629, 1990, Turkey, pp. 299308.
Rees, N.E and Rosenthal, S.S. (1996) Field bindweed. In:
Rees, N.E., Quimby, P.C., Piper, G.L., Coombs, E.M.,
Turner, C.E., Spencer, N.R. and Knutson, L.V. (eds) Biological control of weeds in the West. Western Society of
Weed Science, Bozeman, MT, USA.
Rosenthal, S.S. and Buckingham, R.G. (1982) Natural enemies of Convolvulus arvensis in western Mediterranean
Europe. Hilgardia 5, 119.
Sforza, R. (2002) Candidates for the biological control of teasel, Dipsacus spp. In: Cullen, J.M., Briese, D.T., Kriticos,
D.J., Lonsdale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings of the XI International Symposium on Biological
Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 155161.
Spencer, K.A. (1973) Agromyzidae (Diptera) of Economic
Importance. W. Junk, The Hague, The Netherlands,
418 pp.
Tth, P. (2000) InsectsA Fresh Perspective in the Biological Control of Field Bindweed (Convolvulus arvensis L.).
PhD thesis. Slovak Agricultural University, Nitra, Slovakia, 229 pp.
Tth, P. and Cag, . (2001) Spread of dodder (Cuscuta
spp.) in the agroecosystems of Slovakia: is it an emerging
problem? Acta Fytotechnica et Zootechnica, 4 (Special
Number), 117120.
219
XII International Symposium on Biological Control of Weeds
Tth, P, ern, M. and Cag, . (2004a) First records of
Melanagromyza cuscutae Hering, 1958 (Diptera: Agromyzidae) from Slovakia and its new host plant. Entomologica Fennica 15, 4852.
Tth, P., Cristofaro, M. and Cag, . (2004b) Bionomy, seasonal incidence and influence of parasitoids of the field
bindweed stem borer fly Melanagromyza albocilia (Diptera:Agromyzidae) in Slovakia. In: Cullen, J.M., Briese,
D.T. Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott,
J.K. (eds) Proceedings of the XI International Symposium
on Biological Control of Weeds. CSIRO Entomology,
Canberra, Australia, pp. 351352.
Tth, P. and Cag, . (2005) Organisms associated with the
family Convolvulaceae and their potential for biological
control of Convolvulus arvensis. Biocontrol News and Information 26, 17N40N.
Tth, P., Cristofaro, M. and Cag, . (2005) Seasonal biology of Melanagromyza albocilia (Diptera: Agromyzidae)
and seasonal patterns of field bindweed infestation, under
field conditions in Slovakia. Entomologica Fennica 16,
254262.
Tth, P. and Tthov, M. (2006) Possibilities for biological control of field bindweed (Convolvulus arvensis
L.) by tortoise beetles (Chrysomelidae: Cassidinae).
In: Herda, G, Mazkov, J. and Zouhar, M. (eds) Proceedings of XVII Czech and Slovak Plant Protection
Conference. CAU, Prague, Czech Republic, pp. 528
532.
220
Lewia chlamidosporiformans,
a mycoherbicide for control of
Euphorbia heterophylla: isolate selection
and mass production
B.S. Vieira,1 K.L. Nechet2 and R.W. Barreto1
Summary
The potential of the fungus Lewia chlamidosporiformans Vieira and Barreto as a biological control
agent for wild poinsettia, Euphorbia heterophylla L., a noxious invader of soybean fields in Brazil, is
being assessed. One isolate was selected from nine that were tested as being the most aggressive to
a series of wild poinsettia populations (including one known to be herbicide-resistant). The biphasic
technique was investigated as an option for mass production of conidia of L. chlamidosporiformans.
This method involves the production of mycelia in liquid culture that are later blended and poured
into trays containing a solid medium and incubated under a specific light regime until conidia form.
After 3 days, conidia are harvested once per day by pouring sterile water over the surface of the colonized medium and scraping the surface with a rubber spatula. A semi-synthetic liquid medium (with a
sucrose and asparagin base) was selected as the best for the first phase of growth. A vegetable brothagar medium supplemented with CaCO3 was the best solid medium for fungal growth and sporulation
in the second phase.
Keywords: bioherbicides, pathogenicity, biphasic.
Introduction
Losses caused by weeds represent one of the main
limiting factors in agriculture production worldwide.
Chemical herbicide applications are gradually becoming the dominant method of control of weeds in both
developed and developing countries (Wyse, 1992; Abernathy and Bridges 1994). However, parallel to this,
problems with contamination of water resources, accumulation of chemical residues in the soil, emergence
of herbicide resistence in weed species and threats
to biodiversity are also on the rise. This justifies the
search for alternatives that might allow the reduction or
replacement of chemical herbicide applications such as
through biological control of weeds with plant pathogens (Rosskopf et al., 1999).
Wild poinsettia (Euphorbia heterophylla L.), known
in Brazil as amendoim-bravo or leiteiro, is a native
euphorb of tropical and subtropical America (Lorenzi,
Universidade Federal de Viosa, Departamento de Fitopatologia, CEP
36571-000, Viosa, MG, Brazil.
2
Embrapa Roraima, CEP 69301970, Boa Vista, RR, Brazil.
Corresponding author: B.S. Vieira <[email protected]>.
CAB International 2008
1
2000). In Brazil, it is regarded as one of the worst weeds
in important crops such as corn, sugarcane, common
bean and soybean (Guedes and Wiles, 1976; Arevalo
and Rozanski, 1991). The reduction in the soybean harvest caused by competition with E. heterophylla varies depending on the weed density in an invaded area
and the soybean cultivar, but it is estimated that losses
range from 35% to 62% (Constantin et al., 1997; Voll
et al., 2002). Acetolactate synthase (ALS) inhibiting
herbicides have been the favorite product used in the
control of E. heterophylla in soybean. However, the repetitive use of these products and their residual effect
in the soil led to a continuos selection of populations of
E. heterophylla that are now resistant to these products
(Gazziero et al.1998; Melhorana and Pereira, 1999).
Such a situation offers an ideal opportunity for the use
of a fungus formulated as a mycoherbicide (Charudattan, 2001).
Lewia chlamidosporiformans Vieira and Barreto is
a newly described fungus capable of causing severe inflorescence necrosis, foliar blight and stem canker on
E. heterophylla under natural conditions. Since its discovery, it has been intensively evaluated as a potential
mycoherbicide. This paper reports some of the results
obtained during these studies, namely isolate selection
221
XII International Symposium on Biological Control of Weeds
and development of a method for mass production of
L. chlamidosporiformans inoculum.
Materials and methods
Fungal isolates
Samples of E. heterophylla with symptoms of attack by Lewia were collected in the Brazilian states of
Minas Gerais, Rio de Janeiro, and Rio Grande do Sul.
Direct isolation from sporulating lesions as well
as indirect isolations through surface sterilization and
plating of diseased tissues was performed in vegetablebroth agar (VBA; Pereira et al., 2003). Cultures were
preserved in silica gel as described in Dhingra and Sinclair (1995).
Inoculum production
Conidia of L. chlamidosporiformans of all fungal
isolates were produced for the first isolate-selection experiment using the methodology described by Walker
(1980) with the following modification: ten culture
disks obtained from the margin of 7-day-old cultures
grown in VBA were transferred to a series of Erlenmeyer flasks containing 100 ml of VB, i.e., the same as
described in Pereira et al. (2003) but without agar. The
erlenmeyers were left on a shaker at 140 rpm for 7 days
at room temperature. After this period, the mycelial
mass was blended within the remaining liquid medium
within each flask and poured onto 20 28 cm aluminum trays, each containing 100 ml of solidified VBA.
Trays were kept in a controlled temperature room at
26 2C under a 12-h photoperiod (light from two 40-W
daylight fluorescent lamps and two 40-W fluorescent,
near-ultra-violet light lamps). After 2 days, conidia
were collected by pouring 50 ml of sterile water on the
culture surface and scraping it with a rubber spatula.
The resulting suspension was then filtered through two
layers of cheese cloth, and the final concentration of the
suspension was evaluated and adjusted to the adequate
concentration for use in the experiment.
Table 1.
Code
EKLN16
EKLN19
EKLN247
ERWB274
ERWB280
ETSB
ETRB
ESH
ERH
E. heterophylla plants for the experiments
The populations represented in the experiment were
produced from seeds obtained from different locations
and included plants with the following characteristics:
resistance to the herbicide imazetaphyr and resistance
to Bipolaris euphorbiae (Hansford) Muchovej, a fungus
previously evaluated as a mycoherbicide for wild poinsettia (Yorinori and Gazziero, 1989; Marchiori et al.,
2001, Nechet et al., 2006). Seeds to be used in the experiments (Table 1) were harvested from healthy plants
grown in a greenhouse and stored at 5C until use.
Plants used in the experiments were produced from
pre-germinated seeds that were planted in 500-ml pots
containing sterile soil. The plants were maintained in a
greenhouse (26 2C) and watered daily. Plants were
inoculated at the three- to four-leaf stage.
Screening of fungal isolates
Groups of plants of nine populations listed in Table 1
were inoculated with conidial suspensions representing
each isolate obtained in the survey. Inoculum consisted
of suspensions of 1.0 104 conidia/ml + 0.05% Tween 20
(polyoxyethylene monolauratic) + 0.05% Breakthru
(polyether-polymethyl siloxane copolymer + polyether;
T.H. Goldschmidt, Guarulhos, So Paulo). After inoculation, plants were kept in a mist room at 25C for 24 h
and then moved to a greenhouse (26 2C). Plants inoculated with a suspension with the same components
as described above but without L. chlamidosporiformans conidia served as the control. The number of dead
plants and percentage of diseased leaves (proportion of
number of diseased leaves per total number of leaves)
were evaluated at 5-day intervals for 30 days, and the
area under the disease progress curve (AUDPC) was estimated (Campbell and Madden, 1990).
The experiment was carried out in a completely randomized design with a factorial of nine isolates, nine
plant populations and three replications per treatment.
Each replicate consisted of one pot containing two
plants.
Euphorbia heterophylla populations included in the study.
Origin
Niteri-RJ
Viosa-MG
Itabuna-BA
Nova Laranjeira-PR
Nova Petrpolis-RS
Londrina-PR
Londrina (resistant to B. euphorbiae)
Viosa-MG
Viosa-MG
222
Lewia chlamidosporiformans, a mycoherbicide for control of Euphorbia heterophylla
Evaluation of liquid-culture media
on the mycelial growth of
L. chlamidosporiformans
for 24 h. The experiment had a completely randomized
design with four replications.
This experiment aimed to evaluate the growth of
L. chlamidosporiformans (isolate KLN-06) in a series of five common liquid-culture media of different
compositions (see below for details) at either standard
or double concentration. This test aimed to determine
the composition of a liquid medium, from among the
following, that might be adequate for mass production
of mycelium for the first stage of a biphasic-technique
(Walker, 1980):
1. PS (Dhingra and Sinclair 1995, standard concentration): decoction of 200 g potato; 20 g sucrose;
1 l distilled water;
2. PS 2 (doubled concentration): decoction of 400 g
potato; 40 g sucrose; 1 l distilled water;
3. VBS (VB as mentioned above supplemented with
sucrose)-100 ml vegetable broth; 20 g sucrose;
450 ml distilled water;
4. VBS 2 (doubled concentration)- 200 ml vegetable broth; 40 g sucrose; 450 ml distilled water;
5. Marine ammonium mineral salt (MAMS; standard
concentration): decoction of 200 g castor-bean
plant leaves; 20 g sucrose; 1 l distilled water;
6. MAMS 2 (doubled concentration): decoction
of 400 g castor-bean plant leaves; 40 g sucrose;
1 l distilled water;
7. MANDS (standard concentration): decoction of
200 g cassava leaves; 20 g sucrose; 1 l distilled
water;
8. MANDS 2 (doubled concentration): decoction
of 400 g cassava leaves; 40 g sucrose; 1 l distilled
water;
9. MSSA: semi-synthetic sucrose-asparagin medium
(Alfenas, 1998; normal concentration): 10 g sucrose, 2 g l-asparagin, 2 g yeast extract; 1 g KH2
PO4 ; 0.1 g MgSO47H2O; 0.44 mg ZnSO47H2O;
0.48 mg FeCl36H2O; 0.36 mg MnCl2H2O; 1 l distilled water.
10. MSSA 2 (doubled concentration): 20 g sucrose, 4
g L-asparagin, 4 g yeast extract; 2 g KH2PO4; 0.2g
MgSO47H2O; 0.88 mg ZnSO47H2O; 0.96 mg
FeCl36H2O; 0.72 mg MnCl2H2O; 1 l distilled
water.
Three mycelial plugs from 7-day-old cultures grown
on PDA were transferred to 125-ml Erlenmeyer flasks,
each containing 50 ml of one of the liquid-culture media described above. The flasks with plugs were placed
on a shaker at 100 rpm at room temperature (25C).
Dry weight of mycelia produced was evaluated after
7 days of incubation. The contents from each flask
were vacuum filtered through filter paper until the mycelium was dry. The mycelial mass was then scraped
and weighed after drying in an electric oven at 70C
Effect of different solid-culture media on
sporulation of L. chlamidosporiformans
This experiment aimed to determine the influence
three different solid-culture media (supplemented or not
with CaCO3 at 3 g/l), on the sporulation of L. chlamidosporiformans during the second growth phase of the
biphasic system of mass production. The following culture media were tested:
1. Concentrated PCA: decoction of 200 g potato; 200 g
carrot; 1 l distilled water; 20 g agar;
2. Concentrated PCA + CaCO3
3. VBA (Pereira et al. 2003)
4. VBA + CaCO3
5. FLA: decoction of 200 g wild poinsettia leaves,
20 g sucrose, 1 l distilled water.
6. FLA + CaCO3
The medium utilized during the first phase (mass
production of mycelium in liquid culture) was MSSA
(described above), and the procedure was also as described above. The mycelium was blended inside
of the erlenmeyers, and 100 ml of the resulting suspension was poured into each of 24 aluminum trays
(35 20 cm), containing 100 ml of the solid media
that were being tested. Trays were kept in a controlled
temperature room at 26 2C under a 12-h photoperiod. After 3 days, conidia were harvested once per day
by pouring 50 ml of sterile water over the surface of
the colonized medium and scraping the surface with
a rubber spatula, with a total of four harvests per tray.
The obtained suspension was filtered through three layers of cheesecloth. An aliquot of 20 mL of the conidial
suspension obtained from each tray was removed and
mounted on a microscope slide, and the number of
conidia produced was counted, with results converted
into conidia/ml from each treatment. The experiment
had a completely randomized design with four replications and each tray represented a replicate. Statistical
analysis was made of the sum of conidial concentrations obtained for the four harvests.
Results
Selection of a fungal isolate
Nine isolates were obtained from several locations (Table 2). Among these isolates, only three were
pathogenic to all E. heterophylla accessions that were
screened. Isolate KLN06 caused the highest diseaseintensity levels resulting in larger values of AUDPC
for five of the wild poinsettia populations involved in
the test, including population ERH (resistant to the herbicide imazethaphyr) and was equivalent to other isolates in disease severity caused to the remaining weed
223
XII International Symposium on Biological Control of Weeds
Table 2.
Lewia chlamidosporiformans isolates included in the study.
Code
KLN06
KLN09
KLN14
KLN15
KLN17
KLN18
KLN19
KLN20
RWB280
Origin
Viosa-MG
Araruama-RJ
Italva-RJ
Niteri-RJ
So Miguel do Anta-MG
Viosa-MG
Viosa-MG
Viosa-MG
Nova Petrpolis-RS
populations (Figure 1). The first symptoms appeared
5 days after inoculation with KLN06, and plant death
started appearing 7 days after inoculation in populations EKLN19, ERWB247, ETSB, ERH and ESH.
KLN06 was selected as the most promising isolate for
further studies.
Effect of different solid-culture media on
sporulation of L. chlamidosporiformans
Evaluation of liquid-culture media
on the mycelial growth of
L. chlamidosporiformans
The liquid-culture media that resulted in the hightest
levels of L. chlamidosporiformans mycelial production
was MSSA 2 (Figure 2). Other media yielded inferior results for the production of L. chlamidosporiformans mycelia, reaching values that varied from 1/3 to
1/2 that obtained with MSSA 2. A smaller production
of mycelial mass was obtained for: PS, MAMS and
MAMS 2. The doubled concentration of the ingredients in the liquid-culture media only resulted in significant increase in the yield of mycelial biomass for
Figure 1.
MSSA 2 and PS. Mycelial production for MSSA 2
was triple that obtained for standard MSSA (Figure 2).
MSSA 2 was hence selected for use as medium in the
first phase of biphasic mass production of L. chlamidosporiformans.
Results obtained in this experiment are presented in
Figure 3. Among the solid-culture media that were tested
for sporulation in the second phase of the biphasic
mass production of L. chlamidosporiformans, VBA +
CaCO3 had the best performance. Its use resulted in a
production of 8.1 105 conidia/ml and was followed
by PCA + CaCO3 (7.03 105 conidia/ml) and FLA +
CaCO3 (5,6 105 conidia/ml). There were significant
statistical differences among most culture media being
tested (Figure 3). The supplementation of CaCO3 (3g/l)
significantly increased the production of spores of the
fungus for all solid-culture media being tested, and this
Area under the disease progress curve for the isolates KLN06, KLN09, and KLN17 of Lewia chla
midosporiformans based on disease severity (means of three repetitions; bar standard deviation).
224
3,5
3
2,5
2
1,5
bcd
bc
cd
bc
0,5
BS
BS
V
PS
PS
2
x
M
S
M
A
M
A
x
S
D
M
A
M
S
S
D
N
M
A
M
SS
M
SS
Dry mycelial mass (g)/50 mL of
media
Lewia chlamidosporiformans, a mycoherbicide for control of Euphorbia heterophylla
Liquid culture media
Production of mycelial mass of Lewia chlamidosporiformans in different liquid-culture media (means of four
repetitions, bars standard deviation, means followed by the same letter did not differ under Tukey test at the level
of 5% of probability).
Sporulation (x100000 con/mL)
Figure 2.
10
8
b
c
4
2
0
VBA +
CaCO3
VBA
PCA conc.+ PCA conc.
CaCO3
FLA +
CaCO3
FLA
Solid culture media
Figure 3.
Lewia chlamidosporiformans conidial production on different solid-culture media (means of four repetitions,
bars standard deviation, means followed by the same letter did not differ under Tukey test at the level of 5% of
probability).
was particularly significant for VBA. The addition of
CaCO3 (3g/l) increased conidial production in VBA
16-fold. It also increased conidial production for FLA
by a factor of 9.3 and by a factor of 1.24 for PCA.
Acknowledgements
Seeds used in experiments were provided by the Laboratrio de Herbicida na Planta, Departamento de Fitotecnia, Universidade Federal de Viosa or by J.T. Yorinori (Embrapa Soja). The authors acknowledge CNPq
and CAPES for financial support.
References
Abernathy, J.R. and Bridges, D.C. (1994) Research priority
dynamics in weed science. Weed Technology 8, 396399.
Alfenas, A.C. (1998) Eletroforese de Isoenzimas e Protenas
Afins; Fundamentos Aplicaes em Plantas e Microrganismos. Editora UFV, Viosa,-MG, Brazil, 574 pp.
Arevalo, R.A. and Rozanski, A. (1991) Plantas Daninhas na
cultura do feijo. In Anais do 4 Seminrio sobre Pragas e
Doenas do Feijoeiro. Campinas: Secretaria da Agricultura e Abastecimento. Campinas, So Paulo, pp. 3343.
Campbell, C.L. and Madden, L.V. (1990) Introduction to
Plant Disease Epidemiology. John Wiley & Sons, New
York, NY, 532 pp.
Charudattan, R. (2001) Biological Control of weeds by means
of plant pathogens: significance for integrated weed,
management in modern agro-ecology. BioControl 46,
229260.
Constantin, J., Contiero, R.L., Demeis, M., Ita, A.G. and
Maciel, C.D.G. (1997) Controle de Euphorbia heterophylla e fitotoxicidade dos herbicidas imazamox e imazethaphyr na cultura da soja (Glycine max L. Merril).
In: Resumos do XXI Congresso Brasileiro da Cincia
das Plantas Daninhas, p. 451. Caxambu, Minas Gerais,
Brazil.
Dhingra, O.D. and Sinclair, J.B. (1995) Basic Plant Pathology Methods. CRC Press, New York, NY, 434 pp.
Gazziero, D.L.P., Brighenti, A.M., Maciel, C.D.G., Christofolleti, P.J., Adegas, F.S. and Voll, E. (1998) Resistncia
225
XII International Symposium on Biological Control of Weeds
de amendoim-bravo aos herbicidas inibidores da enzima
ALS. Planta Daninha 16, 117125.
Guedes, L.V. and Wiles, T.L. (1976) Controle de plantas
daninhas em plantio direto de soja: avaliao em escala
comercial em fazendas. In Resumos do XI Seminrio
Brasileiro de Herbicidas e Ervas Daninhas, p.131. Londrina, Paran, Brazil.
Lorenzi, H.J. (2000) Plantas Daninhas do Brasil: Terrestres,
Aquticas, Parasitas, Txicas e Medicinais. Instituto
Plantarum, Nova Odessa, SP, 648 pp.
Marchiori, R., Nachtigal, G.F., Coelho, L., Yorinori, J.T. and
Pitelli, R.A. (2001) Comparison of culture media for the
mass production of Bipolaris euphorbiae and its impact
on Euphorbia heterophylla dry matter accumulation.
Summa Phytopathologica 27, 428432.
Melhorana, A.L. and Pereira, F.A.R. (1999) Eficincia do
herbicida lactofen no controle de Euphorbia heterophylla,
resistente aos herbicidas inibidores da enzima acetolactato sintase (ALS). Documentos-Embrapa Agropecuria
Oeste 3, 1114.
Nechet, K.L., Barreto, R.W. and Mizobuti, E.S. (2006) Bipolaris euphorbiae as a biological control agent for wild
poinsettia (Euphorbia heterophylla): host-specificity and
variability in pathogen and host populations. BioControl
51, 259275.
Pereira, J.M., Barreto, R.W., Ellison, A.C. and Maffia, L.A.
(2003) Corynespora casiicola f. sp. lantanae: a potential
biocontrol agent from Brazil for Lantana camara. Biological Control 26, 2131.
Rosskopf, E.N., Charudattan, R. and Kadir, J.B. (1999) Use
of plant pathogens in weed control. In: Fisher et al. (eds)
Handbook of Biological Control. Academic Press, San
Diego, pp. 891917.
Voll, E., Gazziero, D.L.P., Brighenti, A.A.M. and Adegas, F.S.
(2002) Competio relativa de espcies de plantas daninhas
com dois cultivares de soja. Planta Daninha 20, 1724.
Yorinori, J.T. and Gazziero, D.L.P. (1989) Control of wild poinsettia (Euphorbia heterophylla) with Helminthosporium
sp. In: ed. Delfosse, E.S.(ed. Delfosse, E.S.) Proceedings
of the VII International Symposium on Biological Control
of Weeds, pp.571576. Rome Istituto Sperimentale per la
Patologia Vegetale, Rome, Italy.
Walker, L. (1980) Production of spores for field studies. Advances in Agricultural Technology 12, 15.
Wyse, D.L. (1992) Future of weed science research. Weed
Technology 6, 162165.
226
Sphenoptera foveola (Buprestidae)
as a potential agent for biological control
of skeletonweed, Chondrilla juncea
M.G. Volkovitsh,1 M.Yu Dolgovskaya,1 S.Ya Reznik,1
G.P. Markin,2 M. Cristofaro3 and C. Tronci4
Summary
Skeletonweed, Chondrilla juncea L. is an important invasive weed in the USA, Australia, and Argentina. With the aim of finding new potential agents for biological control of this weed, surveys were
carried out in 2004 to 2005 in its native range in Southern Russia and Kazakhstan, where the bronze
skeleton weed root borer, Sphenoptera foveola (Gebler) (Coleoptera: Buprestidae) was repeatedly
collected from different Chondrilla species. According to the literature and our survey, this buprestid
is widely distributed in sandy deserts of Southern Russia and Kazakhstan. Locally, it could be rather
abundant. Observations suggest that both larvae and adults of S. foveola feed exclusively on plants of
the genus Chondrilla. Adults feed on green stems, larvae feed externally (within latex case) on roots
and, at high population density, can cause significant damage to attacked plants. We conclude that
S. foveola should be considered as a potential agent for biological control of skeleton weed, although
further studies (particularly, host-specificity tests) are necessary to prove this hypothesis. Sphenoptera (Deudora) clarescens Kerremans, another sphenopteran species attacking Chondrilla in Iran and
Turkey, may have a different root-feeding strategy and invites further investigations to evaluate it as
a potential agent for biological control of skeleton weed.
Keywords: bronze skeleton weed root borer, surveys, taxonomy, distribution, biology,
host range, impact.
Introduction
Chondrilla juncea L. (Asteraceae), skeletonweed, is an
important invasive weed in the western USA, Australia,
and Argentina. With the aim of finding potential agents
for biological control of this weed, extensive surveys
have been carried out in 2004 and 2005 in its native
range in Southern Russia and Kazakhstan. Among
other phytophagous insects, the bronze skeleton weed
root borer, Sphenoptera foveola (Gebler, 1825) (Coleoptera: Buprestidae) was repeatedly collected from
different Chondrilla species. In an earlier programme
this buprestid had been considered as a potential candi-
Zoological Institute, St. Petersburg, Russia.
US Forest Service, Forestry Science Laboratory, Bozeman, MT, USA.
3
ENEA-Casaccia; BBCA, Rome, Italy.
4
Biotechnology and Biological Control Agency, Sacrofano, Rome, Italy.
Corresponding author: M.G. Volkovitsh <[email protected]>.
CAB International 2008
1
2
date for biological control (Caresche, 1970; Wapshere,
1973, 1974). However, no attempts were made to test
this species. Another root boring buprestid, Sphenoptera (Deudora) clarescens Kerremans, 1909 distributed in Iran and Turkey was considered and tested,
but dropped as a potential biological control candidate
(Hasan, 1978). From a biological aspect S. foveola is
one of the best studied species of Sphenoptera due to
the ability of its larvae to induce latex secretion from
Chondrilla roots. In the 1930s, an ambitious project
was conducted in the former Soviet Union to use native
Central Asian latex-producing plants (mainly Scorzonera, Chondrilla, and Taraxacum) for rubber production. This project resulted in extensive investigations
of selected plants and associated insects including
S. foveola (Emelianova et al., 1932).
To evaluate the potential perspectives of S. foveola
as a candidate for biological control of C. juncea, field
collections and biological observations were made in
2004 to 2005 in Kazakhstan and Russia.
227
XII International Symposium on Biological Control of Weeds
Methods and materials
The first trip to Kazakhstan was made between June 23
and July 15, 2004 and included two 6-day field trips
and several 1-day visits. Total extent of the route covered was 4230 km, in which 41 sites were surveyed
and S. foveola was found at eight. The second trip from
May 10 to May 22, 2005 included one 7-day field trip
and two short daily visits; total distance covered was
approximately 1500 km with 18 sites being visited at
which S. foveola was found at four. These trips covered
mainly low land and foothill desert areas (elevation,
4001385 m) between Ili River on the west, the artificial Qapshagay Lake on the south and Balqash Lake on
the north (Figure 1). Numerous species of Chondrilla
were abundant in the study area. Adults of S. foveola
were collected by hand from Chondrilla plants or from
the soil under the plants where they were apparently
ovipositing. To collect preimaginal stages, the roots of
different Chondrilla species were excavated, examined
and if necessary, dissected.
To study the geographic distribution of S. foveola,
the buprestid collection of Laboratory of Insect Taxonomy, Zoological Institute, Russian Academy of Sciences, St. Petersbourg, Russia (ZIN) was examined,
Figure 1.
all collecting data were studied and collecting localities mapped (Figure 1). Confirmation of identification
was performed by comparison with syntype from ZIN
collection using a dissecting microscope. Chondrilla
species collected were identified by botanists from Botanical Institute, Russian Academy of Sciences, St. Petersbourg, Russia, and A. Popov, Volgograd, Russia.
Results and discussion
Taxonomy
Taxonomical position of this buprestid is as follows:
Coleoptera: Buprestidae: Chrysochroinae: Sphenopterini (Volkovitsh and Kalashian, 2006).
Sphenoptera (Sphenoptera) foveola Gebler, 1825:
46 (as Buprestis). 1 Syntype: Steppe Kirgiz., Gebler
(Zoological Museum, Helsinki University, Finland);
1 syntype: the same label (ZIN).
Synonym: strandi Obenberger, 1920: 113. 2 syntypes: Tarbagatai, Siberia [Tarbagatai Mts., Kazakhstan] (National Museum, Praha, Czech Republic).
Synonym: foveola var. usta Obenberger, 1927: 51.
1 syntype (female): Astrakhan (National Museum,
Praha, Czech Republic).
The distributional map of bronze skeleton weed root borer, Sphenoptera foveola. Closed triangles refer to our
field collections in 2004 to 2005; open triangles refer to the data from the insect collection of Zoological Institute,
St.Petersburg, Russia.
228
Sphenoptera foveola (Buprestidae) as a potential agent for biological control of skeletonweed, Chondrilla juncea
Larva: Alexeev et al., 1990 (detailed morphological
description).
We looked for but did not find another sphenopteran
species, S. (Deudora) clarescens Kerr., reported by
Hasan (1978), to be found in Iran and Turkey by Kashefi
(2002) and C. Tronci (unpublished communication).
This species tunnels inside the roots of C. juncea which
is different from that of S. foveola which feed externally
on the roots. S. clarescens, according to Hasan (1978),
also has a very wide host range among non-Chondrilla
genera of plants which would make it an unsuitable
biological control agent. We did not find evidence of a
Sphenoptera that fit the description of S. clarescens in
our surveys in either Kazakhstan or southern Russia or
find evidence in museum collection of it having been
collected in any former USSR countries.
Geographical distribution
According to the literature, S. foveola has been
found in southeastern European Russia (Astrakhan
prov.), Armenia, Kazakhstan, Kyrgyzstan, Uzbekistan,
Turkmenistan (Alexeev et al. 1990, Kadyrbekov and
Tleppaeva 2004, Volkovitsh and Kalashian 2006). The
great majority (28 specimens examined) were from
east of the Caspian Sea or the Volga River (Figure 1).
We did not find any specimens from the coastal sandy
semi-desert areas around the west side of the Caspian
Sea but we suspect that it would be found in this area
as well as in the semi-desert and desert habitats along
the valleys of the Kura and Arax rivers in the Armenian
Mountains. This is based upon the finding of two specimens of S. foveola in the ZIN collection from Armenia,
the only two specimens from west of the Caspian Sea.
Biology
Our field observations and the literature (Emelianova
et al., 1932) suggested that adult S. foveola are present in the field from mid-April to mid-October, where
they feed on growing Chondrilla plants in hilly, sandy
deserts (Figures 2, 3a). In the hottest part of the day we
found adult beetles usually sitting, head down on the
stems of Chondrilla plants or in the plants shadow on
the ground probably to protect themselves from overheating. Under laboratory conditions, oviposition was
observed from mid-May until the beginning of October.
In the field, eggs were laid in the sand near the crown
or directly on the crown of the plant. In the laboratory,
oviposition of an individual female lasted 10 to 50 days
during which she laid 11 to 135 eggs. In the field, peak
oviposition activity appeared to occur in June and July.
The eggs take 12 to 14 days to develop in July but more
than 30 days in September. Mortality of eggs laid both
in the laboratory and collected in the field was approximately 25%. By the end of the summer, 70% of the
plants examined had been attacked but in some sites
this could reach 90% to 97% (n > 50).
Newly eclosed, first instar larvae migrate down the
outside of the taproot working their way down between
the grains of sand to a depth of two to three centimeters where they chew into the cortex on the Chondrilla
roots. Their attack provokes an extreme secretion of
latex which when mixed with sand forms a small, porous case one to one and one-half cm long in which the
larvae develop. As the larvae grow, they move spirally
down the outside of the root, further damaging the cortex and releasing more latex which increases the size
of the case, which can now surround the root (Figures
3b, c). A single case can be up to 30 cm long and often when several larvae attack the same plant and their
combined cases fusing forming congealed latex/sand
lump that can be as big as 225 grams. It was noticed that
plants growing in loose stands have both a higher rate
of attack and larger than in plants growing in compact,
hardened sand. The cases contain hollow chambers in
which the larvae move freely and extend from near the
surface to over 30 cm deep. At the deepest points, they
probably experience a constant temperature of 20C to
25C which may be optimal for larval development. In
the field, larvae usually eclose in May or June, feed during the summer, over-winter as larvae inside the case,
and pupate the following spring. Adults emerge from
the pupae at the beginning of summer, usually June and
July. Larvae that hatch from eggs laid in midsummer
over-winter and continue to feed in the following season. However, larvae from the first eggs laid in spring
may complete their development and produce adults by
the end of the summer which then may over-winter.
Host range
According to our field observations,S. foveola fed
and developed mainly on Chondrilla ambigua Fisch.
ex Kar. et Kir., and, occasionally, on C. canescens Kar.
Figure 2.
229
Typical collection site for the bronze skeleton
weed root borer, Sphenoptera foveola, in hilly
desert. Kazakhstan, Site KZ04-08, Almaty Region, Qumbasy sandy desert, 69 km NNW of
Qapshagay, 28.06.2004.
XII International Symposium on Biological Control of Weeds
Figure 3.
The bronze skeleton weed root borer, Sphenoptera foveola: (a) the adult; (b) mature larva; and (c) latex and sand
cases of the mature larvae. The insect was feeding on Chondrilla ambigua, Almaty Region, Kazakhstan.
et Kir. Additionally, S. foveola was reported from the
literature to have been found feeding on C. pauciflora
Ldb. and more rarely on C. brevirostris Fisch. et Mey.
(singularly) (Emelianova et al., 1932), but not on C.
juncea. The genus Chondrilla, is taxonomically divided
into several divisions, section Brachyrhynchus contains
C. ambigua and C. pauciflora. Section Euchondrilla
contains C. brevirostris and C. canescens which also
contains C. juncea L. (Iljin, 1930, Flora SSSR, 1964).
S. foveola feeding on a non-Chondrilla species,
Scorzonera tau-saghyz Lipschiz et Bosse, was reported by Alexeev et al. (1990) although they give no
data to support this claim. This doubtful host use was
repeated by Tleppaeva (1999), Kadyrbekov and Tleppaeva (2004), and Tleppaeva and Ishkov (2004) but
without any additional substantiating data. Our field
observations agree with Emelianova et al. (1932) who
suggested that only Chondrilla species are suitable for
S. foveola larval development, with feeding primarily
concentrated on species of the Brachyrhynchus section
with fewer records from Chondrilla in the Euchondrilla
section.
Impact on the host plant
Chewing by adults on Chondrilla shoots often kills
small branches which break off when dry. However,
most of the damage is done by larvae feeding on the
roots. The damage from a single small larva is usually
not fatal but damage by mid-sized larva particularly if
there are several, often kills the above ground portion
of smaller plants and some times larger stems of bigger
plants. The copious flow of latex exuded by the wounded
roots represents a major loss of nutrients and energy.
This loss of latex if does not kill the plant, must stress
it, which should result in reduced growth, flowering,
and general loss of competitiveness. Also, increased
stress from the loss of latex probably would weaken
the plants making them more susceptible to the three
existing biological control agents already established
in many parts of Chondrilla junceas introduced range
(Julien and Griffiths, 1998).
Discussion
S. foveolas natural range, at least where it is most
abundant, seems to extend from the Volga River and
Caspian Sea eastward through the deserts of Kazakhstan, possibly northern most parts of Turkmenistan and
Uzbekistan, as well as the southern most part of Russia adjacent to Kazakhstan. This area is probably the
centre of origin for the genus Chondrilla since it contains approximately 18 of the 21 known species of this
plant (Flora SSSR, 1964). By contrast, the target weed,
C. juncea is one of the few species not found in this
area. C. junceas range extends eastward across Europe
from Spain, along the borders of the Mediterranean Sea
(Wapshere et al., 1974). It is also found in the Balkans,
Turkey and Iran and along the north shore of the Black
Sea (Wapshere et al., 1976). The recorded eastern edge
of its range ends approximately at the edge of the Caspian Sea and the Volga River (our review of herbarium
species). It is therefore unfortunate that the range of S.
foveola does not naturally extend westward far enough
for it to overlap that of C. juncea.
The wide range of other species of Chondrilla which
S. foveola can attack including the very closely related
species C. brevirostris and C. canescens make us suspect that under the right climate conditions, S. foveola
will probably also attack C. juncea. We are presently
planning feeding studies to determine the suitability
of C. juncea as a new association host for S. foveola.
The fact that S. foveola can attack a number of different
species of Chondrilla should not prohibit it from being
considered as a biological control agent. In all parts of
the introduced range of C. juncea there are no native
or introduced species of Chondrilla which might be
230
Sphenoptera foveola (Buprestidae) as a potential agent for biological control of skeletonweed, Chondrilla juncea
at risk from this attack. Its potential for attacking species other than in the genus Chondrilla, we feel is also
low since we found the only record in the literature being Scorzonera tau-saghyz reported by Alexeev et al.,
(1990). We hope to address the question of its potential
for non-Chondrilla host attack in future host testing.
We are also planning to find and compare S. foveola
with the related species of Sphenoptera clarescens reported attacking Chondrilla in Turkey and Iran.
Acknowledgements
We would like to thank Alexander Popov (Volgograd),
botanist and our guide in Lower Volga and Don Basins
(Southern-East Russia); Dr Roman Yashchenko, President of Tethys Scientific Society (Almaty, Kazakhstan)
for the great assistance in arranging our trips to Kazakhstan. The major sources of funding for this study
are the Idaho Rush Skeletonweed Task Force, the Idaho
Department of Agriculture and the Forest Service,
Rocky Mountain Research Station. We also gratefully
acknowledge the help and support of the USDA ARS
foreign program office in Beltsville, MD, and Dr Walker
Jones, Director of EBCL, Montpellier, France.
References
Alexeev, A.V., Zykov, I.E. and Soyunov, O.S. (1990) Novye
materialy po lichinkam zlatok roda Sphenoptera Sol.
(Coleoptera, Buprestidae) pustyn Zakavkazya, Kazakhstana i Srednei Azii. Izvestiya akademii nauk Turkmenskoi
SSR 3, 3038 (in Russian).
Caresche, L. (1970) The biological control of Skeleton weed,
Chondrilla juncea L. Entomological aspects. In: Simmonds, F.J. (ed.) Proceedings of the First International
Symposium on Biological Control of Weeds. European
Station, CIBC, Delemont, Switzerland, pp 510.
Emelianova, N.A., Pravdin, F.N., Kuzina, O.S. and Lisitsyna, L.I. (1932) Biologia i ekologia Sphenoptera foveola
Gebl. v svyazi s voprosom o naplyvoobrazovanii na khondrille. In: Vtoroi sbornik po kauchukonosam (ed. Kizel,
A.R.), pp 1027. Trudy nauchno-issledovatelskikh institutov promyshlennosti, No. 502. Vsesoyuznyi nauchnoissledovatelskii institute kauchuka i guttaperchi, vypusk 6.
Izdatelstvo Narkomata tyazheloi promyshlennosti, Moskva,
(In Russian with German Summary).
Flora, SSSR (1964) Tome 29 [Asteraceae: Cichorioideae].
(eds. Bobrov, E.G. and Tsvelev, N.N.). Nauka, MoskowLeningrad, 796 pp.
Gebler, F.A. von (1825) Coleoptera Sibiriae species novae
descriptae. Hummel, Essais 4, 4257.
Hasan, S. (1978) Biology of a buprestid beetle, Sphenoptera clarescens [Col.: Buprestidae], from skeleton weed,
Chondrilla juncea. Entomophaga 23, 1923.
Iljin, M.M. (1930) Kriticheskii obzor roda Chondrilla L.
Bulleten otdela kauchukonosov Tsentralnoi nauchnoissledovatelskoi laboratorii Rezinpotreba No. 3, 161.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A World Catalogue of Agents and Their Target Weeds. Fourth Edition. CABI Publishing, Wallington,
U.K. 223p.
Kadyrbekov, R.Kh. and Tleppaeva, A.M. (2004) Faunisticheskii obzor zhukov-ksilofagov (Coleoptera, Buprestidae, Cerambycidae) Kazakhstanskoi chasti Priaralskogo
regiona. Izvestiya NAN RK. Seriya biologicheskaya I meditsinskaya 5, 3743. (In Russian with English Summary).
Kashefi, J. (2002) Report of Research [Turkey]. Rush Skeletonweed Report, USDAARS, Office of International
Research Programs, European Biological Control Laboratory, 14 pp.
Obenberger, J. (1920) Studien ber die Buprestidengattung
Sphenoptera Latr.I. Archiv fr Naturgeschichte 85 (A),
Heft 3, 101138.
Obenberger, J. (1927) Sphenopterinorum revisionis prodromus 2. De subgenere Sphenoptera Sol. s. str. (Col. Buprestidae). Revise podrodu Sphenoptera Sol. s. str. (Col.
Buprestidae). Acta Entomologica Musaei Nationalis
Pragae 5, 399.
Tleppaeva, A.M. (1999) Obzor zhukov-zlatok (Coleoptera,
Buprestidae) Almatinskogo zapovednika. Tethys Entomological Research 1, 183186 (In Russian with English
Summary).
Tleppaeva, A.M. and Ishkov, E.V. (2004) Annotirovannyi
spisok zhukov-zlatok (Coleoptera, Buprestidae) Iliiskoi
doliny. [Annotated list of buprestid beetles (Coleoptera,
Buprestidae) of Ili river valley]. Tethys Entomological Research X, 8186 (in Russian).
Volkovitsh, M.G. and Kalashian, M.J. (2006) Buprestidae:
Chrysochroinae: Sphenopterini. In: Lbl, I. and A. Smetana
(ed.) Catalogue of Palearctic Coleoptera. pp. 5356
[New Acts], 352369. Vol. 3. Apollo Books, Denmark
Stenstrup, 690 pp.
Wapshere, A.J. (1973) Selection and weed biological control
organisms. In: Proceedings of the 2nd International Symposium on Biological control of weeds. CIBC Misc. Publ.
No. 6. pp. 5662.
Wapshere, A.J. (1974) Host specificity of phytophagous organisms and the evolutionary centres of plant genera or
subgenera. Entomophaga 19, 301309.
Wapshere, A.J., Hasan, S. and Caresche, L. (1974) The ecology of Chondrilla in the Eastern Mediterranean. Journal
of Applied Ecology 11, 783799.
Wapshere, A.J., Caresche, L. and Hasan, S. (1976) The ecology of Chondrilla juncea in the Western Mediterranean.
Journal of Applied Ecology 13, 545553.
231
Common buckthorn, Rhamnus cathartica L.:
available feeding niches and the
importance of controlling this invasive
woody perennial in North America
M.V. Yoder,1 L.C. Skinner1,2 and D.W. Ragsdale1
Summary
Common buckthorn, Rhamnus cathartica L., an invasive woody perennial of northern hardwood
forests in North America, has been targeted for classical biological control, and research has been
underway since 2001. In support of biological control research, a survey was conducted for insects
associated with common buckthorn in a portion of its introduced range in the state of Minnesota.
This survey provides baseline information on available feeding niches for potential control agents of
common buckthorn and identifies the natural enemy community that could potentially interfere with
agent establishment. In 2 years of sampling, 356 species representing 111 families and 13 orders were
collected from common buckthorn in Minnesota. There was no significant defoliation observed at
any of the study sites. We surmise that ample feeding niches are available given that most herbivores
collected can be classified as generalists. However, the abundance of parasitoids and predators may
hinder establishment of potential biological control agents. Further research is needed to determine if
biotic resistance could play a significant role in preventing establishment of herbivores in a classical
biological control programme for common buckthorn in North America.
Keywords: Rhamnaceae, arthropod herbivores, natural enemies, biological control.
Introduction
Common buckthorn, Rhamnus cathartica L., is an in
vasive woody perennial that has become established in
northern hardwood forests of North America. It was
introduced as a landscape plant and used as a shelter
belt tree because of its winter hardiness and its ability
to grow in multiple soil types and habitats (Archibold
et al., 1997). In North America, common buckthorn is
one of the most invasive woody perennials in natural
ecosystems (Archibold et al., 1997; Catling, 1997).
Common buckthorn retains its leaves longer than na
tive tree species, creating a competitive advantage
(Harrington et al., 1989). In addition, Archibold et al.
(1997) suggested that common buckthorn might be al
Department of Entomology, University of Minnesota, 219 Hodson
Hall, 1980 Folwell Avenue, St. Paul, MN 55108, USA.
2
Minnesota Department of Natural Resources, 500 Lafayette Road, St.
Paul, MN 55155-4025, USA.
Corresponding author: D.W. Ragsdale <[email protected]>.
CAB International 2008
1
lelopathic, allowing its seedlings to grow below mature
female trees while inhibiting native tree species. Com
mon buckthorn produces a dense branching structure
that attracts nesting songbirds; however, the American
robin, Turdus migratorius L., experiences higher levels
of predation when nesting in common buckthorn com
pared to native species (Schmidt and Whelan, 1999).
Others have documented that common buckthorn cau
ses changes in soil properties, leaf litter composition,
and micro-arthropod communities (Heneghan et al.,
2002, 2004).
Common buckthorn has negative impacts on agri
culture. It is the spring host for oat crown rust, Puccinia
coronata Corda, which can cause severe yield losses
in oats (Harder and Chong, 1983). Common buckthorn
was identified as a suitable overwintering host for soy
bean aphid, Aphis glycines Matsumura, which was first
discovered in North America in 2000 and by 2007 has
spread to 24 states and three Canadian provinces (Rags
dale et al., 2004; Voegtlin et al., 2005).
Common buckthorn is currently on the noxious weed
list in six states and two Canadian provinces (University
232
Common buckthorn, Rhamnus cathartica L.
of Montana-Missoula, 2007; USDA, 2007). Multiple
methods of control have been employed against com
mon buckthorn including cut stump treatments, foliar
herbicide applications, and burning (Archibold et al.,
1997). Such control efforts are expensive and for the
most part are only effective on a small scale because
seedlings tend to re-grow after a burn or a chemical
treatment (Archibold et al., 1997). In the early 1960s,
several potential biological control agents were identi
fied by Malicky et al. (1970), but their study was not
continued. In 2001, biological control research was
resumed by the Minnesota Department of Natural Re
sources in collaboration with CABI Europe Switzer
land to identify and screen potential control agents.
Our first objective was therefore to conduct a survey
of herbivorous insect species associated with common
buckthorn, while our second objective was to identify
which predators and parasitoids could be found on
common buckthorn in Minnesota. These data will pro
vide key information in understanding the availabil
ity of feeding niches for potential biological control
agents and provide insights on what biotic resistance
might be present to interfere with agent establishment
in Minnesota.
Methods and materials
Field sites
In 2004 and 2005, eight common buckthorn sites
were sampled in three different habitat types, i.e., urban
(three sites), rural (two sites), and agricultural (three si
tes), in seven (2004) or six (2005) southern Minnesota
counties (see Yoder 2007 for specific locations). Sites
were characterized for their plant communities by ran
domly sampling ten 1-m2 plots. Data collected in each
plot included: percent cover of common buckthorn, per
cent cover of other plant species, common buckthorn
stem density, other plant species stem density, number
of different plant species, and percent canopy cover.
Canopy cover was estimated using a densiometer. To
characterize mature trees in the forest, which were not
captured by the 1-m2 plots, we counted the number of
trees for each species that were at least 1.5 m in height,
in a 2-m radius around each 1-m2 plot.
Insect sampling
In 2004 and 2005, 12 common buckthorn plants:
four small (<1m in height), four medium (13 m), and
four large (>3 m), were marked for repeated insect
sampling at each site. Sites were visited every 2 weeks
throughout the growing season (15 June15 September
2004; 15 May15 September in 2005). All reachable
branches were visually surveyed and any insect present
was collected, and immediately returned to the labo
ratory for either identification if adults were captured
or reared to adult stage if immature insects were col
lected. In addition to the 12 plants sampled biweekly,
two transects were established at either five (2004) or
six (2005) of the largest sites. The first transect con
sisted of 25 consecutive common buckthorn trees
growing along a path, roadway, or other opening where
trees had full exposure to the sun resulting in common
buckthorn trees that were larger. The second transect
was perpendicular to the first transect and consisted of
another 25 consecutive common buckthorn trees and
included trees growing in the under-story in shade or
filtered sunlight. All trees selected were visually sur
veyed for up to 2 min and all insects observed were
collected and returned to the laboratory.
Adults reared and collected in the field were pre
served and pinned for later identification. Soft-bodied
insects and immature insects that failed to reach the
adult stage were preserved in vials containing 70%
ethanol. Voucher specimens were deposited in the En
tomology Museum at the University of Minnesota.
All adult specimens were categorized as herbivores,
predators, parasitoids, or scavengers. For a species
to be included in the statistical analysis, a minimum
of five specimens per species was required. A quali
tative Sorenson index (Magurran, 1988) was used to
characterize differences in insect assemblages between
habitat types. The equation for the qualitative Sorenson
index (CS) is CS = 2j/(a + b) where j is the number of
species found in both groups, a is the number of spe
cies in group x, and b is the number of species in group
y. We used a quantitative Sorenson index (Magurran,
1988) to characterize differences in insect assemblages
in relation to abiotic factors such as the amount of sun
light (forest edge vs. interior) and biotic factors such as
tree size (small, medium, large). The equation for the
quantitative Sorenson index (CN) is CN = 2jN /(aN + bN)
where jN is the sum of the lower of the two abundances
recorded for a given species found in both groups, aN
is the total number of specimens in group x, and bN is
the total number of specimens in group y. The closer CS
or CN are to 1, the more similar the groups are, and the
closer to 0, the more dissimilar.
Results
Site characteristics
Overall, urban sites had the highest density of com
mon buckthorn and the lowest plant species diversity
(Table 1). Those sites characterized as agricultural sites
had the opposite, with the lowest density of common
buckthorn and the greatest plant diversity (Table 1).
Those sites characterized as rural had an intermediate
percent cover of common buckthorn, but on a stem
density per square metre had common buckthorn den
sities equal to those of the urban landscapes. Plant spe
cies diversity was low in the rural sites, but the percent
cover of other plant species and stem density of other
plant species was intermediate (Table 1). Interestingly,
233
XII International Symposium on Biological Control of Weeds
Table 1.
ite characteristics for three habitat types (urban, rural, and agricultural) surveyed for insect fauna on Rhamnus
S
cathartica, common buckthorn.
Site characteristics
Urban sites
Rural sites
Agricultural sites
61.0 0.1
39.0 0.1
6.1 1.1
11.5 2.4
3.5 0.4
82.0 1.1
48.0 0.1
52.0 0.1
6.1 1.1
28.4 3.4
3.8 0.3
83.0 0.8
31.0 0.1
69.0 0.1
4.3 0.9
26.4 3.6
6.8 0.5
80.0 0.8
6.1 1.5
1.9 0.4
1.5 0.3
2.8 0.7
1.4 0.2
1.1 0.1
1.1 0.3
2.6 0.3
1.4 0.1
All vegetation (1 m )
% Cover of common buckthorn
% Cover of other plant species
Common buckthorn stem density m2
Other plant species stem density m2
Number of other plant species
% Canopy cover
Mature tree survey (12.56 m2)a
Number of common buckthorn trees
Number of other trees
Number of other tree species
2
Trees at least 1.5 m tall in a 2-m radius from center of plot (12.56 m2).
if common buckthorn was excluded from this analysis,
the most common plant species found in either urban
or rural sites was garlic mustard, Alliaria petriolata
L., another invasive plant of hardwood forests. When
surveying the mature tree composition, four sites had
common buckthorn as the dominant mature tree. Ur
ban sites had a significantly higher density of mature
common buckthorn trees when compared to rural or
agricultural sites with one urban location having a ma
ximum of 11 mature buckthorn trees per 1 m2. Agricul
tural sites had the lowest number of mature buckthorn
trees and the highest number of other mature tree spe
cies (Table 1). For all sites, the most dominate mature
tree species, other than common buckthorn, was Ame
rican elm, Ulmus americana L., followed by box elder,
Acer negundo L.
Insect fauna
Over the 2-year study, a total of 1733 arthropods
representing 13 orders, 111 families and 356 species
were collected from common buckthorn. Hemiptera
was the most abundant order, followed by Hymenop
tera, which consisted mostly of parasitoids (Tables 2
and 3). Several species were abundant, each with over
75 specimens collected: Metcalfa pruinosa (Say) (Fla
tidae), Lasius alienus (Frster) (Formicidae), Harmonia axyridis (Pallas) (Coccinellidae), Graphocephala
coccinea (Forster) (Cicadellidae), and Trissolcus sp. a.
(Scelionidae).
For the analysis we used 606 herbivores represent
ing 32 different species, 154 predators representing five
different species, and 140 parasitoids representing four
different species (Tables 2 and 3). An additional 314
species were excluded from analysis because fewer
than five specimens were collected over the 2-year
sampling effort or because species were known to be
saprophagous, mycetophagous, scavengers, or nonfeeding as adults. The Sorenson index (CS) showed
that all three habitat types, agriculture, rural, and urban
landscapes, were very similar in insect species diver
sity (range 0.710.73). The majority of predators were
captured at sites in agriculture habitats (56%) and the
majority of parasitoids were captured in rural habitats
(61%). The quantitative Sorenson index for insect di
versity for forest edge and interior using data collected
from the two perpendicular transects was (CN = 0.54). It
is not surprising that transects are different since more
insects were collected on common buckthorn along the
transect where plants were along a forest edge (62% of
captures) compared to the interior (38% of captures).
When comparing tree sizes, large and small trees had
the least similar insect composition (CN = 0.44); whereas
medium and small trees had the most similar insect
composition (CN = 0.59). Medium trees tended to have
higher diversity and abundance compared to large and
small trees.
In general, there was very little evidence of feed
ing damage on common buckthorn. The most common
type of damage was leaf miner tunnels, followed by
damage caused by lepidopteran larvae. Nine species
were reared in the laboratory after collecting immature
insects from common buckthorn indicating these nine
species are able to complete their development solely
on buckthorn. These included three hemipteran spe
cies, Acanalonia conica (Say), M. pruinosa, and Gyponana quebecensis (Provancher), three orthopterans,
Neoxabea bipunctata (De Geer), Oecanthus fultoni
Walker, and Oecanthus niveus (De Geer), and three
lepidopterans collected as eggs and reared to adult,
which included Choristoneura rosaceana (Harris),
Machimia tentoriferella Clemens, and Spilosoma virginica (Fabricius). The two tortricids, C. rosaceana and
M. tentoriferalla experienced high mortality during
rearing and adult specimens that did emerge often had
abnormal wing development. However, a literature
search revealed that these nine species listed above
can be categorized as generalist herbivores and are not
234
Common buckthorn, Rhamnus cathartica L.
Table 2.
Herbivores collected on Rhamnus cathartica, common buckthorn in Minnesota. Only species for which a mini
mum of five specimens were collected were included, except for Oecanthus spp., because of the high abundance
of immature specimens collected.
Order
Family
Genus species
Orthoptera
Gryllidae
Neoxabea bipunctata (De Geer)
Oecanthus fultoni Walker
Oecanthus niveus (De Geer)
Acanaloniidae
Aphididae
Pentatomidae
Tingidae
Acanalonia conica (Say)
Aphis glycines / nasturtii
Aphis glycines Matsumura
Aphis nasturtii Kaltenbach
Clastoptera obtusa (Say)
Philaenus spumarius (L.)
Empoasca sp. b
Graphocephala coccinea (Forster)
Gyponana quebecensis (Provancher)
Jikradia olitorius (Say)
Cedusa incisa (Metcalf)
Metcalfa pruinosa (Say)
Hyaliodes harti Knight
Hyaliodes vitripennis (Say)
Paraproba capitata (Van Duzee)
Phytocoris spicatus Knight
Euschistus tristigmus (Say)
Corythucha pergandei Heidemann
Chrysomelidae
Curculionidae
Pyrochroidae
Diabrotica longicornis (Say)
Polydrusus sericeus (Schaller)
Pedilus impressus (Say)
Arctiidae
Gracillariidae
Psychidae
Tortricidae
Spilosoma virginica (Fabricius)
Phyllonorycter caryaealbella (Chambers)
Thyridopteryx ephemeraeformis (Haworth)
Choristoneura rosaceana (Harris)
Machimia tentoriferella Clemens
Cecidomyiidae
Cynipidae
Parwinnertzia notmani Felt
Diplopepsis sp. a
Liodora sp.
Fenusa sp.
Subtotal
Hemiptera
Cercopidae
Cicadellidae
Derbidae
Flatidae
Miridae
Subtotal
Coleoptera
Subtotal
Lepidoptera
Subtotal
Diptera
Hymenoptera
Tenthredinidae
Subtotal
Table 3.
Number of specimens
2005
5
1
2
8
11
39
17
26
7
8
5
64
9
6
2
164
6
2
8
4
9
5
392
5
5
4
14
10
6
5
5
3
29
5
9
6
19
34
Total
15
3
3
21
20
39
24
26
7
10
5
85
13
15
10
175
13
5
14
6
17
5
489
6
7
5
18
10
8
5
9
5
37
7
9
6
19
34
Predators and parasitoids collected on Rhamnus cathartica, common buckthorn in Minnesota. Only species for
which a minimum of five specimens were collected were included.
Order
Family
Genus species
Hemiptera
Coleoptera
Nabidae
Cantharidae
Coccinellidae
Lasiomerus annulatus (Reuter)
Podabrus rugulosus LeConte
Coleomegilla maculata DeGeer
Harmonia axyridis (Pallas)
Empididae
Platygasteridae
Scelionidae
Tachypeza sp. a
Leptacis sp. c
Idris sp.
Trissolcus sp. a
Trissolcus sp. b
Subtotal
Diptera
Hymenoptera
Subtotal
2004
10
2
1
13
9
0
7
0
0
2
0
21
4
9
8
11
7
3
6
2
8
0
97
1
2
1
4
0
2
0
4
2
8
2
0
0
0
0
Number of specimens
235
2004
10
5
3
44
52
4
0
18
37
0
55
2005
15
0
2
68
70
3
8
0
39
38
85
Total
25
5
5
112
122
7
8
18
76
38
140
XII International Symposium on Biological Control of Weeds
considered specialist herbivores that only feed on com
mon buckthorn.
Discussion
Urban sites, which had the densest common buckthorn
infestation and lowest plant species diversity, also had
the lowest insect abundance when compared to other
habitat types. All urban sites sampled were located in
highly populated areas where human activities could
easily disturb the natural habitat. In contrast, agricul
tural sites had the highest plant diversity with more in
sects collected at those sites. Predators were collected
at higher rates in agricultural sites than the other sites
possibly drawn there by agricultural pests that would
be found in the adjacent crop fields.
The main objective of this study was to identify ma
jor herbivores present on common buckthorn in Min
nesota. Overall, there were many herbivores collected,
however; most insects collected were represented by
fewer than five specimens suggesting that they were
transient feeders or generalist herbivores that do not
utilize common buckthorn. In reports of herbivores col
lected from R. cathartica in Europe, the most common
insect species found were Lepidopterans (Malicky et al.
1970). Here we show that in Minnesota, defoliators
were common, but unlike the situation in Europe, more
Hemipterans were encountered in Minnesota than Le
pidopterans. During our 2-year study we did not find
any insect feeding internally on buckthorn, and thus
one potential niche that could be exploited successfully
would be an internal feeder such as the stem-boring
beetle, Oberea pedemontana Chevrolat (Coleoptera:
Cerambycidae) which has been identified in Europe
Figure 1.
as a possible biological control agent of R. cathartica
(Gassmann, 2005). Even though we found many gene
ralist herbivores feeding on leaves at no time did de
foliation exceed 5% on any one tree, thus a specialist
herbivore would have an abundant resource to utilize
in Minnesota.
The second objective of this study was to identify
possible sources of biotic resistance if non-native her
bivores were introduced as classical biological control
for common buckthorn. There were numerous parasit
oids and predators, all considered generalists, collected
from common buckthorn. The abundance of parasitoids
and predators may indeed hinder establishment of po
tential biological control agents. Generalist predators
have been known to interfere with biological control
agents released for purple loosestrife control (Sebolt
and Landis, 2004). Currently, there have been a few
species proposed as potential biological control agents
for common buckthorn in North America (Gassmann,
2005). As agent selection continues for common buck
thorn, the species diversity and abundance of natural
enemies collected from buckthorn and documented
here should be considered. In particular, H. axyridis
could play a significant role in preventing establishment
of herbivores since it was the most abundant generalist
predator collected and this coccinellid is known to pre
fer arboreal habitats. This recently introduced cocci
nellid was more common in spring and fall (Figure 1).
Harmonia axyridis could exert strong biotic resistance
on biological control agents especially if a vulnerable
life stage was present when H. axyridis densities were
high. For example, one group of candidate biological
control agents for common buckthorn are the psyllids,
Cacopsylla rhamnicolla and Trichohermes walkeri
Seasonal abundance of Harmonia axyridis observed on Rhamnus cathartica, common buckthorn in Minnesota. Data pooled for 2004 and
2005.
236
Common buckthorn, Rhamnus cathartica L.
(Gassmann 2005) and it is possible that Harmonia axyridis would pose a particular threat to psyllids. This po
tential negative interaction could be studied as part of
the host testing procedure.
Acknowledgements
We would like to thank Dr John Luhman, Dr Leonard
Ferrington, and Gregory Setliff for help on identifica
tions. In addition, we would like to thank all of the
undergraduate researchers for help in the field and la
boratory. This research was funded by the Minnesota
Department of Natural Resources based on funds
appropriated by the Minnesota Legislature as recom
mended by the Legislative Commission on Minnesota
Resources.
References
Archibold, O.W., Brooks, D. and Delanoy, L. (1997) An in
vestigation of the invasive shrub European buckthorn,
Rhamnus cathartica L., near Saskatoon, Saskatchewan.
Canadian Field-Naturalist 111, 617621.
Catling, P.M. (1997) The problem of invading alien trees
and shrubs: some observations in Ontario and a Canadian
checklist. Canadian Field-Naturalist 111, 338342.
Gassmann, A. (2005) Developing biological control of buck
thorns. In: Skinner, L. (ed.) Proceedings of the Symposium
on the Biology,Eecology, and Management of Garlic
Mustard (Alliaria petiolata) and European buckthorn
(Rhamnus cathartica). USDA Forest Service Publication,
FHTET-2005-09, pp. 5557.
Harder, D.E. and Chong, J. (1983) Virulence and distribution
of Puccinia coronata in Canada in 1982. Canadian Journal of Plant Pathology 5, 185188.
Harrington, R.A., Brown, B.J. and Reich, P.B. (1989) Eco
physiology of exotic and native shrubs in southern Wis
consin; I. Relationship of leaf characteristics, resource
availability, and phenology to seasonal patterns of carbon
gain. Oecologia 80, 356367.
Heneghan, L., Clay, C. and Brundage, C. (2002) Rapid de
composition of buckthorn litter may change soil nutrient
levels. Ecological Restoration 20, 108111.
Heneghan, L., Rauschenberg, C., Fatemi, F. and Workman,
M. (2004) European buckthorn (Rhamnus cathartica) and
its effects on some ecosystem properties in an urban wood
land. Ecological Restoration 22, 275280.
Magurran, A.E. (1988) Ecological Diversity and Its Measurements. Princeton University Press, Princeton, NJ, 95 pp.
Malicky, H., Sobhian, R. and Zwlfer, H. (1970) Investiga
tions on the possibilities of a biological control a biolo
gical control of Rhamnus cathartica L. in Canada: host
ranges, feeding sites, and phenology of insects associated
with European Rhamnaceae. Zeitschrift fur angewandte
Entomologie 65, 7797.
Ragsdale, D.W., Voegtlin, D.J. and ONeil, R.J. (2004) Soy
bean aphid biology in North America. Annals of the Entomological Society of America 97, 204208.
Schmidt, K.A. and Whelan, C.J. (1999) Effects of exotic Lonicera and Rhamnus on songbird nest predation. Conservation Biology 13, 15021506.
Sebolt, D.C. and Landis, D.A. (2004). Arthropod predators
of Galerucella calmariensis L. (Coleoptera: Chrysomeli
dae): an assessment of biotic interference. Environmental
Entomology 33, 356361.
University of Montana-Missoula. (2007) Invaders Database
System. Available at: invader.dbs.umt.edu.
USDA (2007) Plants Database. Available at: www.plants.
usda.gov.
Voegtlin, D.J., ONeil, R.J., Graves, W.R., Lagos, D. and
Yoo, H.S. (2005) Potential winter hosts of soybean aphid.
Annals of the Entomological Society of America 98, 690
693.
Yoder, M.V. (2007). Post-release monitoring of two classi
cal biological control agents, Galerucella calmariensis
(L.) and G. pusilla (Duftschmidt), on purple loosestrife,
Lythrum salicaria L. M.Sc. Thesis, University of Minne
sota, 155 p.
237
Evaluation of Fusarium as potential
biological control against Orobanche
on Faba bean in Tunisia
M. Zouaoui Boutiti,1 T. Souissi1 and M. Kharrat2
Summary
A total of 149 fungal strains identified as Fusarium were isolated from infected Orobanche crenata
Forsk. and Orobanche foetida Poir. plants. Their pathogenicity and virulence were assessed in Petri
dish assays using lentils as the medium. Ten isolates were found to reduce the number of tubercles attached to the host plant. Among them, two isolates that caused necroses on tubercles of Orobanche in
the Petri dish assays were identified as Fusarium F6 and F10. They reduced the number of tubercles
of O. crenata by 97% and 98%, respectively. Inoculums of F6 and F10 were produced on barley
grains and were tested in sterilized and non-sterilized soil in separate pot experiments, using O. crenata and O. foetida as parasitic plants. Both isolates reduced the number of O. crenata and O. foetida
by 68% and 88%, respectively, and their dry matter by 82% to 88%. A similar experiment conducted
using formulated inoculums of the two isolates showed that the formulation improved the efficiency
of the fungi, and reductions in the number and dry matter of tubercles to 100% were observed. These
results suggest that Fusarium isolates have the potential to be used as biological control agents against
O. crenata and O. foetida on faba bean in Tunisia.
Keywords: pathogens, pathogenicity, virulence, parasitic plants, broomrapes.
Introduction
Broomrapes, Orobanche spp., of the family of Orobanchaceae are troublesome root parasitic weeds
that cause severe damage to vegetables, legumes and
sunflower (Parker and Riches, 1992). Approximately,
16 million hectares of arable land in the Mediterranean region as well as in west Asia are currently endangered by Orobanche infestation (Sauerborn, 1994). In
Tunisia, Orobanche crenata Forsk. distributed in the
north-east and O. foetida Poir. in the north-west, are
the main species that cause losses in leguminous crops,
especially on faba bean (Kharrat and Halila, 1994).
Losses in faba bean fields can reach 80% (Kharrat,
2002). Difficulties in controlling Orobanche are due to
National Agronomic Institute of Tunisia, 43 Avenue Charles Nicolle,
1082 Tunis-Mahragne, Tunisia <[email protected], tsouissi@
alinto.com>.
2
National Agricultural Research Institute of Tunisia, Rue Hdi Karray,
2080 Ariana, Tunisia <[email protected]>.
Corresponding author: M. Zouaoui Boutiti <[email protected]>.
CAB International 2008
1
the numerous tiny seeds that retain their viability in the
soil for 6 to 20 years. Germination of Orobanche seed
requires a stimulant excreted by the host plant and produces germ tubes that attach to the host plant (Raynal
et al., 1989). The germ tube develops a haustorium and
forms a tubercle. The haustorium represents the physical and morphological contact between the parasite
and the host. It supplies the parasite with water, mineral
nutrients and organic materials from its host (Kroschel,
2001).
So far, no efficient control measures for Orobanche
spp. have become available to farmers (Mller-Stver,
2001). Single methods such as delayed sowing and
use of resistant varieties have shown unsatisfactory results. The use of chemical products such as glyphosate
requires care to avoid phytotoxicity. Thus, Orobanche
represents a difficult target for selective chemical control. Control of Orobanche may be possible by integrating control measures. The integration of biological control with other Orobanche management methods is of
increasing research interest. Several investigators have
reported the use of fungi as potential biological agents
against Orobanche (Waston and Waymore, 1992).
238
Evaluation of Fusarium as potential biological control against Orobanche on Faba bean in Tunisia
Research has been conducted in several countries in
cluding Algeria, Egypt, Germany, Maroc and Chili
(Klein et al., 1999; Zermane et al., 1999; Mller-Stver,
2001; Boari and Vurro, 2004). Fusarium oxysporum f.
sp. orthoceras (Appel and Wollenw.) Bilai obtained
from diseased O. cumana tested in soil with sunflower
as a host plant was able to reduce the number of attached
and emerged broomrape seedlings by about 90% (Bedi
and Donchev, 1991). F. oxysporum f. sp. orthoceras on
O. cumana was also tested by Thomas et al. (1998).
Recently, F. arthroporioides and F. oxysporum isolated
in Israel from O. aegyptiaca were shown to be effective
in reducing broomrape growth (Amsellem et al., 2001).
The pathogenicity of two isolates, Ulocladium botrytis
Preuss and F. oxysporum Schlecht. f. sp. Orthoceras,
were tested by Mller-Stver (2001). The two fungi cause
necroses on both O. cumana and O. crenata. Currently,
the development of an appropriate formulation which allows successful application of fungal propagules
will determine the success of Fusarium in agriculture
applications. The encapsulation of fungal propagules
in a solid matrix Pesta was used by (Mller-Stver,
2001). A 70% reduction of Orobanche emergence was
obtained when wheat flour kaolin granules containing
chlamydospore rich biomass was applied.
Considering the importance of O. crenata and
O. foetida in Tunisia and the lack of research on fungi
associated to Orobanche spp., the main objective of this
study was to screen and evaluate the potential of fungi
isolated from Orobanche with potential as biological
control agents against the parasitic weed, in laboratory
and green house experiments.
Materials and methods
Field surveys
Field surveys were carried out from April 2004 to
May 2005 in northern Tunisia, especially in the region
of Nabeul. Underground stages of Orobanche with
symptoms of fungal infections such as browning and
rotting were collected. The plants were conserved in
laboratory until use.
Isolation
Isolations were made from pieces of tubercles and
stems with fungal symptoms. Diseased tissues were excised, washed with distilled water, sterilized in 1% sodium hypochlorite with Tween 20 for 5 min and rinsed
four to five times with sterile distilled water. After drying on filter paper, pieces were placed in Petri dishes on
potato dextrose agar (PDA) medium supplemented
with 100 ppm of streptomycin. The Petri dishes were
incubated in the dark at 22C until fungal development
occurred. Repeated sub-culturing was done to obtain
pure cultures. Isolates were conserved on special nutrient poor agar (SNA) at 5C for short term storage and
in liquid nitrogen for long-term storage.
Bioassays
The isolated fungi were evaluated to assess their
phytotoxic ability on the growth of the underground
stages of Orobanche. In these Petri dishes bioassays,
seeds of O. crenata were used as the parasitic weed
and those of lentil were used as the host plant. The methods followed those of Kroschel (2001). Plastic Petri
dishes were filled with washed sterile sand, watered
and covered with filter paper. Orobanche crenata seeds
were sterilized with sodium hypochlorite, rinsed with
distilled water and sprinkled on the filter paper at the
densities of 25 seeds per square metre. The Petri dishes
were covered with black plastic and incubated in the
dark for conditioning at 22C for 10 days. To enhance
pre-conditioning in the Orobanche seed, 100 ppm of
gibberellic acid was added. Pre-germinated lentil seed
lings were inserted into sand in the Petri dishes through
holes made in the surface of the filter paper.
To test the pathogenicity of fungal isolates, fresh col
onies of the isolated fungi growing on SNA medium
were used. For each Petri dish, the black plastic was
removed and filter paper containing O. crenata seeds
and lentil seedlings were sprayed with 10 ml of the
spore suspension at 106 spores per millilitre. The Petri dishes were incubated in the green house at 25C
and 16 h/18 h photoperiod for 5 weeks. Two replicates
were used per treatment. The number of germinated,
attached seeds and the number of tubercles formation
were recorded.
Test of specificity
The isolates which reduced the number of tubercles of Orobanche in Petri-dish assays were selected
and were tested for their host specificity. A range of
plants was used that included tomato (Lycopersicon
esculentum Mill), carrot (Daucus carota L), Faba
bean (Vicia faba L. and Vicia faba L. minor), pea (Pisum sativum L.), chickpea (Cicer arietinum L.), lentil
(Lens culinaris L), wheat (Triticum aestivum L.) and
barley (Hodeum vulgare L.) These plants were grown in pots in a greenhouse. Ten days after emergence
roots were washed with distilled water and immersed
in the inoculums at the concentration of 106 spores per
millilitre, for 5 to 10 min. Then plants were transplanted in pots and observed weekly for 1 month for the
development of symptoms. Four plants of the same
test plant species were used per pot with four pots per
tested plant.
Pot assays
Selected fungal isolates that were able to reduce
Orobanche seed germination in the Petri dish assays
were tested in sterilized and non-sterilized soils in pots
in a greenhouse. Both O. crenata and O. foetida were
used as parasitic weeds and V. faba as the host plant
during this experiment.
239
XII International Symposium on Biological Control of Weeds
Preparation of inoculum: Barely grains were used as
a solid inoculum. Ten ml of sterile water were added to
fresh colonies of the fungi growing on SNA medium.
The resulting suspension, containing spores and mycelium, was used to inoculate the organic substrate which
was then incubated for 14 days at 25C.
Inoculation and pot trials: Two pot experiments, using sterilized and non sterilized soils, were conducted
under greenhouse conditions to study the ability of
fungi to control O. crenata and O. foetida in V. faba
plants. In both experiments, plastic pots with 750 g
capacity were filled with a mixture of soil and sand
at a ratio of 2:1. Orobanche seeds were sprinkled
onto the soil surface. The inoculum was added and
mixed into the soil together with the seeds. Each pot
was provided with either 11 mg of O. crenata seed or
9 mg of O. foetida seed, or 7.5 g of the solid inoculum.
Three seeds of V. Faba minor were planted per pot
and thinned afterwards to obtain one host plant per
pot. The pots were fertilized and watered each week.
Five controls were prepared: the host plant alone
(H), the host plant with O. crenata or host plant with
O. foetida, the host plant with non-inoculated substrate (H + NIS) and host plant with inoculated substrate (H + NIS). The experiments continued for
5 months and terminated when the host plant in the
control was dead. The parameters used to assess the
effect of fungi on the control of O. crenata and O.
foetida were faba bean height (cm) and dry matter
weight (g) and Orobanche number and dry matter
weight (g).
Formulation in pot experiments
Wheat flour kaolin granules were prepared after the
methods of Connik et al. (1991, 1996). Durum wheat
flour 32 g, kaolin 6 g and sucrose 2 g, were blended
and poured into a dish. Inoculums were added as chlamydospore in 23 ml PDB (potato dextrose broth). The
Table 1.
Isolates
Control
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
LSD(5%)
mixture was kneaded with gloved hands and passed
through a small, hand-operated pasta maker. Obtained granules were incorporated in sterilized and nonsterilized soils. The granules were added and mixed
into the soil together with the seeds. Each pot was filled
with either 11 mg of O. crenata or 9 mg of O. foetida
and 1 g of granules per kg of soil. Three seeds of V. faba
minor were planted per pot and thinned afterwards to
obtain one host plant per pot. In total, five pot experiments were carried out.
Statistical analysis
All pots experiments were conducted in totally randomized design. Statistical analyses were performed
using analysis of variance (ANOVA) with alpha 0.05 in
GEN-STAT software.
Results
Field survey and isolation of fungi
One hundred and forty nine isolates were obtained
from infected O. crenata plants collected during field
surveys. All isolates were found to belong to the genus
Fusarium after microscope examination.
Bioassays
Ten isolates of the genus Fusarium reduced the germination of O. crenata 27% to 93% and eight reduced
attachment (22% to 79%) to the host plant by germinated
Orobanche (Table 1). The two exceptions, F4 and F7,
did not differ statistically between the treatments and
the non-inoculated control. The number of tubercles
developed by O. crenata was reduced by 78% to 98%
(Table 1). Symptoms of necroses were observed on
Orobanche inoculated tubercles.
Isolates F6 and F10 were the most efficient in reducing the percentage of tubercles of O. crenata by
ffect of a conidial suspension of Fusarium on the development of the underground stages of Orobanche crenata
E
in Petri dishes including; the number of O. crenata seeds that germinated, the number of germinating seeds that
attached to the host and the number of tubercles formed.
Number
germinated
Percent reduction
in germination
Number
attached
Percent reduction
in attachments
Number of
tubercles
Percent reduction
in tubercles
36.5
7.5
3.5
5
3
5
26.5
3.5
4
2.5
23
5.74
79
90
86
92
86
27
90
89
93
37
27
5.5
18
10.5
29
9
13.5
24.5
13
13.5
21.5
4.03
79
33
61
67
50
9
52
50
22
47
7.5
10
2
7.5
5
1
4.5
4
7
1.5
3.5
84
78
95
84
89
98
90
91
85
97
240
Evaluation of Fusarium as potential biological control against Orobanche on Faba bean in Tunisia
97% and 98% respectively. These isolates were used
in sterilized and non sterilized soil on O. crenata and
O. foetida using faba bean as a host plant.
Test of specificity
Isolates F6 and F10 were selected for use in specificity test because in preliminary tests no symptoms
and no death were observed on test plant species.
Pot assays
Sterilized soil: F6 and F10 tested in sterilized soil on
O. crenata and O. foetida reduced the number germinated and dry matter weights of both parasitic plants.
In O. crenata the number of tubercles was reduced by
70% to 87% compared to the infested the controls (H +
O + NIS), and the dry matter of tubercles was reduced
by 88% (Table 2). Inoculation with F6 and F10 isolates resulted in 36% to 38% increase in height of faba
bean compared to the Orobanche infested control (H +
IS). The dry weight of the host plant was also significantly increased by 120% to 129% compared to the
Orobanche-infected control (Table 2).
Isolates F6 and F10 reduced the number of O. foetida
by 68% to 77% whereas the dry matter was reduced
by 81% to 84% compared to the infested control (H +
O + NIS) (Table 2). The height and the dry matter of
Table 2.
Faba bean was also increased by 35% to 45% and 79%
to 82%, respectively, compared to the infested control
(H + IS).
Non-sterilized soils: Isolates F6 and F10 incorporated
with barley grains as inoculum substrate into the soil
reduced the number of both O. crenata and O. foetida
by more than 90% compared to the infested Orobanche
control (Table 3). So, the dry matter of both O. crenata
and O. foetida was increased by 100%, which did not
statistically differ from both control of O. crenata and
O. foetida (H + O + NIS) (Table 3).
There were no significant effects on the height or
dry matter of faba bean compared to either Orobanche
species (Table 3). High significantly, no tubercles were
produced by either species.
Formulation experiments in pots
There were no tubercles produced and therefore no
tubercle dry matter for O. crenata and O. foetida when
wheat flour kaolin granules containing chlamydospore
rich biomass was applied in sterilized or non sterilized
soils (Figure 1).
Discussion
The use of Petri dishes allowed observation of the underground stages of Orobanche (germination, attachments
and tubercles) which would not have been possible
ffect of isolates F6 and F10 on the number of tubercles and tubercle dry weight of Orobanche crenata and
E
O. foetida and on the height and dry weight of faba bean, in sterilized soil.
Treatments
Plant height
(cm)
Percent
increase
Plant dry
matter (g)
Percent
increase
No. of
tubercle
Vica faba minor, faba bean
Control (H)
Control (H + IS)
Control (H + F6)
Control (H + F10)
Control (H + O)
Control (H + O +
NIS)
(F6)
(F10)
LSD (5%)
64.8
65.4
89.4
89
51.4
53.8
Control (H)
Control (H + IS)
Control (H + F6)
Control (H + F10)
Control (H + O)
Control (H + O +
NIS)
(F6)
(F10)
LSD (5%)
64.8
65.4
89.4
89
50.2
54.2
37
36
73.4
78.8
11.33
35
45
73
74.2
11.33
37
36
Percent
reduction
Tubercles dry
matter (g)
Percent
reduction
Orobanche crenata
5.39
5.65
9.41
8.6
2.66
2.77
66
52
5.2
4.8
120
129
0.6
1.4
2.18
87
0.23
70
0.25
1.14
Orobanche foetida
88
88
5.39
5.65
9.41
8.6
3.32
3.41
66
52
5.6
4.4
2.12
2.28
6.08
6.18
1.81
79
82
1.4
1
2.18
68
77
0.43
0.36
1.14
81
84
36
6.11
38
6.35
1.81
Vica faba minor, faba bean
2.02
1.93
H Faba bean only; H + NIS faba bean plus non-inoculated barley grains; H + O faba bean plus Orobanche; H + O + F Faba bean plus noninoculated barley grains plus Orobanche.
241
XII International Symposium on Biological Control of Weeds
Table 3.
ffect of isolates F6 and F10 on the number of tubercles and tubercle dry weight of Orobanche crenata and
E
Orobanche foetida and on the height and dry weight of Faba bean, in non sterilized soil.
Treatments
Plant height Percent
(cm)
increase
Plant dry
weight (g)
Percent
increase
No. of
tubercle
Vica faba minor, faba bean
Control (H)
Control (H + IS)
Control (H + F6)
Control (H + F10)
Control (H + O)
Control (H + O +
NIS)
(F6)
(F10)
LSD (5%)
56.4
49.8
64
51.8
39.8
37.6
Control (H)
Control (H + IS)
Control (H + F6)
Control (H + F10)
Control (H + O)
Control (H + O +
NIS)
(F6)
(F10)
LSD (5%)
56.4
49.8
64
51.8
38.6
30.4
28
4
48.4
48.6
NS
59
59
45.6
45.4
NS
28
4
Percent
reduction
Tubercles dry
weight (g)
Percent
reduction
Orobanche crenata
1.9
1.8
1
1.2
0.8
0.9
21
1.9
20
1.6
NS
Vica faba minor, faba bean
2
2,2
0.025
0.025
100
0
100
0
NS
Orobanche foetida
111
77
0
0
0.59
100
100
1.9
1.8
1
1.2
0.8
0.6
1.8
2.6
0.025
0.025
1.6
1
NS
166
66
0
0
0.59
100
100
0
0
NS
100
100
H Faba bean only; H + NIS faba bean plus non-inoculated barley grains; H + O faba bean plus Orobanche; H + O + F faba bean plus non
inoculated barley grains plus Orobanche.
3
2,5
2
1,5
1
Figure 1.
H+OF+F10
H+OF+F6
H+OF+ST
H+OF
H+OC+F10
H+OC+F6
H+OC+ST
0,5
H+OC
Nomber of Orobanche
tubercles
3,5
Pesta formulation effect on the number of Orobanche crenata and O. foetida tubercles.
during a pot experiment or under field conditions. In
Petri dishes the fungus reduced the germination of O.
crenata by 27% to 93% compared to the control. Thomas et al., (1999), found that Fusarium oxysporum f.
sp. ortoceras colonized seeds of O. cumana, so the germination of inoculated seed was significantly reduced.
Zermane et al. (1999) found that the germination of
O. crenata was reduced by 50.3% when they used
F. oxysporum in root chamber assays.
The percentage of attachment of O. crenata to the
lentil host plant was also reduced by 79% as a result of
fungus inoculum. Bedi and Donchev (1991) suggested
that the black pigment of seeds protects the seed from
fungal attack and they believed that the infection by the
fungus occurred after seed germination. Consequently,
the number of tubercles formation was significantly
reduced by 98% in Petri dishes assays. Mller-Stver
(2001) found that the mortality of O. aegyptiaca tu-
242
Evaluation of Fusarium as potential biological control against Orobanche on Faba bean in Tunisia
bercles was significantly increased after inoculation
in a root chamber. Accordingly, the same phenomena
were also observed by Bouzoukov and Kouzmanova
(1994). Thomas et al. (1998) suggested that F. oxysporum inoculated in a root chamber, attacked germination
and tubercles formation. Cohen et al. (2002) explained
the mortality process of tubercles; they suggested that
the hyphen penetrated the outer cells layer within 24 h,
reaching the centre of the tubercles by 48 h and infected
nearly all cells by 72 h. Most of the infected tubercles
had died by 96 h. We observed necroses on inoculated
tubercles and the same were observed by Linke et al.
(1992).
Of the ten isolates tested, two F6 and F10 were the
most effective in reducing the percentage of tubercles
in Petri dishes. Tests in with these two isolates in sterilized and non-sterilized soil showed that they could
significantly reduce the number and dry weight of O.
crenata and O. foetida. Similarly, Sauerborn et al. (1994)
found a reduction in tubercle number of O. cumana parasitizing sunflower. Mller-Stver (2001) observed a
decrease of the total O. cumana dry matter per pot as
a consequence of the application of fungi. The same
phenomena were also observed by Thomas et al. (1998)
for O. cumana and F. oxysporum f. sp. orthoceras.
Faba bean height and dry weight increased when
F6 and F10 were used on barley grains as inoculum
substrate. Zonno and Vurro (2002) using F. oxysporum
and F. solani on O. ramosa, with tomato as the host
plant, suggested that both isolates permitted growth of
a larger and healthier tomato root system compared to
their controls. The same reduction in numbers and dry
weights of O. crenata and O. foetida were observed by
us in non-sterilized soil.
These results suggested that Fusarium was able to
grow and compete successfully with other microorgan
isms present in the soil (Abbasher et al., 1996). The use
of Fusarium to control O. crenata and O. foetida in soil
has not previously been considered for biological control. The potential of mycoherbicides for use against
the parasitic plants has been investigated (Garcia Garza
et al., 1998) and our studies indicate that this may be
possible using Fusarium to control Orobanche species.
Future research will be done to identify isolate F6 and
F10 using the morphological and molecular technique.
References
Abbasher, A.A., Sauerborn, J. and Kroschel, J. (1996) Evaluation of Fusarium semitectum var. majus for control of
Striga hemonthica. In: Moran, V.C. and Hoffman, J.H.
(eds) Proceeding of the IX International Symposium on
Biological Control of Weeds. University of Cape Town,
Stellenbosch, South Africa, pp. 115120.
Amsellem, Z., Kleifed, Y., Kereny, Z., Hornok, L., Goldwas
ser, Y. and Gressel, J. (2001) Isolation, identification
and activity of mycoherbicidal pathogens from juvenile
broomrapes plants. Biological Control 21, 274284.
Bedi J.S. and Donchev, N. (1991) Results on mycoherbicide
control of sunflower broomrape (Orobanche cumana)
under glass house and field conditions. In: Fifth International Symposium on Parasitic Weeds. Nairobi, Kenya,
pp. 7682.
Boari, A. and Vurro. M. (2004) Evaluation of Fusarium spp.
and other fungi as biological control agents of broomrapes
(Orobanche ramose). Biological Control 30, 212219.
Bouzoukov, H. and Kouzmanova. I. (1994) Biological control
of tobacco broomrape (Orobanche spp) by means of some
fungi of the genus Fusarium. Biology and management of
Orobanche. In: Pieterse A.H., Verkleig, j.a.c. and Borg,
s. j. Ter. (eds) Proceedings of the Third International
Workshop on Orobanche and Related Striga Research.
Royal Tropical Institute, Amsterdam, The Netherlands,
pp. 532538.
Cohen, B.A., Amsellem, Z., Simcha, L.Y. and Gressel, J.
(2002) Infection of tubercles of parasitic weed Orobanche
aegyptiaca by mycoherbicidal Fusarium species. Annals
of Botany 90, 567578.
Connik, W.J., Boyette, C.D. and McAlpine, J.R. (1991) Formulation of Mycoherbicides using a pasta-like process.
Biological Control 1, 128287.
Connik, W.J., Daigle, D.J., Boyette, C.D. Williams, K.S.,
Vinyard, O.T. and Quimby, P.C. (1996) Water activity and
other factors that affect the viability of Colletotrichum
truncatum conidia in wheat flour-kaolin granules (Pesta). Biological Science and Technology 6, 277284
Garcia Garza, J.A., Fravel, D.R., Bailey, B.A. and Hebbar, P.K. (1998) Dispersal of Fusarium oxysporum f.sp
erythroxyli and F. oxysporum f .sp melonis by ants. Biological Control 88, 158190.
Kharrat, M. and Hallila, M.H. (1994) Orobanche species
on Faba bean (Vicia faba L) in Tunisia; problems and
management. Biology and management of Orobanche.
In: Pieterse A.H., Verkleig, j.a.c. and Borg, s. j. Ter.
(eds) Proceedings of the third international workshop on
Orobanche and related Striga research. Royal Tropical
Institute, Amsterdam, The Netherlands, pp. 638644.
Kharrat, m. (2002) etude de la virulence de lcotype de Bja
de lO. foetida sur les diffrentes espces de lgumineuses. Le devenir des lgumineuses alimentaires dans le Magreb. In: Kharrat, M., Andaloussi, A., Maatoughi, M.E.H.,
Sadiki, M. and Bertenbreiter, W. (eds) Proceedings du 2me
sminaire du rseau REMAFEVE/REMALA. Hammamet,
Tunisie, pp. 8896.
Klein O., Kroschel, J. and Sauerborn, J. (1999) Efficacit
des Lchs supplmentaires de Pytomyza Orobanchia
Kalt (DIPTERA/Agromizidae) pour la lutte biologique
contre lOrobanche au Maroc. In: Abderabihi, J.M. and
Betz, H. (eds) Advances in Parasitic Weed at OnFarm
Level. Kroschel, Vol II. Joint Action to Control Orobanche
in Wana Rgion. Margraf, Verlag, Weikerheim, Germany,
pp. 161171.
Kroschel J. (2001) A technical manual of Parasitic weeds.
Technische Zusammenarbreit. GTZ, University of Hohenhein, 256 pp.
Linke, K.H., Scheibel, C., Saxena, M.C. and Sauerborn, J.
(1992) Fungi occurring on Orobanche spp. and their preliminary evaluation for Orobanche control. Tropical Pest
mangement 38, 127130.
Mller-Stver, D. (2001) Possibilities of biological control
of Orobanche crenata and O. cumana with U. Botrytis
243
XII International Symposium on Biological Control of Weeds
and Fusarium oxysporum f. sp. orthoceras. APIA Verlag,
Laubach, Germany, 166 pp.
Parker, C. and Riches, C.R. (1992) Orobanche species: The
Broomrapes. In: Parasitic Weeds of Worlds. The University
of Arizona Press, Tucson, AZ, pp. 111164.
Raynal, G., Gondran, J., Bournoville, R. and Courtillot, M. (1989)
Enemies et Maladies des Prairies. INRA, Paris, 241 pp.
Sauerborn, J. (1994) Orobanche species. In: Labrada, R.,
Caseley, J.C. and Parker, C. (eds) Weed management for
developing countries. FAO, Rome, Italy, pp. 150155.
Sauerborn, J., Abbasher, A.A. and Kroschel, J. (1994) Biological control parasitic weeds by phytopathogenic fungi.
Biology and management of Orobanche. Biology and
management of Orobanche. In: Pieterse A.H., Verkleig,
j.a.c. and Borg, s. j. Ter. (eds) Proceedings of the Third
International Workshop on Orobanche and Related
Striga Research. Royal Tropical Institute, Amsterdam, The
Netherlands, pp. 545549.
Thomas, H., Sauerborn, J., Mller-Stver, D., Ziergler, A.,
Bedi, JS. and Kroschel, J. (1998). The potential of Fu-
sarium oxysporum f.sp. orthoceras as biological control
agents for Orobanche cumana in sunflower. Biological
Control 13, 4148.
Thomas, H., Sauerborn, J., Mller-Stver, D. and Kroschel,
J. (1999) Fungi of Orobanche in Nepal with potential as
biological control agents. Biocontrol Science and Technology 3, 379381.
Waston, K. and Waymore, L.A. (1992) Les bioherbicides. In:
Morin, G. (ed.) Lutte Biologique. Quebec, Canada, pp.
361374.
Zermane, N., Kroschel, J., Salle, G. and Bouznad, Z.
(1999) Prospects for biological control of Parasitic weed
Orobanche in Algeria. In: Kroschel. J., Abderabihi, M. and
Betz, H. (eds) Advances in Parasitic Weed at On-Farm
Level. Vol II. Joint Action to Control Orobanche in Wana
Rgion. Margraf-Verlag, Weikerheim, Germany, pp.173,
83.
Zonno, M.C. and Vurro, M. (2002) Inhibition of germination
of Orobanche ramosa seeds by Fusarium toxins. Phytoparasitica 30, 519524.
244
Abstracts: Theme 3 Target and Agent Selection
Prospective biological control agents for
Nassella neesiana in Australia and New Zealand
F.E. Anderson,1 J. Barton2,3 and D.A. McLaren4,5
CERZOS-UNS, Camino La Carrindanga Km 7, 8000, Baha Blanca, Argentina
2
Landcare Research, Private Bag 92170, Auckland, New Zealand
3
Present address: 467 Rotowaro Road, RD 1, Huntly 3771, New Zealand
4
Department of Primary Industries Frankston, PO Box 48, Frankston 3199, Australia
5
Cooperative Research Centre for Australian Weed Management, Waite Campus, PMB 1,
Glen Osmond SA 5064, Australia
1
Chilean needle grass (Nassella neesiana, Poaceae), which is native to South America, costs agriculture
millions and is threatening indigenous grasslands in Australia and New Zealand. Field observations
and laboratory experiments have been undertaken in Argentina to find fungal pathogens suitable as
biocontrol agents. Three rust species have been selected: Uromyces pencanus, Puccinia graminella
and Puccinia nassellae. All three have been observed causing severe damage to their host in the field
and are believed to be quite host specific. Attempts to elucidate their life cycles experimentally have
failed to-date, and this is discussed. U. pencanus is the most promising of the three because reliable
methods have been developed for culturing and storing inoculum and applying it to plants. There have
been some technical difficulties with the other two rusts. An isolate of U. pencanus has been found
which can attack six of seven Australian accessions, and it has been selected for host-specificity testing. A different isolate will be needed for New Zealand populations of the weed. Mixed infections by
these rusts are not uncommon in the field. Studies will continue on all three prospective candidates, as
a combination may eventually be needed to achieve the desired level of control.
Biological control of Cirsium arvense
by using native insects
G.A. Asadi, R. Ghorbani, M.H. Rashed and H. Sadeghi
Department of Agronomy, Faculty of Agriculture, Ferdowsi University of Mashhad,
PO Box 91775-1163, Mashhad, Iran
Cirsium arvense is considered as one of the worlds worst weeds and the third most troublesome weed
in Europe. It has become increasingly problematic in ecological areas where conventional control measures are restricted. Thus control that exploits both plant competition and herbivory by specialized native insects may be an inexpensive and sustainable alternative control measure. To date, augmentation
or conservation of native agents has received little attention compared to other approaches, but interest is growing. Reasons for this are that future progress in classical biological control of C. arvense
will depend on the identification of new, host-specific herbivores from the native range and better
predictions and evaluations of non-target impacts. Surveys are being conducted for herbivores on
C. arvense in North Khorasan. We are beginning this project but already realize the importance of
Cassida rubiginosa, a univoltine shield beetle that feeds on foliage of C. arvense. For successful use
of this insect as a biological control agent, knowledge is required about (1) the insect densities required to obtain the desired control level and (2) the factors preventing this insect from attaining such
population levels. This information may lead to the development of strategies to increase population
densities of the agents.
CAB International 2008
245
XII International Symposium on Biological Control of Weeds
The degree of polymorphism in Puccinia punctiformis
virulence and Cirsium arvense resistance:
implications for biological control
M.G. Cripps,1 G.R. Edwards,1 N.W. Waipara,2 S.V. Fowler3 and G.W. Bourdt4
1
Field Service Centre, PO Box 84, Lincoln University, Canterbury, New Zealand
2
Landcare Research, Private Bag 92170, Auckland, New Zealand
3
Landcare Research, PO Box 69, Gerald Street, Lincoln, New Zealand
4
AgResearch, PO Box 60, Gerald Street, Lincoln, New Zealand
Cirsium arvense (Californian thistle) is one of the worst weeds in New Zealand. The host-specific rust
fungus, Puccinia punctiformis, is known to have detrimental effects on this weed; however, its usefulness for biological control in New Zealand has not been fully explored. A collection of C. arvense
ecotypes and rust pathogen isolates from across New Zealand were used in a reciprocal interactions
experiment in order to elucidate different host/pathotype infection combinations. Here, we report on
the degree of polymorphism in this host/pathogen system and the possible implications for biological
control.
Field exploration for saltcedar natural enemies in Egypt
M. Cristofaro,1 F. Di Cristina,2 E. Colonnelli,3 A. Zilli4 and W.M. Amer5
ENEA-Casaccia, BIOTEC, via Anguillarese 301, 00123 Rome, Italy
2
BBCA, Via del Bosco 10, 00060 Sacrofano, Rome, Italy
3
University of Rome La Sapienza, via delle Giunchiglie 56, Rome, Italy
4
Museo Civico di Zoologia, via U. Aldrovandi 18, Rome, Italy
5
Botany Department, Faculty of Science, Cairo University, Cairo, Egypt
1
Genus Tamarix, saltcedar, consists of 90 different species, and 8 of them have been introduced into the
United States in the 1800s. Among them, only two species are considered a real threat to the natural
ecosystems of the southwestern USA: Tamarix parviflora and Tamarix ramosissima. These weeds can
be found primarily in Colorado, Utah, Kansas, Texas, New Mexico, Wyoming and Arizona (Brock,
1994; Di Tomaso, 1998). Once established, saltcedar can out-compete stressed native plants and cover
large areas of formerly native habitat, resulting in a less productive and less diverse environment. Very
promising results were achieved in the biological control domain by the release of the gregarious leaf
beetle Diorhabda elongata. This work aimed to survey the entomofauna associated to Tamarix species
in Egypt.
246
Abstracts: Theme 3 Target and Agent Selection
The phytophagous insects associated with spotted
knapweed (Centaurea maculosa Lam.)
in northeast Romania
A. Diaconu,1 M. Talmaciu,2 M. Parepa1 and V. Cozma1
Institute of Biological Research, Bd. Carol I, 20-A, 700505 Iasi, Romania
University of Agronomy Sciences and Veterinary Medicine, M. Sadoveanu Alley, 700490, Iasi, Romania
1
Spotted knapweed is a Eurasian species that has become a problem weed, especially in mountain
rangelands in North America, where approximately 7 million acres are invaded by this plant. In the
second half of the past century, studies have been conducted with the purpose to introduce several
natural enemies from the region of origin for the biological control of spotted knapweed. Until the
present, 16 biological control agents have been introduced, of which 13 were insect species. In studies
conducted in 2005 and 2006 at multiple sites in northeast of Romania, 20 insect species were
obtained, belonging to the orders Lepidoptera (seven), Diptera-Brachicera (six), Coleoptera (five) and
Hymenoptera-Cinipidae (two). There is an important role for species that attack new shoots in the
reduction of spotted knapweed populations such as Apion sp. (Curculionidae), Napomyza lateralis
(Fallen) (Diptera-Agromizidae) and Tephritidae species (Diptera) and some lepidopteran species.
Parkinsonia dieback: a new association
with potential for biological control
N. Diplock,1 V. Galea,1 R.D. van Klinken2 and A. Wearing1
1
School of Agronomy and Horticulture, University of Queensland, Gatton Campus,
Gatton, QLD, Australia
2
CSIRO Entomology, 120 Meirs Road, Indooroopilly, QLD, Australia
A case study is being carried out investigating the effect of a native fungal pathogen attacking an
invasive woody weed (Parkinsonia aculeata) in rangeland Australia. This is a new association causing impact on parkinsonia that does not appear to be occurring in its native range. Observations have
shown that this dieback is capable of killing whole stands of parkinsonia in small pockets across the
country. Field transects in a naturally occurring dieback site are being monitored to investigate the
movement of this disease through a stand of adult parkinsonia trees. Field and glasshouse trials are
being conducted to observe the effect of isolates taken from diseased plants. Trials so far indicate that
two of these isolates are capable of causing disease in healthy adult plants when applied to a stem
wound. Six months after inoculation, plants have been observed with large spreading stem lesions and
significant reductions in plant vigour. These results are promising with potential for biological control
opportunities for parkinsonia.
247
XII International Symposium on Biological Control of Weeds
Ecology, impact and biological control of the weed
Tradescantia fluminensis in New Zealand
S.V. Fowler,1 N.W. Waipara,2 J.H. Pedrosa-Macedo,3 R.W. Barreto,4
H.M. Harman,2 D. Kelly,6 S. Lamoureaux5 and C.J. Winks2
Landcare Research, PO Box 40, Lincoln, New Zealand
Landcare Research, Private Bag 92170, Auckland, New Zealand
3
Universidade Federal do Paran, Curitiba, Brazil
4
Departamento de Fitopatologia, Universidade Federal de Viosa, Viosa, MG 36571-000, Brazil
5
Agresearch, PO Box 60, Lincoln, New Zealand
6
Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch, New Zealand
1
Tradescantia fluminensis (wandering Jew, family Commelinaceae) is a serious environmental weed
in many frost-free regions of New Zealand. The weed commonly forms dense, monospecific mats
that completely prevent native forest regeneration. It also reduces indigenous biodiversity, alters litter
decomposition and changes soil nutrient availability. Conventional control is difficult and very costly,
so a biological control programme was initiated. A survey of invertebrates and plant pathogens present
in New Zealand found that damage attributed to either insect herbivory or disease was minimal with
little biocontrol potential. Surveys were subsequently initiated in the plants native range in SE Brazil
to identify potential agents associated with plant damage for classical biocontrol. Molecular studies are
under way to determine the clonal status of the plant in New Zealand and to establish more accurately
its geographic origin. In addition, the weed population dynamics will be modelled, with the aim of
further understanding the ecology of this forest invader in New Zealand and improving the prospects
of successful biological control.
Potential for biological control of Rhamnus cathartica
and Frangula alnus in North America
A. Gassmann,1 I. Tosevski1 and L.C. Skinner2
CABI Switzerland Centre, CH-2800 Delmont, Switzerland
Minnesota Department of Natural Resources, St. Paul, MN 55155-4025, USA
1
Rhamnus cathartica and Frangula alnus are small trees of Eurasian origin, which have become invasive in North America. Some 1,000 insect samples collected at 97 buckthorn sites in Europe indicate
that the insect-species richness is higher on R. cathartica than on F. alnus and includes more species
that are host-specific at the species or genus level. Lepidoptera (22 species) largely dominate, followed
by Hemiptera (8 species), Diptera (4 species), Acarina (4 species) and Coleoptera (1 species). Although
there is no clear pattern in terms of direction of dispersal, it appears that Rhamnus and Frangula are
predominant in the Old Word and New World, respectively, and this most probably explains a significant proportion of the variation in the insect-species richness on the two target plants. Minimizing potential non-target effects will likely require the selection of agents which are specific to either
R. cathartica or F. alnus. There are in Europe several arthropod species which are monophagous on
R. cathartica or the host range of which will be limited to a few species in the genus Rhamnus. Biological control of F. alnus with species- or genus-specific agents will undoubtedly be more difficult and
will require additional field surveys.
248
Abstracts: Theme 3 Target and Agent Selection
Arundo donax (giant reed): an invasive weed of the
Rio Grande Basin
J. Goolsby,1 A. Kirk,2 W. Jones,2 J. Everitt,1 C. Yang,1 P. Parker,3 D. Spencer,4
A. Pepper,5 J. Manhart,5 D. Tarin,5 G. Moore,5 D. Watts5 and F. Nibling6
USDAARS, Kika de la Garza Subtropical Agricultural Research Center, Weslaco, TX, USA
2
USDAARS, European Biological Control Laboratory, Montpelier, France
3
USDAAPHIS, Edinburg, TX, USA
4
Invasive and Exotic Research Unit, Davis, CA, USA
5
Texas A&M University, Dept. of Biology, College Station, TX, USA
6
Bureau of Reclamation, Denver, CO, USA
Arundo donax L., giant reed, is an exotic and invasive weed of riparian habitats, irrigation canals and
transportation drainages of the southwestern USA and northern Mexico. Giant reed dominates these
habitats, which leads to: loss of biodiversity; catastrophic stream bank erosion; damage to bridges;
increased costs for chemical and mechanical control along irrigation canals. Most importantly, this
invasive weed competes for water resources in an arid region where these resources are critical to the
environment, agriculture and urban users. A. donax is a good target for biological control because it
has no close relatives in North or South America, and several insects from Mediterranean Europe are
known to be monophagous. Our research program includes: (1) remote sensing and ecohydrology
to determine the distribution and water use of giant reed in the Rio Grande River Basin; (2) use of
microsatellites to determine the origin(s) of the invasive North American vegetative clones; (3) field
studies in the native range; (4) pre-release quarantine impact studies on candidate agents, integrating
ecohydrology and plant architecture to select the most promising agent(s) for full host-range testing
and potential release as biological control agents.
Potential agents from Kazakhstan for
Russian Olive biocontrol in USA
R.V. Jashenko,1 I.D. Mityaev1 and C.J. DeLoach2
1
Tethys Scientific Society, Institute of Zoology, 93 Al-Farabi Street, Almaty 050060, Kazakhstan
2
USDAARS, Grassland, Soil and Water Research Laboratory, 808 East Blackland Road,
Temple, TX 76502, USA
The Almaty, Kazakhstan biological control research group has been involved in Russian Olive biocontrol studies since 2006. This group has two goals: (1) to find effective biological agents (among insects)
of Russian Olive and (2) to study their biological features under native conditions. Our research shows:
there are about 30 insect species mentioned as strict specific natural enemies of Elaeagnus angustifolia: ten homopterans, two hemipterans, nine beetles, one fly and eight lepidopterans. The three most
preferable potential Russian Olive biocontrol agents for introduction into the USA are one beetle and
two psyllids: (1) Altica balassogloi Jcbs. (Coleoptera, Chrysomelidae) damages foliage and shoots,
distributed in south and southeastern Kazakhstan (Arys, Ili, Karatal, Charyn rivers riparian forests);
(2) Trioza magnisetosa Log. (Homoptera, Psylloidea), damages foliage (usually on young trees), distributed in south, central and west Kazakhstan; (3) Trioza furcata Low (Homoptera, Psylloidea), damages foliage (50100% loss of foliage), distributed in central, south and west Kazakhstan. Preliminary
studies indicate that the best agent for biocontrol of Russian Olive in the USA is A. balassogloi.
249
XII International Symposium on Biological Control of Weeds
Biology of the Rumex leaf defoliator sawfly
Kokujewia ectrapela Konow
(Hymenoptera: Argidae) in Urmia region
Y. Karimpour
Department of Plant Protection, Faculty of Agriculture, Urmia University, Urmia-IR, Iran
The Kokujewia ectrapela Konow (Hym. Argidae) belongs to the Caspian fauna. It has been collected
in Russia, Transcaucasia and Iran. The larvae of this sawfly were found living on Rumex spp. (Polygonaceae) and were considered as a potential biological control agent of weedy Rumex spp. Field surveys
were conducted to determine the life history of K. ectrapela in Urmia region (3731 N, 4501 E).
K. ectrapela completes four generations within the growing season and hibernates as a pupal stage inside the protective cocoon in the plant litter surrounding the dock plants. First generation appears from
late April to end June, and the last generation appears from late August to late September. After emergence and copulation, the females attach their eggs to the margins of Rumex leaves in a single row. The
average fecundity was 182 15 eggs per female. Eggs hatch within 56 days at mean daily temperature
25C. Young larvae begin consumption of leaves on the area in between small veins. However, once
larvae are mature, they consume the entire leaf leaving only the midrib and major veins. The larvae of
each generation occurred on host plants for 1020 days depending on natural conditions. Fully grown
larvae of first, second and third generations pupate within silken brownish cocoons spun using host
plant material. The developmental time of K. ectrapela from egg to emerging adult is 2535 days.
What defines a host? Growth rate
the paradox revisited
M.K. Kay
Ensis, 49 Sala Street, Rotorua 3012, New Zealand
Paradoxically, plants that provide a poor diet as a defence against herbivores will potentially lose more
foliage biomass to defoliators compensating by feeding longer to complete their development. It is
generally believed that rapid larval growth is beneficial for the avoidance of natural enemies and that
the trade-off for plants providing slow growth is the higher herbivore mortality resulting from their
increased exposure to natural enemies. This argument has been found wanting in that natural enemies
often prefer healthy prey from prime hosts and that herbivores may benefit from enemy-free-space
by feeding on novel hosts. In fact, it appears that a slow growth rate has far more insidious effects
on defoliators. Using bioassays of polyphagous Lymantriidae, a slow growth rate has been shown to
correlate with the reduction of final body mass, mating opportunities, sex ratios, fecundity, dispersal
abilities and pathogen resistance. Host specialization may well be driven by natural selection for improved growth rate. Unfortunately, many potential biological control agents are rejected when they
feed on non-target hosts in the lab., even though they would have little chance of maintaining viable
populations on those hosts.
250
Abstracts: Theme 3 Target and Agent Selection
Selection of fungal strains for biological control of
important weeds in the Krasnodar region of Russia
T.M. Kolomiets,1 E.D. Kovalenko,1 Zh.. Mukhina,3 S.N. Lekomtseva,2
.V. Alexandrova,2 O.o. Skatenok, I.Uj. Samokhina,1 L.F. Pankratova,1
D.K. Berner4 and S.A. Volkova3
All Russian Research Institute of Phytopathology, 143050, Moscow Region,
Bolshie Vyazemi, Russia
2
M.J.Lomonosovs Moscow State University, Moscow, Russia
3
All Russian Scientific Research Institute of Rice, Krasnodar, Belozerniy
4
USDAARS, Foreign DiseaseWeed Science Research Unit, Fort Detrick, MD, USA
1
Fungi, collected in different districts of the Krasnodar Region of the Russia Federation, were collected
and isolated from diseased weed samples. Weeds sampled included species in the genera Centaurea, Salsola, Vincetoxicum, Carduus, Cirsium, and Echinochloa. Fungal isolates were selected based
on biological tests and the potential of the fungi for classical and/or inundative biological control
of the weeds. A live collection of plant pathogens, isolated from the collected herbarium material,
was formed. This collection consists of the following genera isolated from Centaurea solstitialis and
Centaurea diffusa: Acremonium kiliense, Alternaria alternata, Alternaria radicina, Alternaria brassicae, Aspergillus, Cladosporium, Coniochaeta, Embellisia chlamydospora, Epicoccum sp., Fusarium
culmorum, Fusarium heterosporum, Fusarium oxysporum, Fusarium sporotrichioides, Humicola sp.,
Periconia igniaria, Phoma sp., Rhizoctonia, Sclerotinia, Sordaria, Ustilaginoides ochracea, Verticillium dahliae. The facultative fungi P. igniaria E.W. Mason et M.B. Ellis (Teleomorph Didymosphaeria
igniaria C. Booth) and Phoma glomerata (Corda) Wollenw. et Hochapfel. and the obligate pathogen
Puccinia hieracii var. hieracii (syn. P. jaceae) represented the greatest interest as potential agents for
biological control of C. solstitialis.
Vegetative expansion and seed output of
swallow-worts (Vincetoxicum spp.)
L.R. Milbrath,1 K.M. Averill2 and A. DiTommaso2
USDAARS, U.S. Plant, Soil and Nutrition Laboratory, Tower Road, Ithaca, NY 14853, USA
2
Department of Crop and Soil Sciences, Cornell University, Ithaca, NY 14853, USA
Swallow-worts (Vincetoxicum rossicum, pale swallow-wort, and V. nigrum, black swallow-wort) are
herbaceous, perennial, twining vines related to milkweeds (Apocynaceae). Pale swallow-wort is native
to Ukraine and southeastern European Russia; black swallow-wort is native to southwestern Europe.
Both species are becoming increasingly invasive in the northeastern United States and southeastern
Canada. They grow in both high and low light environments in a variety of disturbed and undisturbed
habitats. The success of a classical biological control program for swallow-worts will be dependent
on the availability of critical biological and ecological data about the target species, such as which life
stage(s) are important for population growth and most sensitive to control efforts, which in turn will
affect the selection of candidate biological control agents. Assessments of the rate of vegetative expansion and reproductive output of isolated swallow-wort plants have begun at several sites in New York
State, including old-field and forest understory habitats within sites. In 1 year, the number of tillers per
pale swallow-wort plant increased by 45% in old fields and 19% in the forest understory. Follicle (seed
pod) production was generally lower in the forest understory than old-field habitats. Monitoring will
continue for at least the next 2 years.
251
XII International Symposium on Biological Control of Weeds
A new biological control program for common tansy
(Tanacetum vulgare) in Canada and the USA
A.S. McClay,1 M. Chandler,2 U. Schaffner,3 A. Gassmann3 and G. Grosskopf3
McClay Ecoscience, 15 Greenbriar Crescent, Sherwood Park, AB, Canada T8H 1H8
Minnesota Department of Agriculture, 601 North Robert Street, Saint Paul, MN 55101, USA
3
CABI Switzerland Centre, 1, rue des Grillons, Delmont CH-2800, Switzerland
1
Common tansy (Tanacetum vulgare L., Asteraceae) is an invasive herbaceous perennial native to
Europe. It was introduced into North America as a culinary and medicinal herb. Now widely naturalized in pastures, roadsides, waste places and riparian areas across Canada and the northern USA, tansy
is also spreading in forested areas. It contains several compounds toxic to humans and livestock if
consumed, particularly -thujone. Tansy reduces the productivity of pastures, displaces native vegetation in natural areas and can be a problem in regeneration of logged areas. It is listed as a noxious
weed in several states and provinces. Common tansy is a good target for biological control, as it is a
perennial plant growing in stable habitats and has few native North American congeners. A biological
control program for common tansy is funded and coordinated by a CanadianUS consortium led by
the Alberta Invasive Plant Council and the Minnesota Department of Agriculture. CABI Switzerland
Centre is identifying and testing potential agents for efficacy and host specificity. Potential agents
include the stem-mining weevil Microplontus millefolii, the leaf-feeding beetle Cassida stigmatica, the
rhizome-mining moths Isophrictis striatella and Dichrorhampha spp., and the stem, rosette and flower
head-galling gall midge Rhopalomyia tanaceticola.
Surveys in Argentina for the biological control of
Brazilian peppertree in the USA
F. McKay,1 G. Cabrera Walsh,1 M.I. Oleiro1 and G.S. Wheeler2
1
USDAARS, South American Biological Control Laboratory, Bolivar 1559, B1686EFA,
Hurlingham, Buenos Aires, Argentina
2
USDA/ARS Invasive Plant Research Laboratory, Fort Lauderdale, FL, USA
Brazilian peppertree (Schinus terebinthifolius: Anacardiaceae) is a perennial tree native to Argentina,
Brazil and Paraguay that was introduced in the USA during the late nineteenth century as ornamental.
In Hawaii, it has been mentioned as one of the most significant invasive plants, threatening federally
listed endangered native plants, and in southern Florida, it has been ranked among the most important
threats to biodiversity. Although most surveys have been conducted in Brazil, Argentina seems to be
the most likely centre of dispersion of the genus, with 22 of the 30 species of Schinus. Consequently,
in June 2004, an agreement for cooperation on Brazilian peppertree research was initiated between the
USDAARS Invasive Plant Research Laboratory, Fort Lauderdale, FL, and the USDAARS South
American Biological Control Laboratory. Recent surveys in Argentina showed that Brazilian peppertree (BP) populations harbour many phytophagous insects not previously noticed. Although most of
them feed on several species within the Anacardiaceae, three of these insects, a leaf-blotcher moth, an
eriophyid leaf mite and a thrips that feeds on shoots, have shown a very narrow field-host range. The
potential of these natural enemies as biocontrol agents of BP, and the specificity problems implied by
working in the Anacardiaceae family are discussed.
252
Abstracts: Theme 3 Target and Agent Selection
Natural enemies of balloon vine and
pompom weed in Argentina: prospects for
biological control in South Africa
F. McKay,1 M.I. Oleiro,1 A. McConnachie2 and D.O. Simelane3
1
USDAARS, South American Biological Control Laboratory, Bolivar 1559, B1686EFA,
Hurlingham, Buenos Aires, Argentina
2
ARC-PPRI Weeds Division, Private Bag X6006, Hilton 3245, South Africa
3
ARC-PPRI Weeds Division, Private Bag x134, Queenswood 0121, South Africa
As part of a new South African strategy for targeting weeds at an early stage of invasion (emerging
weeds), a cooperative research agreement has been signed between the Plant Protection Research Institute (PPRI), South Africa and the USDAARS South American Biological Control Laboratory. The
objective of this collaboration is to search for and study host-specific natural enemies of the emerging weeds, balloon vine (Cardiospermum grandiflorum: Sapindaceae) and pompom weed (Campuloclinium macrocephalum: Asteraceae). Balloon vine, a perennial woody climber, originally from
tropical and sub-tropical America, and pompom weed, an ornamental herb native to South and Central
America and Mexico, are considered serious invasive weeds in South Africa. Several exploratory trips
were conducted in Argentina to survey for potential natural enemies of these weed species. Among
the natural enemies found on balloon vine, the seed-feeding insects Cissoanthonomus tuberculipennis (Coleoptera: Curculionidae) and Lisseurytomella flava (Eulophidae) constitute the most promising
candidates. Surveys on pompom weed revealed the presence of the stem-galling thrips, Liothrips sp.
(Thysanoptera), Cochylis n. sp. (Tortricidae) and Adaina sp. prob. simplicius Grossbeck (Pterophoridae), two flower-feeding moths that cause considerable damage to the developing seeds. The potential
of these insects as biocontrol agents is currently being assessed in Argentina and South Africa.
Tamarix biocontrol in US: new biocontrol
agents from Kazakhstan
I.D. Mityaev,1 R.V. Jashenko1 and C.J. DeLoach2
1
Tethys Scientific Society, Institute of Zoology, 93 Al-Farabi Street, Almaty 050060, Kazakhstan
2
United States Department of Agriculture, Agricultural Research Service, Grassland,
Soil and Water Research Laboratory, 808 East Blackland Road, Temple, TX 76502, USA
According to research of Kazakhstan biological control research group (since 1994), the four most
preferable potential Tamarix biocontrol agents for introduction into the USA (after Diorhabda elongata) are: (1) the stem-galling moth, Amblypalpis tamaricella (Lepidoptera: Gelechiidae); (2) the fo
liage and flower galling psyllid,Crastina tamaricina (Homoptera: Psylloidea, Aphalaridae); (3) the gall
midge, Psectrosema noxium (Diptera: Cecidomyiidae); and (4) the foliage-feeding weevil, Coniatus
steveni (Coleoptera: Curculionidae). All can heavily damage Tamarix in Kazakhstan, and all have some
protection from predators and from drowning. The best agent among four species is A. tamaricella; it
inhabits riparian forests and deserts in south and southeastern Kazakhstan, and heavy infestations are
capable of killing entire trees.
253
XII International Symposium on Biological Control of Weeds
Biological control of aquatic weeds by Plectosporium
alismatis, a potential mycoherbicide in Australian rice
crops: comparison of liquid culture media for their ability
to produce high yields of desiccation-tolerant propagules
C. Moulay,1 S. Cliquet,1 K. Zeehan,1 G.J. Ash2 and E.J. Cother3
Biopesticide Research, Laboratoire de microbiologie applique de Quimper (LUMAQ),
Universit de Bretagne Occidentale, 2, rue de lUniversit, Quimper 29000, France
2
E.H.H Graham Centre for Innovative Agriculture, School of Agricultural and Veterinary Sciences,
Charles Sturt University, Boorooma Street, PO Box 588, Wagga Wagga, NSW 2678, Australia
3
NSW Department of Primary Industries, Agricultural Institute, Orange, NSW 2800, Australia
1
In our laboratory, a methodology for the development of a stable, effective Plectosporium alismatis mycoherbicide is currently being investigated. In this work, we compared our standard optimized
medium to other liquid media for their ability to support high conidial and chlamydospore yields and
subsequent tolerance of conidia and chlamydospores to air-drying. When grown in a casamino acidsglucose based liquid medium, P. alismatis developed hyphae and produced high yields of conidia
(1 107 conidia ml1) and dry weights (220 mg dry weights per erlen), while no chlamydospore was
formed. In a nitrate-glucose based medium, growth was poor, P. alismatis producing aggregated
hyphae that contained chlamydospores (6.5 104 chlamydospores per millilitre). The addition of
nitrate in the casamino acids-glucose based medium restored partially chlamydospore formation
(1 104 chlamydospores per millilitre). Although our standard, optimized medium produced 2 105
chlamydospores per millilitre, less than 10% chlamydospores and 50% conidia remained viable after
15 days storage at 25C, while 50% chlamydospores produced in the nitrate-glucose based medium
were still viable after 30 days in the same storage conditions; moreover, these chlamydospores sporulated through a microcycle conidiation.
Herbivorous insects from Brazil for classical
biocontrol of Tradescantia fluminensis
J.H. Pedrosa-Macedo,1 S.V. Fowler,2 M. Silvrio,1 K. Doetzer,1
M. Livramento1 and L. Suzuki1
Laboratorio Neotropical de Controle Biologico de Plantas, CIFLOMA-Setor de Ciencias Agrarias,
Universidade Federal do Paran, Rua Bom Jesus 650, Juveve 80.035-010, Curitiba PR, Brasil
2
Landcare Research, PO Box 40, Lincoln, New Zealand
Tradescantia fluminensis (wandering Jew; family Commelinaceae) is a South American plant that is
an exotic invasive weed in New Zealand and elsewhere. Field surveys and preliminary host range tests
are underway in its native range in Brazil for herbivorous arthropods with classical biocontrol potential
for New Zealand. Species found inflicting locally high levels of damage to T. fluminensis include two
chrysomelid beetles (Buckibrotica cinctipennis and Lema sp nr guerini), a coleophorid moth (Idioglossa sp), a thrips Scirtothrips sp (Thripridae) and a sawfly. Other herbivorous insects located include
several additional chrysomelid beetles (one a leaf-mining species), a gall midge, and the noctuid moth
Mouralia tinctoides. This noctuid is native to South, Central and North America and appears to attack
a range of plant species in the family Commelinaceae. However, the Commelinaceae (or even the order Comminales) contains no native New Zealand plant species and no plants of significant economic
benefit to the country. Hence, host specificity to species, genus or even subfamily/family is, in theory,
not essential in this programme. This lack of close plant relatives has also resulted in a list of plants for
host-range testing that is unusual because it does not include any New Zealand native species.
254
Abstracts: Theme 3 Target and Agent Selection
Nigrospora oryzae, a potential bio-control agent for
Giant Parramatta Grass (Sporobolus fertilis) in Australia
S. Ramasamy,1,4 D. Officer,3,4 A.C. Lawrie1 and D.A. McLaren2,4
RMIT University, Bundoora West Campus, PO Box 71, Bundoora 3083, Australia
Department of Primary Industries, Frankston Centre, PO Box 48, Frankston 3199, Australia
3
NSW Agriculture, PMB 2, Grafton NSW 2480, Australia
4
CRC for Australian Weed Management, Australia.
1
Giant Parramatta Grass (GPG) is an aggressive perennial tussocky grass from tropical Asia that is a
declared noxious weed in Australia. It invades native pastures and reduces animal production. Its potential distribution is estimated at 23.7 million hectares in Australia. A fungus was isolated from dead
shoot tips and flag leaves of Sporobolus fertilis (Steud.)Clayton (Poaceae) in Australia. The fungus was
identified as Nigrospora oryzae (Berk and Broome) Petch based on the morphological characteristics
and fruiting bodies. Its identity was confirmed by DNA sequencing using primers ITS 1 and ITS 4
to the internal transcribed spacer (ITS) region. N. oryzae was investigated as a potential bio-control
agent for GPG. Forty healthy plants of uniform size were selected for the experiment. Twenty plants
were inoculated to run-off with spore suspension (106/ml in 0.1% Tween 20) and the control plants
with Tween 20 alone. Necrosis and cessation of growth occurred in inoculated plants but not in control
plants. This is the first report of N. oryzae on GPG in Australia, and further trials are warranted to test
its potential as a bio-control agent for this noxious weed.
Biological control and ecology of the submerged
aquatic weed Cabomba caroliniana
S.S. Schooler,1 G.C. Walsh2 and M.H. Julien3
CSIRO Entomology, 120 Meiers Road, Indooroopilly, QLD 4068, Australia
USDAARS South American Biological Control Laboratory, Bolivar 1559, Hurlingham,
Buenos Aires, Argentina
3
CSIRO Entomology European Laboratory, Campus International de Baillarguet,
34980 Montferrier sur Lez, France
1
Cabomba caroliniana is a submerged aquatic plant from South America that is becoming a serious
weed worldwide. It spreads by seed and by fragmentation and has an extremely wide climatic range,
invading lakes and ponds from tropical (Darwin, Australia: latitude 12) to cold temperate regions
(Peterborough, Canada: latitude 45). There are currently no effective methods of managing cabomba
infestations, and funding has been allocated to research biological methods. Surveys have examined
cabomba in its native range and have identified several potential biological control agents. The most
promising are a stem-boring weevil and two aquatic Pyralid moths. We have also examined the effects
of depth and season on the dynamics of cabomba populations in Australia. We found that cabomba
exhibits no clear seasonal patterns in biomass at three lakes in southeast Queensland. The plant has
greatest biomass in 23 m depth of water (mean = 185.6 g m2, SD = 118.8 g m2), but rooted plants
were found down to depths of 6 m. This study indicates that host plant resources will be available for
biological control agents throughout the year, which is likely to result in more stable and potentially
more effective biological control.
255
XII International Symposium on Biological Control of Weeds
Hindsight is 20/20: improved biological control
of Chromolaena odorata (Asteraceae) for
seasonally dry regions
L.W. Strathie,1 C. Zachariades,1 O. Delgado2 and C. Duckett3
ARC-Plant Protection Research Institute, Private Bag X6006, Hilton 3245, South Africa
Museo del Instituto de Zoologa Agrcola Francisco Fernndez Ypez MIZA, Universidad Central
de Venezuela, Facultad de Agronoma, Maracay, Estado Aragua, Apartado 4579, Venezuela
3
Department of Entomology, Blake Hall, 93 Lipman Drive, Rutgers University,
New Brunswick, NJ 08901, USA
1
Currently established biocontrol agents on Chromolaena odorata achieve a measurable degree of control in high rainfall regions but have limited or no success in regions that experience a distinct dry,
fire-prone season. Earlier consideration of this limiting factor when selecting agents may have enabled
greater control than has been achieved. Insect species from seasonally dry climates within the Neotropical native range of C. odorata that have soil-dwelling or diapausing stages have thus become foci
within the South African research programme and have relevance for control programmes elsewhere.
Field host-range surveys were conducted within the native range for three potential agents, in conjunction with laboratory investigations, and for one species, with molecular and morphological taxonomic
studies of specimens collected from several Asteraceae. These data provide convincing evidence regarding the unsuitability of the root-feeder Longitarsus sp. (Coleoptera: Chrysomelidae) and the likely
suitability of the stem-galler Conotrachelus reticulatus (Coleoptera: Curculionidae) and stem-borer
Carmenta sp. nov. (Lepidoptera: Sesiidae) for biological control of C. odorata. The value of integrating native host range investigations with laboratory studies is discussed.
Surveys for herbivores of Casuarina spp. in Australia for
development as biological control agents in Florida, USA
G.S. Taylor,1 G.S. Wheeler2 and M.F. Purcell3
Australian Centre for Evolutionary Biology and Biodiversity, The University of Adelaide, PMB 1,
Glen Osmond, South Australia 5064, Australia
2
USDAARS, Invasive Plant Research Laboratory, 3225 College Avenue, Fort Lauderdale, FL 33314, USA
3
CSIRO Entomology, USDAARS, Office of International Research Programs, Australian Biological
Control Laboratory, 120 Meiers Road, Indooroopilly, Queensland, Australia 4068
1
The Australian pines, Casuarina equisetifolia, Casuarina glauca and Casuarina cunninghamiana,
have become serious invasive weeds in southern Florida. With rapid growth and thick litter accumulation, they dramatically alter the habitat and inhibit growth of native flora and associated fauna. In coastal
dunes, C. equisetifolia interferes in nesting of the endangered sea turtle and American crocodile. C.
glauca occurs extensively in seasonally inundated sub-coastal areas. In addition to seed production, it
suckers profusely, creating dense, ecologically barren monocultures. Despite its weed status, C. cunninghamiana is being considered by the citrus industry for windbreaks. Being dioecious, propagation
of male plants would limit dispersal, but also have implications for biocontrol. Surveys for potential
agents commenced in Australia in 2004 and includes collection of insects and galls and cones for rearing cone-feeders or granivores. Potential agents include seed-feeding and gall-forming Hymenoptera,
Lepidopteran defoliators, plant-hoppers, psyllids, scales and curculionids. Many are narrowly host specific, even within the Casuarinaceae, but their potential to damage their hosts is little known, especially
in the natural environment where their populations may be moderated by predators and parasitoids.
Future studies aim to investigate the systematics and coevolution of insect herbivores by exploring the
phylogenetic congruence between potential biocontrol agents and their plant hosts.
256
Abstracts: Theme 3 Target and Agent Selection
Differential host preferences of Diorhabda elongata:
implications for biological control of Tamarix
H.Q. Thomas
Department of Entomology, University of California-Davis, 1 Shields Avenue,
Davis, CA, 95616 USA
In 20042006, releases of Diorhabda elongata in California resulted in poor establishment and efficacy for control of Tamarix spp. Previous data suggests differential host preferences favouring invasive Tamarix ramosissima could have contributed to establishment failure at California sites where
Tamarix parviflora is present. In September 2006, a new population of D. elongata from T. parviflora
in Greece was imported to determine if higher preference for T. parviflora exists and whether the program could be improved. Here, I present the results of previous experiments justifying importation of
this colony. An open-field host-choice test was conducted in July 2006 between T. ramosissima and
T. parviflora using the original colony. D. elongata showed marked ovipositional preference for mixed
and T. ramosissima treatments over T. parviflora (F = 6.57, df = 2,10, P = 0.015). Adult presence was
also significantly higher on these treatments than on T. parviflora alone (repeated measures MANOVA
F = 7.93, df = 2,14, P = 0.005). A caged test conducted in Greece showed that beetles from the collection site of the original colony favour Tamarix smyrnensis over T. parviflora, whereas T. smyrnensis is
often synonymized with T. ramosissima (F = 12.66, df = 1,4.32, P = 0.02). Tests will now be conducted
to compare the host ranges of the new population and the original colony.
Hybridization potential of Saltcedar leaf beetle,
Diorhabda elongata, ecotypes
D.C. Thompson,1 B.A. Petersen,1 D.W. Bean2 and J.C. Keller3
Department of Entomology, Plant Pathology and Weed Science, New Mexico State University,
Las Cruces, NM 87003, USA
2
Colorado Department of Agriculture, Biological Pest Control, Palisade Insectary,
750 37.8 Road, Palisade, CO 81526, USA
3
Former address: USDA ARS Exotic and Invasive Weeds Research Unit,
Western Regional Research Center, 800 Buchanan Street, Albany, CA 94710, USA
Saltcedar (Tamarix spp.) is an invasive riparian shrub/tree in the western USA that displaces native
plants, increases soil salinity and wildfires and lowers water tables. Diorhabda elongata feeds exclusively on saltcedar in Europe and Asia. The biological control potential of seven ecotypes is being tested:
Fukang and Turpan, China; Chilik, Kazakhstan; Posidi and Crete, Greece; Karshi, Uzbekistan; and
Tunis, Tunisia. The Fukang and Kazakhstan ecotypes are being released and defoliating large acreages
of saltcedar in the northern half of the western USA, while all other ecotypes are being released in the
southern half. Although different ecotypes have been released in the western USA, the effects of hybridization between ecotypes need to be understood before deliberately mixing ecotypes. All ecotypes
will hybridize in controlled settings when not given a choice of mate. Some of these hybrids produce
sterile offspring which could disrupt long-term population dynamics in field populations. The Greek
ecotypes were tested with other ecotypes. Tests were conducted in a controlled, small caged environment, differing in sex ratio, in a controlled, larger caged environment, differing in ecotype numbers,
and in a large controlled, open greenhouse. Pure-breeding and cross-breeding was observed in testing
Greek ecotypes with other ecotypes. Implications for biological control will be discussed.
257
XII International Symposium on Biological Control of Weeds
Pathogens as potential classical biological control
agents for alligator weed, Alternanthera philoxeroides
M.G. Traversa,1 M. Kiehr,1 R. Delhey,1 A.J. Sosa2 and M.H. Julien3
Alligator weed (Alternanthera philoxeroides) is an evergreen species native of South America. It is
an invasive plant in Australia, USA, China and other countries. To identify possible candidates for
the biological control of this plant, surveys of fungal pathogens were carried out in Buenos Aires
and northwestern and northeastern provinces in Argentina between November 2004 and May 2005.
Thirty sites were surveyed, and at least 12 fungal species were collected. Colletotrichum orbiculare
and Colletotrichum cf. capsici, associated with stem lesions and leaf spots, were widely distributed and
showed a high incidence and impact in the plant populations. Fusarium sp., associated with concentric
large leaf spots, also had high incidence. The white rust Albugo bliti was collected on A. philoxeroides and on the closely related Alternanthera aquatica, but its impact seems to be limited. Phoma sp.,
Phomopsis sp. and other fungi have also been identified. The fungi likely to be the most promising as
candidates for classical biological control of alligator weed are Colletotrichum orbiculare, Colletotrichum capsici and Fusarium sp.
A survey for fungal pathogens with potential for
biocontrol of exotic woody Fabaceae in Argentina
M.G. Traversa, M. Kiehr and R. Delhey
Laboratorio de Patologa Vegetal, Departamento de Agronoma, Universidad Nacional del Sur,
CC 738 (8000), Baha Blanca, Buenos Aires, Argentina
Several exotic woody plants in the Fabaceae are aggressive invaders in native ecosystems in Argentina,
especially in the Pampas. They transform local communities, replace native plant species and cause
economic damage. It is assumed that a lack of natural enemies contributes to the success of these
invaders. As a first step towards their eventual biocontrol, we studied the fungal pathogens naturally
associated with these exotic plants in the southern Pampas region. A very aggressive dieback was
observed on Spartium junceum, Acacia baileyana, Acacia mearnsii and less so on Genista monspessulana. A Phomopsis sp. was always found associated with this dieback. Fusarium sacchari var. subglutinans was also isolated from S. junceum plants showing dieback. Another Fusarium sp. was isolated
from A. mearnsii plants with similar symptoms. The rust Uromyces genistae-tinctoriae (uredinia only),
frequently infected with the hyperparasite Sphaerellopsis filum, was identified on G. monspessulana.
The powdery mildew fungus Erysiphe rayssiae (anamorph only) was observed on S. junceum, where
it causes some damage on re-growth. Further studies are necessary to determine whether some of these
naturally occurring pathosystems could be manipulated to help control these exotic plant invaders,
e. g. via the inoculative approach.
258
Abstracts: Theme 3 Target and Agent Selection
Applied biocontrol, a landscape comparison of
two Dalmatian toadflax agents
S.C. Turner
British Columbia Ministry of Forests and Range, 515 Columbia Street, Kamloops,
British Columbia, Canada V2C 2T7
British Columbians have a strong societal sense for protecting their environment and its resources.
These values are reflected in their provincial government which has a stewardship responsibility to
approximately 885,600 km2 or 93% of the province (MSRM, 1997). The IAPP Application can be accessed on-line at http://www.for.gov.bc.ca/hfp/invasive/index.htm. It houses activities for management
of invasive alien plants in British Columbia: planning, inventory, mechanical, chemical and biological
control, the monitoring of each of these activities and biological control agent dispersal. The IAPP Application is structured to track sites and their characteristics as geographic locations. Invasive species
that invade the site are then recorded. Multiple invasive species with multiple surveys can be inputted
on a single site. This allows recording of the change in the invasive plant community over time as
well as the level of success of our treatment efforts. A compilation of this data allows assessment of
the current set of biocontrol agents in the province for a target plant species. By comparing the spread
of Dalmatian toadflax to the habitat requirements of the biocontrol agents, it is possible to determine
whether sufficient, suitable agents exist in the province or whether subsequent screening of agents
must be pursued.
Survey of European natural enemies of
Swallow-worts (Vincetoxicum spp.)
A.S. Weed,1 R. Casagrande1 and A. Gassmann2
Department of Plant Sciences, University of Rhode Island, Kingston, RI, USA
2
CABI Switzerland Centre, Delmont, Switzerland
Two swallow-worts (Vincetoxicum nigrum and V. rossicum) originating from Europe have become established in the eastern USA and Canada. Their population expansion and aggressive growth threaten
native plant species, alter ecological processes and cause problems in agricultural settings. The lack
of herbivory on these plants by native insects in North America and the difficulty in controlling these
weeds has spawned interest in a biological control program. During 2006, surveys for potential biocontrol agents in Central and Eastern Europe revealed the herbivores: Eumolpus asclepiadeus and Chrysolina aurichalcea (Chrysomelidae); Euphranta connexa (Tephritidae); and Abrostola asclepiadis and
Hypena opulenta (Noctuidae). Caterpillars of H. opulenta are leaf-feeders, and this multivoltine species successfully develops on both target weeds. This species was not previously reported developing
on Vincetoxicum. Host- range testing has shown that both chrysomelids feed on the leaves of the target
weeds as adults and the root-feeding larvae of E. asclepiadeus feed and develop on both target weeds.
Future research will continue with host-range and specificity testing of each species to evaluate their
potential as biological control agents of Vincetoxicum.
259
XII International Symposium on Biological Control of Weeds
Climate matching and field ecology of
Australian Bluebell Creeper
A.M. Williams,1,2 H. Spafford Jacob1,2 and E. Bruzzese2,3
Faculty of Natural and Agricultural Sciences, The University of Western Australia,
35 Stirling Hwy, Crawley, WA, 6009 Australia
2
Cooperative Research Centre for Australian Weed Management, Glen Osmond, Australia
3
Primary Industries Research Victoria, PO Box 48, Frankston, VIC 3199, Australia
1
Billardiera heterophylla (Lindl.) L.Cayzer and Crisp (bluebell creeper) has become a serious environmental weed in Victoria, South Australia and Tasmania. Unlike its growth habit in southwest Western
Australia, where bluebell creeper is indigenous, the invasive populations smother existing vegetation,
out-competing and threatening local flora and fauna. Chemical and mechanical control measures have
had limited success, so investigations were carried out to determine if bluebell creeper is suitable for
biological control. A range of genetic and ecological experiments have been conducted on bluebell
creeper to complement the surveys for natural enemies in the indigenous range. Herbivoreplant associations occurring within the invasive range were also surveyed to predict any competition against
the potential biological control agents. This paper will discuss climate matching and field ecological
studies that were performed. These studies were conducted to understand the differences between the
indigenous and invasive bluebell creeper populations. Potential implications towards success of a biological control program and recommendations on the range of the survey area for potential biological
control agents is discussed.
260
Theme 4:
Pre-release Specificity and
Efficacy Testing
Session Chair: Hariet Hinz
261
This page intentionally left blank
The importance of molecular tools in
classical biological control of weeds:
two case studies with yellow starthistle
candidate biocontrol agents
G. Antonini,1 P. Audisio,2 A. De Biase,2 E. Mancini,2 B.G. Rector,3
M. Cristofaro,4 M. Biondi,5 B.A. Korotyaev,6 M.C. Bon,3
A. Konstantinov7 and L. Smith8
Summary
Molecular analyses can play a primary role in the process of host-specificity evaluation at species
and population levels. In this paper, we present two examples of their application with new candidate
biological control agents for yellow starthistle (YST), Centaurea solstitialis L. One is Ceratapion
basicorne (Illiger) (Coleoptera: Apionidae), a root-crown-boring weevil, which, although highly
host specific, could develop on safflower in laboratory tests. Molecular genetic analyses allowed us
to identify larvae from test plants in a field experiment, which conclusively demonstrated that safflower was not at risk. As a result, in 2006, the insect was petitioned for release in the USA. Another
is the flea beetle, Psylliodes chalcomera Illiger (Coleoptera: Chrysomelidae), normally associated
with Carduus. It was discovered feeding on YST and Onopordum in Russia. Laboratory experiments
showed that some populations preferred different host plants. Because it was not possible to differentiate between these insect populations by morphological characters, molecular genetic analyses were
conducted to infer phylogenetic relationship between these populations. Results showed a significant
divergence between some populations, enabling some to qualify for further evaluation as prospective
biological control agents. The ability of molecular genetics to enable us to distinguish cryptic species
may greatly increase the number of potential biological control candidates available.
Keywords: Ceratapion basicorne; Psylliodes chalcomera; COI.
Introduction
1
Biotechnology and Biological Control Agency, Via del Bosco 10,
00060 Sacrofano, Rome, Italy.
2
University of Rome La Sapienza, Department of Animal and Human
Biology, Viale dellUniversit 32, 00185 Rome, Italy.
3
European Biological Control Laboratories, USDA-ARS, 34980 Montferrier sur Lez, France.
4
ENEA C.R. Casaccia, s.p. 25, Via Anguillarese 301, 00060 S. Maria di
Galeria, Rome, Italy.
5
Universit degli StudiLAquila, Dipartimento di Scienze Ambientali,
Via Vetoio snc, 67010 Coppito, LAquila, Italy.
6
Zoological Institute RAS, Laboratory of Insect Systematics, Universitetskaya nab., 1 St. Petersburg, 199034, Russia.
7
Systematic Entomology Laboratory, USDA, c/o Smithsonian Institution, P.O. Box 37012, National Museum of Natural History, MRC-168,
Washington, DC 20013-7012, USA.
8
USDA-ARS, 800 Buchanan Street, Albany, CA 94710, USA.
Corresponding author: G. Antonini <[email protected]>.
CAB International 2008
The main goals of classical biological control of weeds
are to introduce and establish natural enemies that will
effectively control pest populations and not harm nontarget species. However, in relation to the former, and
considering the successes achieved in the past, the outcome of a biological control programme is far from
predictable. In general, only about 60% of introductions have resulted in establishments and about 30%
in successful pest suppression (Crawley, 1989; Julien,
1989; Cruttwell-McFadyen, 2000). We believe that better understanding of natural enemies and pest groups
will help increase rates of biological control agent introduction and success. Molecular markers can provide
biological control practitioners with a valuable tool to
identify cryptic species and biotypes, explore population structure and trace the origin of pests and natural
263
XII International Symposium on Biological Control of Weeds
enemies (Unruh and Woolley, 1999; Ehler et al., 2004).
The application of molecular methods can provide
substantial help especially for insects and pathogens,
where the absence of morphological distinction is frequent and accurate taxonomic identification is a key
point. In addition to applied issues, molecular markers
can be used to refine the phylogeny of host plants and
natural enemies. Phylogenetic information is essential in understanding the evolutionary history of host
plant used by insects. It provides the principal basis for
the choice of the plant species to be evaluated in host-
specificity testing (Wapshere, 1974; Briese, 1996).
Recently, insect molecular systematics has undergone a remarkable revolution. Advances in methods
of data generation and analysis have led to the accumulation of large amounts of DNA sequence data from
most of the major insect groups (Caterino et al., 2000).
Furthermore, the diminishing number of specialist taxonomists and the intrinsic difficulty of identifying herbivores and pathogens present a very big obstacle to
biological control projects. Molecular methods, including genetic barcoding, in combination with classical
taxonomy, can improve our ability to identify natural
enemies (Hebert et al., 2003).
The yellow starthistle (YST), C. solstitialis L. (Asteraceae, Cardueae), is an invasive alien weed in western USA and Canada (Maddox et al., 1985; Pitcairn et
al., 2006). Conventional control strategies have been
insufficient because of the extent of infestation and
economic and environmental costs of herbicides. The
weed is the target of a multidisciplinary biological control programme that involves classical and molecular
approaches (Cristofaro et al., 2002; De Biase et al.,
2003; Smith, 2004; Smith et al., 2005). During the past
years, two promising natural enemies for YST were
identified and studied. One of these is Ceratapion basicorne (Illiger) (Coleoptera: Apionidae), a root-crownboring weevil. Although this insect was highly host
specific in extensive laboratory trials, it could complete
its development on safflower, Carthamus tinctorius L.
(Asteraceae) (Smith and Drew, 2006; Smith, 2007).
Therefore, open-field tests were carried out in eastern
Turkey to determine if, in natural conditions, C. basicorne would attack safflower at locations where the
insect is abundant on YST (Uygur et al., 2005; Smith et
al., 2006). Because the insect pupates inside the plant,
it was necessary to harvest plants before the adults
emerged so that they could be identified. However,
some plants decayed first, so we had to preserve the
larvae, which cannot be identified to species based on
morphology. We were able to develop a method using
molecular marker systems to identify immature stages
of species of the genus Ceratapion attacking YST and
safflower in open-field conditions.
The other agent, Psylliodes chalcomera Illiger (Coleoptera: Chrysomelidae), is a Eurasian flea beetle, which
attacks the developing stem and leaves of YST (Dunn
and Rizza, 1976; Cristofaro et al., 2004). P. chalcomera
is also reported as P. chalcomerus by several cataloguers and specialists on flea beetles in Europe; however,
we are reporting P. chalcomera in this paper because it
is the species name historically used in the biological
control of weed programme in North America (Gruev
and Dberl, 1997; Campobasso et al., 1999; Gruev and
Dberl, 2005). In Italy, P. chalcomera is associated
with the musk thistle, Carduus nutans L. (Asteraceae)
(Dunn and Campobasso, 1993). Indeed, a population of
the beetle from Italy was introduced in the USA to control musk thistle in 1997 (Piper and Nechols, 2004), but
its establishment is unknown. More recently, two sympatric populations were found on different host plants,
YST and Onopordum acanthium L. (Asteraceae), in
southern Russia (Cristofaro et al., 2004). Although Italian and Russian populations are not morphologically
distinguishable, laboratory host range, life history and
field experiments showed significant differences between these host-plant populations (Cristofaro et al.,
unpublished data). In this paper, we report results of
molecular analyses carried out to (1) identify the species of Ceratapion larvae developing in safflower and
YST in field experiments in Turkey and (2) clarify the
taxonomical status of populations of Psylliodes associated with different host plants in the field. For the two
insects, we sequenced two different portions of the
mitochondrial DNA (mtDNA) cytochrome oxidase I
(COI). Many aspects of the structure and evolution of
mtDNA have made it a valuable tool to measure genetic
variation. These include its simplicity of isolation, high
copy number, lack of recombination, relative degree of
conservation of sequence and structure across metazoa
and wide range of mutational rates in different regions
of the molecule. The latter has been successfully used
both in insect-population studies and species distinction (Moritz et al., 1987; Wolstenholme, 1992; Simon
et al.; 1994). These two examples show how molecular
genetic tools can be used to solve critical problems in
the development of new biological control agents.
Methods and materials
Insect specimens
A field experiment was conducted in Eastern Turkey
to determine if C. basicorne would attack safflower at
natural sites with high populations of the insect (Smith
et al., 2006). Larvae of C. basicorne develop in the
upper root and lower stem of YST and pupate inside
the plant. However, other species of Ceratapion also
occurred at one of these sites. To identify the insects,
we collected plants just as the insect was beginning
to pupate. We kept them in individual bags until they
reached the adult stage. Because many of the plants deteriorated before adults emerged, we preserved larvae in
99.9% ethanol for molecular identification. Five adults
of C. basicorne and four of C. scalptum reared from
experimental field plots in Eastern Turkey (Askale,
264
The importance of molecular tools in classical biological control of weeds
Table 1.
Specimens of Ceratapion used for the genetic
analysis (CB, adult Ceratapion basicorne; CS,
adult C. scalptum; CON, adult C. onopordi; LY,
larva emerged from Centaurea solstitialis; LC,
larva emerged from Carthamus tinctorius).
Species
C. basicorne
C. scalptum
C. onopordi
Locality
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Horasan (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Askale (Erzurum, Turkey)
Morlupo, 31 km N of Rome, Italy
Morlupo, 31 km N of Rome, Italy
Morlupo, 31 km N of Rome, Italy
ID no.
CB2.1
CB2.6
CB3.6
CB4.1
LY6
CB2.15
CS1.2
CS2.1
CS2.2
CS2.3
LC5
LC8
LC9
LC10
CON1
CON2
CON3
395629.8N, 403520.1E; Horasan, 40735.9N,
422953.2E) sites were collected and preserved in
99.9% ethanol. The adults were identified using morphology. One larva dissected from C. solstitialis (LY6)
and four larvae dissected from safflower (labelled LC)
(Askale site) were collected and preserved in 99.9%
ethanol. Larval identity was afterwards confirmed by
comparing their sequences to the adult sequences. Three
adults of Centaurea onopordi were chosen as an out-
group and included in subsequent analyses (Table 1).
Genetic analyses for P. chalcomera were carried
out on 58 morphologically identified adults killed in
acetone. Over-wintering adults were collected on YST
and Onopordum sp. in Russia (at sites near Krasnodar,
45755.5N, 364029.6E) and on Onopordum and
C. nutans in Italy (at sites near LAquila, 4235N,
1339E, and Rome, 4189N, 1250E, respectively).
DNA extraction, amplification
and sequencing
Specimens were dried at 37C in 1.5 ml microcentrifuge tubes, then left for 10 min at 20C and, after the
addition of 5 ml of a solution of proteinase K (20 mg/
ml) with 200 l of lysis-digestion buffer (PK buffer:
10 mM EDTA 0.5 M, 100 mM Tris 1 M pH 7.5, 300
mM NaCl 3 M, 2% sodium dodecyl sulphate), were
homogenized using a pestle and incubated at 65C
for at least 3 h. The mixture was extracted with 25:24
phenol/chloroform or 25:24:1 phenol/chloroform/isoamyl alcohol protocols (Sambrook and Russell, 2000),
and the DNA was precipitated in two volumes of 95%
EtOH and 1/10 volume of 3 M sodium acetate. The
DNA was then pelleted, washed once with 80% EtOH
and resuspended in 50 ml of Tris-EDTA buffer 1, pH
7.5. Larvae of C. basicorne were washed and ground
in pure water. An aliquot of 2 ml of this suspension was
used as the template for direct polymerase chain reaction (PCR; without preliminary DNA extraction). A
part of the COI gene, 726 bp at 3 end for C. basicorne,
and 510 bp starting at the 5 end for P. chalcomera, was
amplified and sequenced using universal C1-J-2183
and TL2-N-3014 for C. basicorne and N2-J-1006 for
P. chalcomera (Simon et al., 1994; Antonini et al., unpublished data) and specific primers [C1-SQ-2011(-)]
(De Biase et al., unpublished data). We used a Perkin
Elmer GeneAmp PCR System 2400 thermal cycler and
the following amplification conditions for both species:
95C denaturation (1 min), 55C annealing (30 s) and
72C extension (1 min) for 33 cycles, followed by 7 min
elongation step at 72C. The PCR products were purified using ExoSAP-IT (USB Corporation 2000) and
sent to an external sequencing service (BMR-Genomics,
Padova, Italy).
Data analysis
All sequences were aligned by using the Staden
Package v. 1.6.0 (Bonfield et al., 2005). Phylogenetic
analyses were performed using the software packages
PAUP* v. 4.0b10 (Swofford, 2001) and MEGA v. 3.1
(Kumar et al., 2004).
Results
The COI sequence data set from the three species consisted of 726-bp aligned positions. There was almost no
haplotype diversity within the population representative
of each of the three species. However, a high genetic divergence (11.3%) was observed between C. basicorne
and C. scalptum, which was of the same range as observed between two differentiated species C. scalptum
and C. onopordi (Table 2). The tree (Figure 1) shows
three clusters strongly supported at 100% bootstrapping confidences. A fragment of nearly 1000 bases was
amplified for 58 individuals of P. chalcomera, and 510
aligned positions were used for the analyses. Tamura
and Nei (1993) genetic distances (TN) were computed
Table 2.
Genetic distances (p distances) based on COI se
quence data between three species of Ceratapion.
C. scalptuma
C. scalptum
C. basicorne 0.113 (0.011)d
C. onopordi 0.139 (0.012)
C. basicorneb
C. onopordic
0.144 (0.012)
All adults of C. scalptum and larvae dissected from Carthamus
tinctorius.
b
All adults of C. basicorne and larva dissected from Centaurea
solstitialis.
c
All adults of C. onopordi.
d
Standard errors, in brackets, are estimated by bootstrap method
(replications = 1000 and random number seed = 75,349)
a
265
XII International Symposium on Biological Control of Weeds
LC5-C1-3014RC.ab1
CS2.3-C1-3014RC.ab1
CS2.2-3014RC.ab1
100
CS1.2-3014RC.ab1
LC8-C1-3014RC.ab1
LC9-C1-3014RC.ab1
CS2.1-3014RC.ab1
100
LC10-C1-3014RC.ab1
CB2.6-3014RC.ab1
100
CB3.6-C1-3014RC.ab1
CB4.1-3014RC.ab1
65
LY6-C1-3014RC.ab1
CB2.1-3014RC.ab1
CB2-15-3014RC.ab1
100
CON1RC.ab1
CON2RC.ab1
CON3RC.ab1
0.01
Figure 1.
Figure 2.
NJ tree from the analysis of COI sequence data showing phylogenetic relationships among adult and larval specimens of three Ceratapion species (CB, adult
C. basicorne; CS adult C. scalptum; CON, adult C. onopordi; LY, larva emerged
from Centaurea solstitialis; LC, larva emerged from Carthamus tinctorius).
Unrooted NJ tree based on Tamura and Nei distances from COI sequences showing some differentiation of
Psylliodes chalcomera into genetically distinct populations that are associated with different host plants
[YST, yellow starthistle; ONO, Onopordum sp. (Scotch thistle); ONI, O. illyricum; CAR, Carduus nutans
(musk thistle)].
266
The importance of molecular tools in classical biological control of weeds
among all pairs of sequences (table not shown; values
ranged from 0.000 to 0.058). Genetic relationships
among populations from different host plants and locations were inferred by tree construction using neighbour joining (NJ) of Tamura and Neis (1993) genetic
distance measure. The tree (Figure 2) shows two clusters strongly supported at 98% and 97% bootstrapping
confidences. One cluster united haplotypes from Italy
and Russia on YST, ONO and ONI and the second
united haplotypes from Russia on ONO only. Another
cluster supported at 89% confidence was formed with
Russian populations from ONO and from YST.
Discussion
C. solstitialis is one of the most important weeds in
western USA and is a target of a classical biological
control programme that started in the 1960s. Despite
the introduction of several seed-head insects, the plant
is not yet under control over most of its range (Pitcairn
et al., 2004). Recently, more effort has been placed in
the discovery, selection and assessment of biological
control candidates attacking the weed during early phenological stages because it is expected that stressing
the immature plants may increase mortality or reduce
the number of flower heads available for attack by the
established flower-head insects (Smith, 2004; Uygur,
2004). C. basicorne and P. chalcomera are the most
promising natural enemies attacking immature YST.
As field observations, collections and open field
tests suggested the presence of sibling and closely related species for P. chalcomera and C. basicorne, respectively, we combined classical biological control
strategies with the application of molecular investigations. In this paper, we have demonstrated that DNA
markers are a powerful tool to better understand the
system in which biological control of YST is being
attempted. Molecular analyses allowed identification
of immature insect stages that were recovered from
YST and safflower test plants as C. basicorne and C.
scalptum, respectively and exclusively. The ability to
identify the species of immature insects in the field
experiments was crucial in successfully evaluating the
risk of introducing C. basicorne in the USA (Smith,
2006). Although relationship between the populations
could not be definitively established in this preliminary
study, the phylogeographic pattern of P. chalcomera
s.l. gave evidence of two haplotype groups of Russian
specimens from ONO and from YST. Other haplotypes
that were distributed both in Italy and Russia and found
on different host plants grouped together (Figure 2).
Our findings suggest different hypotheses regarding
the relationships of these genetic pools. Russia (ONO)
and Russia (YST) seem to be distinct gene pools well
characterized and disjunct [Russia (YST) vs Russia
(ONO 2)TN = 0.017]; however, incomplete lineage sorting or even hybridization events cannot be excluded,
e.g. Russia (YST) vs Russia (ONO 1)TN = 0.008 and
Russia (ONO 1) vs Russia (ONO 2)TN = 0.017. On the
other hand, the structure of the first group [Italy (CAR)]
of clustering individuals from two very distant locations (Russia and Italy) and feeding on three distinct
host plants (YST, CAR and ONI) seems to suggest the
existence of generalists shifting among different host
plants. The shifted forms could eventually evolve in local specialized entities like Russia (ONO). However,
mechanisms driving those shifts and their stability are
still unclear.
In conclusion, available data still suggest the existence of distinct genetic entities within the Psylliodes
spp. cf. chalcomera taxon, which are not distinguishable by morphological traits. Such findings seem also
to reflect a trophic specialization at least at local level
with two of them feeding on a single host plant (Cristofaro et al., 2004). We will develop behavioral and
non-lethal genetic assays to identify individuals of the
Russian YST-specific population before they can be released (Fumanal et al., 2005). Furthermore, because a
Carduus-specific population has already been released
in the USA, it will be important to establish if the YSTspecific population can hybridize with it and to assess
the effect on host specificity of the progeny. The development of microsatellites will be essential to identify
hybrid individuals and male gene flow and to finally
solve this interesting applied and taxonomic problem
of P. chalcomera.
Furthermore, these two examples clearly show
the importance of combining the classical biological
control approach with molecular methods as a tool to
discriminate between species and populations and to
identify immature insects to ensure the selection of the
appropriate agents. In spite of their power, the use of
molecular techniques is often limited by the lack of
time, costs and skills. We believe that this problem can
be overcome by cooperation between groups with different expertise, and this will improve our ability to discover safe effective biological control agents. Because
of the existence of cryptic species of insects, molecular
genetic tools enable us to identify host-specific populations from insect species previously thought to be
more polyphagous. This may greatly increase the potential number of prospective biological control agents
that exist in nature.
Acknowledgements
We thank G. Coletti, L. Serrani and M. Trizzino for their
precious work in the laboratory and E. Colonnelli for
advice and helpful and fast identifications of C. onopordi specimens. We also thank M. Yu. Dolgovskaya,
S. Reznik and M. Volkovich (ZIN-RAS, Saint Petersburg, Russia) for their fundamental work with P. chalcomera. We thank R. Hayat, L. Gltekin and H. Zengin
(Plant Protection Department, Atatrk University of Erzurum, Turkey), without whom the field tests in Turkey
could not have been done. A special thanks to C. Tronci
267
XII International Symposium on Biological Control of Weeds
(BBCA) for his cooperative suggestions during the
whole work and helpful comments on the manuscript.
References
Bonfield, J., Beal, K., Jordan, M., Cheng, Y. and Staden, R.
(2005) The Staden Package Manual. Medical Research
Council, Laboratory of Molecular Biology, Cambridge,
UK.
Briese, D.T. (1996) Phylogeny: can it help us to understand
host-choice by biological control agents? In: Moran, V.C.
and Hoffmann, J.H (eds) Proceedings of the IX International Symposium on Biological Control of Weeds. University of Cape Town, Stellenbosch, South Africa, pp.
6370.
Campobasso, G., Colonnelli, E., Knutson, L., Terragitti, G.
and Cristofaro, M. (1999) Wild plants and their associated insects in the Palearctic region, primarily Europe
and the Middle East. US Department of Agriculture, Agricultural Research Service, ARS-147, Washington, DC,
USA, 243 pp.
Caterino, M.S., Cho, S. and Sperling, F.A.H. (2000) The current state of insect molecular systematics: a thriving Tower
of Babel. Annual Review of Entomology 45, 154.
Crawley, M.J. (1989) The successes and failures of weed biocontrol using insects. Biocontrol News and lnformation
10, 213223.
Cristofaro, M., Hayat, R., Gltekin, L., Tozlu, G., Zengin,
H., Tronci, C., Lecce, F., Sahin, F. and Smith, L. (2002)
Preliminary screening of new natural enemies of yellow
starthistle, Centaurea solstitialis L. (Asteraceae) in Eastern Anatolia. In: zbek, H., Gl, S. and Hayat, R. (eds)
Proceeding of the Fifth Turkish National Congress of Biological Control. Atatrk University Publication, Erzurum,
Turkey, pp. 287295.
Cristofaro, M., Dolgovskaya, M. Yu., Konstantinov, A., Lecce,
F., Reznik, S. Ya., Smith, L., Tronci, C. and Volkovitsh,
M.G. (2004) Psylliodes chalcomera Illiger (Coleoptera:
Chrysomelidae: Alticinae), a flea beetle candidate for
biological control of yellow starthistle Centaurea solstitialis. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lons
dale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings
of the XI International Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia,
pp. 7580.
Cruttwell-McFadyen, R.E. (2000) Successes in biological
control of weeds. In: Spencer, N.R. (ed.) Proceedings of
the X International Symposium on Biological Control of
Weeds, Montana State UniversityBozeman, Bozeman,
MT, pp. 314.
De Biase, A., Antonini, G. and Audisio, P. (2003) Genetic
analyses on taxonomic status and mtDNA variation in
natural populations of Psylliodes spp. cfr. chalcomera
(Coleoptera, Chrysomelidae, Alticinae). In: Cristofaro,
M. and Tronci, C. (eds) Technical Annual Report, Biotechnology and Biological Control Agency ONLUS, Sacrofano (Rome), Italy, p. 54. Available as PDF at: http://
www.bbca.it/docs/BBCA0304.pdf
Dunn, P.H. and Campobasso, G. (1993) Field test of the weevil Hadroplonthus trimaculatus and the flea beetle Psylliodes chalcomera against musk thistle (Carduus nutans).
Weed Science 41, 656663.
Dunn, P.H. and Rizza, A. (1976) Bionomics of Psylliodes
chalcomera, a candidate for biological control of musk
thistle (Carduus nutans). Annals of the Entomological Society of America 69, 395398.
Ehler, L.E., Sforza, R. and Mateille, T. (2004) Genetics, Evolution and Biological Control. CABI Publishing, Wallingford, UK, 288 pp.
Fumanal, B., Martin, J.F. and Bon, M.C. (2005) High throughput characterization of insect morphocryptic entities by
a non-invasive method using direct-PCR of fecal DNA.
Journal of Biotechnology 119, 1519.
Gruev, B. and Dberl, M. (1997) General distribution of the
flea beetles in the Palaearctic Subregion (Coleoptera,
Chrysomelidae: Alticinae). Scopolia 37, 298.
Gruev, B. and Dberl, M. (2005) General distribution of the
flea beetles in the Palaearctic Subregion (Coleoptera,
Chrysomelidae: Alticinae). Supplement Pen soft Publishers, Sofia-Mosca, 239 pp.
Hebert, P.D.N., Cywinska, A., Ball, S.L. and deWaard, J.R.
(2003) Biological identifications through DNA barcodes.
Proceedings of the Royal Society of London, Series B,
Biological Sciences 270, 313321.
Julien, M.H. (1989) Biological control of weeds worldwide:
trends, rates for success and the future. Biocontrol News
and Information 10, 299306.
Kumar, S., Tamura, K. and Nei, M. (2004) MEGA3: Integrated
software for molecular evolutionary genetics analysis
and sequence alignment. Briefings in Bioinformatics 5,
150163.
Maddox, D.M., Mayfield, A. and Poritz, N.H. (1985) Distribution of yellow starthistle (Centaurea solstitialis) and
Russian knapweed (Centaurea repens). Weed Science 33,
315327.
Moritz, C., Dowling, T.E. and Brown, W.M. (1987) Evolution of animal mitochondrial DNA: relevance for population biology and systematics. Annual Review of Ecology
and Systematics 18, 269292.
Piper, G.L. and Nechols, J.R. (2004) Psylliodes chalcomera.
In: Coombs, E.M., Clark, J.K., Piper, G.L. and Cofrancesco, A.F., Jr (eds). Biological Control of Invasive Plants
in the United States.. Oregon State University Press, Corvallis, OR, pp. 362363.
Pitcairn, M.J., Piper, G.L. and Coombs, E.M. (2004) Yellow
starthistle. In: Coombs, E.M., Clark, J.K., Piper, G.L. and
Cofrancesco, A.F., Jr (eds) Biological Control of Invasive Plants in the United States,. Oregon State University
Press, Corvallis, OR, pp. 421435.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, J.
(2006) Yellow starthistle continues its spread in California. California Agriculture 60, 8390.
Sambrook, J. and Russell, D. (2000) Molecular Cloning. A
laboratory Manual, 3rd edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Simon, C., Frati, F., Breckenbach, A., Crespi, B., Liu, H. and
Flook, P. (1994) Evolution, weighting, and phylogenetic
utility of mitochondrial gene sequences and a compilation
of conserved polymerase chain reaction primers. Annals
of the Entomological Society of America 87, 651701.
Smith, L. (2004) Prospective new agents for biological control of yellow starthistle. In: Proceedings 56th Annual
California Weed Science Society, 1214 January, 2004,
Sacramento, California. California Weed Science Society,
Salinas, CA, pp. 136138.
268
The importance of molecular tools in classical biological control of weeds
Smith, L., Cristofaro, M., Dolgovskaya, R.Yu., Tronci, C. and
Hayat, R. (2005) Status of new agents for biological control of yellow starthistle and Russian thistle. In: Skurka,
G. (ed) Proceedings California Invasive Plant Council
Symposium, vol. 9, California Invasive Plant Council,
Berkeley, California, October 68, 2005. Chico State University, Chico, CA, pp. 2226.
Smith, L. (2006) Risk assessment of Ceratapion basicorne, a
rosette weevil of yellow starthistle. In: Hoddle, M.S. and
Johnson, M.W. (eds) Proceedings of the Fifth California Conference on Biological Control, 25-27 July 2006.
Center for Biological Control, University of California,
Berkeley, CA, pp. 4754.
Smith, L. (2007) Physiological host range of Ceratapion
basicorne, a prospective biological control agent of Centaurea solstitialis (Asteraceae). Biological Control 41,
120133.
Smith, L. and Drew, A.E. (2006) Fecundity, development
and behavior of Ceratapion basicorne (Coleoptera: Apionidae), a prospective biological control agent of yellow
starthistle. Environmental Entomology 35, 13661371.
Smith, L., Hayat, R., Cristofaro, M., Tronci, C., Tozlu, G. and
Lecce, F. (2006) Assessment of risk of attack to safflower
by Ceratapion basicorne (Coleoptera: Apionidae), a prospective biological control agent of Centaurea solstitialis
(Asteraceae). Biological Control 36, 337344.
Swofford, D.L. (2001) PAUP*. Phylogenetic Analysis Using
Parsimony (*and Other Methods). Version 4b10. Sinauer
Associates, Sunderland, MA.
Tamura, K. and Nei, M. (1993) Estimation of the number of
nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10, 512526.
Unruh, T.R. and Wooley, J.B. (1999) Molecular methods in
classical biological control. In: Bellows, T.S. and Fisher,
T.W. (eds) Handbook of Biological Control: Principles
and Applications of Biological Control. Academic Press,
San Diego, CA, pp. 5785.
Uygur, S. (2004) Density of Centaurea solstitialis L. and its
natural enemies Ceratapion spp. in southern Turkey. Turkish Journal of Agriculture and Forestry 28, 333339.
Uygur, S., Smith, L., Uygur, F.N., Cristofaro, M. and Balciunas, J. (2005) Field assessment in land of origin of host
specificity, infestation rate and impact of Ceratapion basicorne a prospective biological control agent of yellow
starthistle. Biocontrol 5, 525541.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
Wolstenholme, D.R. (1992) Animal mitochondrial DNA:
structure and evolution. International Review of Cytology
141, 173216.
269
Fungal pathogens of Schinus
terebinthifolius from Brazil as potential
classical biological control agents
A.B.V. Faria,1 R.W. Barreto1 and J.P. Cuda2
Summary
The Brazilian peppertree, Schinus terebinthifolius Raddi, is a shrub or small tree, native to Brazil,
Paraguay and Argentina. It has been introduced into other regions of the world as an ornamental or as
a source of condiment. It became an aggressive invader of several exotic ecosystems, particularly in
oceanic islands such as Hawaii and Mauritius, as well as in Florida, USA, and Australasia. Although
research involving the use of insect natural enemies from the plants centre of origin as biological
control agents for S. terebinthifolius has been going on for some time, no systematic survey had been
undertaken for fungal pathogens until recently. A study of the mycobiota associated with this plant
was initiated in 2001, which concentrated on the southeastern states of Brazil. Eleven fungal taxa
have been found thus far, several of which are new to science, namely Hainesia lythri (Desmaz.)
Hhn., Irenopsis sp., Meliola sp., Oidium sp., Phyllosticta sp. nov., Pleomassaria sp. nov., Pilidium
concavum (Desmaz.) Hhn., Pseudocercospora sp. nov., Septoria sp. nov., Stenella sp. and one new
coelomycete genus. These fungi were associated with various symptoms, viz. leaf spots, black mil
dew and powdery mildew. The Septoria leaf-spot fungus appears to have particularly good potential
as a classical biological control agent. Pathogenicity has been demonstrated to biotypes of S. terebinthifolius from Brazil, Florida and Hawaii. Its host range is now being tested in Florida and Hawaii
with the goal of possibly introducing the fungus into those areas.
Keywords: classical biological control, Brazilian peppertree, surveys.
Introduction
The Brazilian peppertree, Schinus terebinthifolius
Raddi, known in Brazil as aroeira, is a small tree in
the family Anacardiaceae native to Brazil, Argentina
and Paraguay (Elfers, 1988; Binggeli, 1997; Taylor,
1998; Cuda et al., 2006). In Brazil, S. terebinthifolius
is found along the east coast from the state of Pernam
buco in the north to Rio Grande do Sul in the south
(Lorenzi and Matos, 2002). It occurs in habitats rang
ing from sand dunes to rainforests and semi-deciduous
highland forests, often growing on river margins and in
swampy areas (Binggeli, 1997).
Schinus terebinthifolius is generally regarded as a
valuable plant in Brazil where it is used for medicinal
1
Universidade Federal de Viosa, Departamento de Fitopatologia, Vi
osa, MG, 36571-000, Brazil.
2
UF/IFAS Entomology and Nematology Department, Biological Weed
Control, Building 970, Natural Area Drive, PO Box 110620, Gaines
ville, FL 32611-0620, USA.
Corresponding author: R.W. Barreto <[email protected]>.
CAB International 2008
purposes (Lorenzi and Matos, 2002) and as a source
of tannin for treating fishing nets and fishing lines.
The wood is used for fencing, firewood and charcoal,
and the plant is also used as an ornamental and as a
source of forage for goats and is valued by beekeep
ers (Baggio, 1988; Lorenzi, 1992). In several parts of
the world, its fruits are used as a spice (pepper ros),
and there are also several other medicinal uses listed
for the plant (Cuda et al., 2006). However, S. terebinthifolius has become a noxious weed in many regions
of the world where it has been introduced, such as
Samoa, Fiji, French Polynesia, the Marshall Islands,
New Caledonia Mauritius, and the USA, particularly
Hawaii and Florida (Cronk and Fuller, 1995). It was
probably introduced into Florida before 1850 as an
ornamental plant (Mack, 1991). By the 1920s, it had
already become widely distributed, and in the 1960s, it
became recognized as an important component of the
natural vegetation forming dense monocultures (Mor
ton, 1978; Elfers, 1988; Binggeli, 1997; Anon, 2000,
2001a,b; Cuda et al., 2006). Presently, the distribution
of S. terebinthifolius in the USA includes central and
270
Fungal pathogens of Schinus terebinthifolius from Brazil as potential classical biological control agents
southern Florida, southern Arizona, southern Califor
nia, Texas, Louisiana, Hawaii and Puerto Rico (Anon,
2001a,b; Cuda et al., 2006); it also occurs in the Ba
hamas (Elfers, 1988; Anon, 2001a,b). In Florida, S.
terebinthifolius is common in areas where the soil is
dry to moderately well-drained along roadsides and
in the vicinity of lakes, and it is invading private and
public gardens (Anon, 2002). Although it is a common
pioneer of disturbed sites, such as abandoned farmland
and waste areas, it is also capable of invading wellpreserved natural areas, such as the drier areas of the
Everglades and the coastline of peninsular Florida.
Once established, S. terebinthifolius displaces native
herbaceous communities due to its dense shading habit
and the alleopathic substances it produces (Gogue et
al., 1974; Elfers, 1988; Binggeli, 1997; Anon, 2002;
Morgan and Overholt, 2005; Cuda et al., 2006). The
plant is readily dispersed by birds (Lorenzi, 1992; Pa
netta and McKee, 1997) and is capable of vigorous re
generation after fire, cutting or frost, making its control
particularly difficult (Binggeli, 1997). Invasions by S.
terebinthifolius can result in the loss of local biodiver
sity (Anon, 2002; Cuda et al., 2006).
Both chemical and mechanical controls of S. terebinthifolius have been adopted with some success in
Florida and Hawaii but only in areas that are cultivated
or otherwise intensively managed. Neither of those
strategies is appropriate for control in environmentally
sensitive natural areas, such as the Florida Everglades
(Anon, 1998; Anon, 2000; Cuda et al., 2006). For such
areas, biological control was recognized early on as
the ideal strategy for managing S. terebithifolius. The
initial search for insect natural enemies was led by en
tomologists who conducted surveys in Brazil and Ar
gentina (Hight et al., 2002). Several promising insects
that were found in association with S. terebinthifolius,
including a seed-feeding bruchid beetle, a stem-boring
moth and a leaf-rolling tortricid, were eventually intro
duced into Hawaii (Yoshioka and Markin, 1991). Later,
additional natural enemies were discovered by Bennett
et al. (1990) for possible introduction into Florida. A
stem-feeding thrips, Pseudophilothrips ichini Hood
(Thysanoptera: Phlaeothripidae), was regarded as a
particularly promising candidate for biological control
of S. terebinthifolius because it was found to be highly
host-specific and damaging to the flowers and young
shoots (Cuda et al., 2006). Its release in Florida was
approved by the US Governments Technical Advisory
Group for Biological Control Agents of Weeds (TAG)
in May 2007.
There are now numerous examples of the success
ful use of fungal pathogens as classical weed biologi
cal control agents (Charudattan, 1991; Watson, 1991;
Julien and White, 1997). Surveys for fungi associated
with important weeds native to Brazil, which were
initiated in the mid-1990s, have yielded a plethora of
fungi (e.g. Barreto and Evans, 1994, 1995a,b,c, 1998;
Barreto et al., 1995; Pereira and Barreto, 2005; Mon
teiro et al., 2003; Pereira and Barreto, 2005, Pereira
et al., 2007; Seixas et al., 2007) including many that
were new to science. Two such fungi have already been
introduced as biological control agents into other parts
of the world, namely Colletotrichum gloeosporioides
(Penz.) Sacc. f.sp. miconiae, which was introduced
into Hawaii and French Polynesia for the control of
Miconia calvescens D.C. (Seixas et al., 2007), and Prospodium tuberculatum (Speg.) Arthur, which was intro
duced into Australia for biological control of Lantana
camara L. (Ellison et al., 2006). Until recently, no sur
veys of the fungi-attacking S. terebinthifolius in Brazil
were performed, and very little information exists in
the literature about this plants mycobiota. A list of all
pathogenic fungi recorded from S. terebinthifolius is pre
sented in Table 1. This paper presents a preliminary ac
count of the first survey for fungal pathogens associated
Table 1.
271
Fungi recorded in association with Schinus
terebinthifolius worldwide.
Fungi
Anamorphic fungi
Alternaria sp.
Cercospora schini Syd.
Corynespora sp.
Diplodia sp.
Helminthosporium sp.
Macrophoma sp.
Phyllosticta sp.
Sphaeropsis tumefaciens
Hedges
Verticillium albo-atrum
Reinke & Berthier
Ascomycota
Botryosphaeria ribis
f. achromogena
Gross. & Duggar
Botryosphaeria ribis
Gross. & Duggar
Diaporthe sp.
Irenopsis coronata (Speg.)
F.L. Stevens
Meliola brasiliensis Speg.
Meliola coronata Speg.
Nectria cinnabarina
(Tode) Fr.
Seuratia millardetii
(Racib.) Meeker
Basidiomycota
Armillaria mellea (Vahl)
P. Kumm
Armillaria tabescens
(Scop.) Emel
Ganoderma orbiforme
(Fr.) Ryvarden
Distribution in association
with. S. terebinthifolius
USA (Farr et al., 1985)
Argentina (Chupp, 1953)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
South America
(Vigas, 1961)
South America
(Vigas, 1961;
Hansford, 1961)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
USA (Farr et al.,1985)
USA (Farr et al., 1985)
USA (Farr et al., 1985)
South America
(Vigas, 1961)
XII International Symposium on Biological Control of Weeds
with this plant in Brazil and an assessment of the po
tential of each species as a biocontrol agent for S. terebinthifolius.
Materials and methods
The area surveyed for fungal pathogens of S. terebinthifolius covered only a small part of southeastern Bra
zil and was concentrated in the municipality of Viosa
and neighbouring regions in the state of Minas Gerais
(MG), the municipality of Rio de Janeiro and north
ern coastal parts of the state of Rio de Janeiro (RJ) and
parts of the valley of the Rio Paraba do Sul in the state
of So Paulo. Pathogen collections were conducted
during the period from September 2001 to May 200.
Although the botanical literature recognizes five va
rieties of S. terebinthifolius (Barkley, 1944), only S.
terebinthifolius var. acutifolius, S. terebinthifolius var.
terebinthifolius, and S. terebinthifolius var. raddianus
are known to occur in the USA [only vars. terebinthifolius and raddianus occur in Florida (Cuda, 2002; Cuda
et al., 2006)]. Separation of such varieties in the field
in Brazil was complicated by the common occurrence
of intermediate forms. For practical reasons, all collec
tions were simply referred as being from S. terebinthifolius without identification of host variety. Due to the
difficulty in separating the varieties morphologically,
the Florida populations of S. terebinthifolius were re
cently characterized genetically using microsatellites
and chloroplast DNA sequence comparisons (Williams
et al., 2005).
Samples of diseased material were collected, dried
in a plant press and taken to the lab for further ex
aminations. Lesions were observed under a dissecting
microscope and slides were prepared and mounted in
lactophenol, lactophenol cotton blue or other mount
ing media, as required, and examined under a light mi
croscope (Olympus BX 50) fitted with a camera and
a drawing tube. Isolations were performed by either
directly transferring fungal structures from sporulat
ing lesions onto plates containing vegetable broth agar
(VBA), as described by Pereira et al. (2003), or by
transferring selected, surface-sterilized fragments of
diseased tissues onto VBA plates. Pure cultures were
kept in a refrigerator at 5C in tubes containing potato
carrot agar until further use or permanent deposition
in the culture collection at the Plant Pathology Depart
ment of the Universidade Federal de Viosa (Brazil).
Pathogenicity was tested for selected fungal spe
cies either by brush-inoculating young (7-month old)
healthy plants with a spore suspension adjusted to 2
106 spores per milliliter in a 0.05% Tween 80 solution
or by depositing onto healthy plant parts 5-mm diam
eter culture disks cut from actively growing mycelium
on VBA plates. For fungi suspected of being oppor
tunistic wound pathogens, healthy plant parts were
wounded with sterile scissors or needles before inocu
lation. Inoculated plants were kept in a dew chamber
at 25C for 48 h and then transferred to a greenhouse
where they were examined daily for evidence of dis
ease symptoms. Healthy plants treated in an equivalent
manner but not exposed to fungal inoculum served as
controls. Schinus terebinthifolius plants used in these
tests were grown from seed originating either from
Florida (Gainesville, Campus of the University of Flor
ida), Hawaii or Brazil (Viosa).
One fungal species collected and tested during these
surveys was selected for further study. It was a spe
cies of the genus Septoria, which was associated with
severe outbreaks of leaf spots in S. terebinthifolius
populations, followed by extensive defoliation. A pre
liminary assessment of the host range of this Septoria
sp. was performed with several local members of the
Anacardiaceae: Anacardium occidentale L. (cashew);
Mangifera indica L. (mango); Schinus molle L. (Pe
ruvian peppertree); Spondias lutea L. (Brazilian plum)
and Tapirira guianensis Aubl. (tatapiririca). Inocula
tions of the test plants were performed as described
above.
Results and discussion
In total, 11 fungal taxa were found attacking S. terebinthifolius during the surveys: three ascomycetes (Pleomassaria sp. possibly associated with stem dieback,
Irenopsis sp. and Meliola sp. associated with black
mildew symptoms); four anamorphic coelomycetes,
all of which were associated with leaf spot symptoms
[H. lythri (Desmaz.) Hhn. and P concavum (Desmaz.)
Hhn. (synanamorphic species), Phyllosticta sp., Septoria sp. and a possible new genus of coelomycete] and
four anamorphic hyphomycetes (Oidium sp., a pow
dery mildew, a Pseudocercospora sp. associated with
leaf spots and a Stenella sp. associated with yellowing
and premature leaf senescence; Table 2).
Eight of the 11 fungi found to be associated with
S. terebinthifolius were isolated in pure culture. Some
were extremely slow growing such as Pleomassaria
sp., which took nearly a month to produce a visible
colony, while others were relatively fast growing such
as H. lythri and P. concavum. The ability of the fungi
to sporulate also varied widely, but most fungi did not
sporulate readily in culture. Pathogenic status was only
preliminarily investigated for seven fungal species:
coelomycete gen. nov., H. lythri, P. concavum, Pseudocercospora sp., Septoria sp. and Stenella sp. Specific
details on each fungal taxon are provided below.
Coelomycete gen. nov.
This fungus was compared with a series of possible
genera of other coelomycetes to which it resembled
but did not match any of them. We concluded that it
represents a new fungal genus that will be described
in a separate publication on the taxonomy of the fungi
on S. terebinthifolius. This new fungus was restricted
272
Fungal pathogens of Schinus terebinthifolius from Brazil as potential classical biological control agents
Table 2.
Synopsis of observations on fungi collected on Schinus terebinthifolius during surveys in
Brazil.
Fungus
Disease
Purported
specificity
Uncertain
Culturability
Leaf spot
Damage
to host
Moderate
Coelomycete
(gen nov.)
Hainesia lythri
Irenopsis sp.
Meliola sp.
Oidium sp.
Phyllosticta sp.
Pleomassaria sp.
Pilidium concavum
Pseudocercospora sp.
Leaf spot
Black mildew
Black mildew
Powdery mildew
Leaf spot
Branch dieback
Leaf spot
Leaf spot
Insignificant
Insignificant
Insignificant
Moderate
Moderate
Uncertain
Insignificant
Severe
Non-specific
High
High
High
Uncertain
Uncertain
Non-specific
High
Cultivable
Not cultivable
Not cultivable
Not cultivable
Cultivable
Cultivable
Cultivable
Cultivable
Septoria sp.
Stenella sp.
Leaf spot
Leaf spot
Severe
Moderate
High
High
Cultivable
Cultivable
Cultivable
Biocontrol
potential
Uncertain
None
None
None
Moderate
Uncertain
Uncertain
None
None (see
comments)
Very high
Uncertain
in its distribution, being found in only one location on
plants growing on a hill in a coastal sand dune area
in the state of Rio de Janeiro. It was always associ
ated with a very distinctive circular leaf spot, and the
disease was regarded as severe on the plants it was at
tacking, resulting in significant defoliation. Its patho
genicity was proven by Kochs postulates, but disease
levels obtained under controlled conditions were less
severe than those observed in the field. Perhaps, the
fungus requires particular environmental conditions,
such as those occurring where it was repeatedly col
lected, to become epidemic and to cause severe disease
levels. It is difficult to determine at this stage whether
this fungus has any potential as a classical biological
control agent.
summer, but it is interesting to note that P. concavum
was associated with non-senescent leaves; this sug
gests that, in the case of S. terebinthifolius, P. concavum may act as a true pathogen. Isolations of H. lythri
from diseased tissues resulted in typical H. lythri colo
nies in pure culture, while isolations of P. concavum
also resulted in typical Hainesia colonies in culture.
This has already been reported for isolates from other
hosts (Shear and Dodge, 1921; Palm, 1991), and it is
accepted that H. lythri and P. concavum are genetically
connected by being synanamorphs of the same asco
mycete species, Discohainesia oenotherae (Cooke &
Ellis) Nannf. (Palm, 1991). Although interesting and
a new report for S. terebinthifolius, these two records
have no significance for biological control.
H. lythri and P. concavum
Irenopsis sp. and Meliola sp.
These fungi are known to be opportunistic patho
gens, depending on wounds for infection; this study
confirmed this relationship. Only wounding of tis
sues before inoculation allowed for disease to develop
during the pathogenicity tests. These two species are
known to be generalists having been described from
a wide range of hosts (Shear and Dodge, 1921; Palm,
1991). Among these are several species belonging to
the Anacardiaceae, e.g., Rhus glabra L., Rhus typhina
L., Toxicodendrun radicans (L.) Kuntze and Rhus aromatica Ait. (Greene, 1950). This information alone
would make these fungi of little interest for classical
biological control because they might pose a threat
to native plants in the area where introduction would
take place. P. concavum has been frequently reported
in association with old lesions of H. lythri, but the two
species are normally not found simultaneously on the
same lesions or during the same season. However,
this association is not observed in all hosts (Shear and
Dodge, 1921; Palm, 1991). In one study, H. lythri and
P. concavum were found occurring together during the
These are two ascomycetes belonging to the family
Meliolaceae, a family of fungi that is widely distributed
in the tropics and contains highly host-specific, obligate
parasites that cause diseases known as black mildews
(Hansford, 1961; Kirk et al., 2001). Irenopsis sp. and
Meliola sp. often were found attacking the same S. terebinthifolius individuals in the field and sometimes oc
curring together on the same leaf. Their colonies were
indistinguishable with the naked eye. A comparison of
the morphology of these fungi with known black mil
dew species associated with the Anacardiaceae led to
the conclusion that both species on S. terebinthifolius
are new to science. These will be described in a separate
publication. Although pathogenic, diseases caused by
the Meliolaceae are generally considered to be too weak
to be of any interest for use in weed biological control.
Oidium sp.
This fungus is an anamorphic form from the Erysi
phaceae, an important family of ascomycetes that are
273
XII International Symposium on Biological Control of Weeds
obligate biotrophs causing powdery mildews. Fungi in
this group can be highly host-specific and cause heavy
losses in some crop plantations. Unfortunately, in the
case of this fungus, it was found only on a few occa
sions, always during the dry and colder season of the
year, and damage to S. terebinthifolius was regarded
as only moderate. Perhaps, further studies will reveal
more aggressive strains of this fungus that may be of
interest to be used in classical introductions. The com
plete identity of this fungus remains obscure at this
stage. A preliminary DNA analysis surprisingly placed
this taxon close to Oidium tuckeri Berk., the aetiologi
cal agent of the powdery mildew of grapes, than to the
Erysiphales known to attack the Anacardiaceae.
branches, particularly the lower ones. The aetiology of
this disease has thus far been elusive. Isolations from
inner parts of diseased branches usually yield a range
of sterile fungi, which do not yield disease symptoms
when inoculated onto healthy plants. Pleomassaria was
thought to perhaps be the aetiological agent involved
in such diebacks. The fungus was isolated in pure cul
ture but grew extremely slowly and did not sporulate.
Nevertheless, work on this fungus should be continued
to clarify its possible involvement in this spectacular
disease.
Pseudocercospora sp.
Phyllosticta is a genus that contains some important
pathogens of crop plants (Nag Raj, 1992), but there
also are saprophytes and endophytes in this genus. The
species of Phyllosticta recorded on the Anacardiaceae
are difficult to separate taxonomically and are, in gen
eral, poorly described in earlier publications. These
are Phyllosticta schini Thm, Phyllosticta rhois West.,
Phyllosticta rhoina Kalch., Phyllosticta toxicodendri
Thm and Phyllosticta toxica Ell. & G. Martin (Sac
cardo, 1884, 1902). A precise comparison with the
fungus found on S. terebinthifolius, based on such de
scriptions, was impossible. A review of the genus was
published by van der Aa and Vanev (2002), and these
authors considered that all the taxa listed above, except
P. toxica, should be excluded from Phyllosticta. It is
likely that the species collected in this survey represents
a new taxon for the genus, but further investigation is
required to confirm this. Obtaining an isolate of Phyllosticta sp. from S. terebinthifolius proved difficult. An
isolate was finally obtained but in the latter stages of
this work, which did not allow for confirmation of its
pathogenicity. The leaf-spot disease on S. terebinthifolius with which Phyllosticta sp. was consistently asso
ciated was often severe. Further studies on this fungus
as a possible biological control agent should be given
high priority.
Until this study, there were no members of the genus
Pseudocercospora known to infect S. terebinthifolius
or any other species in this genus. We found two kinds
of Pseudocercospora spp. associated with leaf spots on
S. terebinthifolius. They have distinct morphological
features and may deserve to be treated as separate spe
cies, to be named and described in a separate publica
tion. In terms of disease symptoms produced by these
two species on S. terebinthifolius, they were essentially
indistinguishable. Pathogenicity to S. terebinthifolius
was demonstrated for one isolate. Because isolates of
the two fungi did not sporulate in culture, Kochs pos
tulates were performed with culture disks serving as in
oculum. Disease symptoms produced with this method
were not very severe and were different from those
observed in the field. Pseudocercospora leaf spot was
one of the most common diseases of S. terebinthifolius
in the surveyed areas in Brazil and sometimes led to
significant levels of defoliation. Unfortunately, it ap
pears that Pseudocercospora spp. have no potential for
classical biological control, as on two separate visits
to Florida, one of us (RWB) collected leaf spots in the
Everglades and near the town of Plantation, bearing
Pseudocercospora colonies. Unless an especially viru
lent Pseudocercospora strain is obtained from Brazil,
introductions of this fungus would probably be super
fluous and innocuous, at least in Florida, as the fungus
is already present in areas where S. terebinthifolius is a
problem but is not reducing infestations.
Pleomassaria sp.
Septoria sp.
This fungus is a new species for this ascomycete ge
nus, which will be described separately. Its pathogenic
status is still uncertain. The fungus was found only
once during examinations of branches showing dieback
symptoms (in Guaraciaba, state of Minas Gerais). Such
diebacks are very commonly observed in the field in
Brazil and are often very debilitating to the host plants.
We have observed that most of the striking differences
between healthy versus unhealthy plants when they are
growing in exotic situations, such as in Florida, com
pared to plants from the native range, such as in Brazil,
are the result of this dieback of a large proportion of
This fungus was compared with other members of
the genus Septoria described in the literature on mem
bers of the Anacardiaceae and is clearly distinct. It
represents another new taxon discovered during this
survey to be fully described in a separate publication.
Although it grew slowly, Septoria sp. sporulated abun
dantly on VBA. Its pathogenicity to S. terebinthifolius
was demonstrated, and abundant lesions formed and
coalesced leading to substantial defoliation of inocu
lated plants. A preliminary host-range study performed
with this fungus indicated that it is highly host-spe
cific, which is often observed for other members of
Phyllosticta sp.
274
Fungal pathogens of Schinus terebinthifolius from Brazil as potential classical biological control agents
Septoria as discussed in the review by Priest (2006).
Fungi in this genus already have been used for classical
biological control of weeds. Three examples involved
introductions of species of Septoria into the USA, all
coincidentally in Hawaii. Septoria passiflorae Syd.
was introduced from Colombia in 1995 for biological
control of Passiflora tarminiana Coopens (=Passiflora
mollissima, Passiflora tripartita, P. tripartita var. tripartita) (Norman, 1995). Another Septoria sp. was
introduced from Ecuador as a biological control agent
against L. camara L. in 1993 (Trujillo and Norman,
1995). Septoria hodgesii Gardner was regarded by
Gardner (1999) to have potential for biological control
of Myrica faya (Ailton) Wilbur. It was also introduced
in Hawaii in 1997 but did not establish, probably due to
unsuitable environmental conditions at the release sites
(E. Killore, personal communication). Excellent con
trol of L. camara and P. tarminiana was reported after
the introductions of Septoria spp. against these weeds
(Trujillo, 2005). Likewise, Septoria sp. collected on S.
terebinthifolius appears to have good potential as a clas
sical biological control agent. It not only caused con
siderable damage through defoliation of infected plants
in the field in Brazil but was found to be pathogenic
to plants grown from seeds of S. terebinthifolius from
Hawaii and Florida. More importantly, it appears to be
host-specific, as it did not infect any of the other five
species of Anacardiaceae (i.e. cashew, mango, Peruvian
peppertree, Brazilian plum and tatapiririca) included in
the preliminary host-range test performed during this
study. Isolates of this fungus are now under additional
evaluation in approved quarantine laboratories located
in Hawaii (HDOA-Biological Control Labs, Honolulu)
and in Florida (FLDACS, DPI Pathogen Containment
Laboratory, Gainesville).
Stenella sp.
Most members of the genus Stenella are plant
pathogens causing leaf-spot diseases. There are over 20
species described in the literature (Kirk et al., 2001),
but none was described in association with S. terebinthifolius or any other member of the Anacardiaceae.
This fungus appears to be a new taxon of cercosporoid
fungus, also to be described later. Observations in the
field strongly indicated that this is a pathogenic fun
gus that forms extensive brown colonies on adaxial
leaf surfaces, accompanied by abaxial yellowing and
premature dropping of infected leaves. Unfortunately,
pathogenicity was not proven during attempts to fulfil
Kochs postulates. One possibility is that the use of cul
ture disks of this fungus as inoculum was inadequate
for that purpose or that an incompatible combination
of fungal isolate and host genotype led to such a fail
ure. This is, therefore, still considered in this study as
an unresolved issue, and the subject of its potential for
biological control of S. terebinthifolius will be pursued
in a subsequent study.
The list of fungi already recorded in association with
S. terebinthifolius (Table 1) contained 16 records from
the USA, but only four from the centre of origin of the
plant in the Neotropics. This survey, although prelimi
nary and covering a small part of the native distribution
of S. terebinthifolius, raised the number of fungi known
to attack this species in the Neotropics to 14. This type
of result is not uncommon and was already observed for
other weeds native to Brazil such as L. camara (Barreto
et al., 1995) and Chromolaena odorata (L.) R.M. King
& H. Rob. (Barreto and Evans, 1994). After becoming
invasive in new, exotic situations, these plants become
ubiquitous and an abundant substrate for saprophytic or
generalist fungi. Fungal collections become frequent,
and many fungalhost associations are then described
and published. The majority of the records from the
USA are probably explained by this approach. For
instance, fungal species such as Armillaria mellea,
Armillaria tabescens, Ganoderma orbiforme, Botryosphaeria ribis, and Nectria cinnabarina (Table 1) are
well-known generalist pathogens. Conversely, fungi
already recorded from the Neotropics or those newly
recorded in this study are (with the clear exception
of H. lythri and P. concavum) likely to be specialized
host-specific pathogens.
Acknowledgements
This work forms part of a research project submitted
as a MSc dissertation to the Departamento de Fitopa
tologia/Universidade Federal de Viosa by A. B. V.
Faria. The authors thanks the Conselho Nacional de
Desenvolvimento Cientfico e Tecnolgico (CNPq),
Coordenao de Aperfeioamento de Pessoal de Nvel
Superior (CAPES) and the Fundao de Amparo Pes
quisa do Estado de Minas Gerais (FAPEMIG) for fi
nancial support. This research work was partly funded
by grants awarded to the University of Florida by the
Florida Department of Environmental Protection and
the South Florida Water Management District.
References
Aa, H.A. van der and Vanev, S. (2002) A Revision of the Species Described in Phyllosticta. CBS, Utrecht, The Neth
erlands, 510 pp.
Anon. (1998) UFL University of Florida. Chemical Con
trol. In: Current Control Technologies. Available at: http://
aquat1.ifas.ufl.edu (accessed 7 December 2001).
Anon. (2000) IPIF Institute of Pacific Island Forestry. Pa
cific Island Ecosystems at Risk (Pier). Available at: http://
www.hear.org/pier3/scteer.htm (accessed 7 December
2001).
Anon. (2001a) CCNRD Collier County Natural Resourses
Department. Brazilian Peppertree (Schinus terebinthifolius).
In: Exotic Plant Management. Available at: http://www.co.
collier.fl.us/natresources/exotics/bp.htm (accessed 7 De
cember 2001).
275
XII International Symposium on Biological Control of Weeds
Anon. (2001b) RCNJ Ramapo College of New Jersey.
Schinus terebinthifolius. Available at: http://www.orion.
ramapo.edu (accessed 7 December 2001).
Baggio, A.J. (1988) Aroeira como potencial para usos mlti
plos na propriedade rural. Boletim de Pesquisa Florestal
17, 1324.
Barkley, F.A. (1944) Schinus L. Brittonia 5, 160198.
Barreto R.W. and Evans H.C. (1994) The mycobiota of the
weed Chromolaena odorata in southern Brazil with par
ticular reference to fungal pathogens for biological con
trol. Mycological Research 98, 11071116.
Barreto R.W. and Evans H.C. (1995a) The mycobiota of the
weed Mikania micrantha in southern Brazil with particu
lar reference to fungal pathogens for biological control.
Mycological Research 99, 343352.
Barreto R.W. and Evans H.C. (1995b) Mycobiota of the weed
Cyperus rotundus in the state of Rio de Janeiro, with an
elucidation of its associated Puccinia complex. Mycological Research 99, 407419.
Barreto R.W. and Evans H.C. (1995c) Fungal pathogens of
weeds collected in the Brazilian tropics and subtropics
and their biocontrol potential. In: Delfosse E.S. and Scott
R.R. (eds) Proceedings of the Eighth International Symposium on Biological Control of Weeds. DSIR-CSIRO,
Melbourne, Australia, pp. 679691.
Barreto, R.W. and Evans, H. (1998) Fungal pathogens of Euphor
bia heterophylla and E. hirta in Brazil and their potential
as weed biocontrol agents. Mycopathologia 141, 2136.
Barreto R.W., Evans H.C. and Ellison C.A. (1995) The myco
biota of the weed Lantana camara in Brazil, with particu
lar reference to biological control. Mycological Research
99, 769782.
Bennett, F.D., Crestana, L., Habeck, D.H. and Berti-Filho,
E. (1990) Brazilian peppertree prospects for biologi
cal control. In: Delfosse, E.S. (ed.) Proceedings VII. International Symposium on Biological Control of Weeds.
Ministero dellAgriculture e delle Foreste, Rome/CSIRO,
Melbourne, Australia, pp. 293297.
Binggeli, P. (1997) Schinus terebinthifolius Raddi (Anacardiaceae). In: Invasive Woody Plants. Available at:
http://members.lycos.co.uk/woodyPlantEcology/docs/
websp17.htm (accessed 7 December 2001).
Charudattan, R. (1991) The mycoherbicide approach with
plant pathogens. In: Tebeest, D.O. (ed.) Microbial Control
of Weeds. Chapman and Hall, New York, pp. 2457.
Chupp, C. (1953) A Monograph of the Fungus Genus Cerco
spora. Ithaca, New York, 667 pp.
Cronk, Q.C.B. and Fuller, J.L. (1995) Plant Invaders. Chap
man and Hall, London, UK, 241 pp.
Cuda, J.P. (2002) Proposed field release of Pseudophilothrips
ichini (Hood) (Thysanoptera: Phlaeothripidae), a nonin
digenous thrips from Brazil for classical biological con
trol of brazilian peppertree, Schinus terebinthifolius Raddi
(Sapindales: Anacardiaceae). Internal Report. Entomology and Nematology Department, University of Florida,
Gainesville, FL, 52 pp.
Cuda J.P., A.P. Ferriter, V. Manrique and J.C. Medal (eds)
(2006) Floridas Brazilian peppertree management plan,
2nd edn. Recommendations from the Brazilian peppertree
Task Force, Florida Exotic Pest Plant Council. Available
at: http://ipm.ifas.ufl.edu/reports/BPmanagPlan.pdf.
Ellison C.A., Pereira J.M., Thomas S.E., Barreto R.W. and
Evans, H.C. (2006) Studies on the rust Prospodium tu-
berculatum, a new classical biological control released
against the invasive weed Lantana camara in Australia.
1. Life-cycle and infection parameters. Australasian Plant
Pathology 35, 309319.
Elfers, S.C. (1988) Element stewardship abstract for Schinus terebinthifolius, Brazilian Pepper-tree. In: Wild Land
Invasive species. The Nature Conservancy. Available at:
http://conserveonline.org/docs/2001/05/schiter.PDF (ac
cessed 7 December 2001).
Farr, D.F., Bills, G.F., Chamuris, G.P. and Rossman, A.Y.
(1989) Fungi on Plants and Plant Products in the United
States. APS Press, The American Phytopathological Soci
ety, St. Paul, MN, 1252 pp.
Gardner, D.E. (1999) Septoria hodgesii sp. nov.: a potential
biocontrol agent for Myrica faya in Hawaii. Mycotaxon
70, 247253.
Greene, H.C. (1950) Notes on Wisconsin parasitic fungi.
XIV. The American Midland Naturalist 44, 630642.
Gogue, G.J., Hurst, C.J. and Bancroft, L. (1974) Growth
inhibition by Schinus terebinthifolius. HortScience 9,
301.
Hansford, C.G. (1961) The Meliolineae, a monograph. Beiheft Sydowia 2, 1806.
Hight, S.D., Cuda, J.P. and Medal, J.C. (2002) Brazilian
peppertreee. In: Van Driesche, R., Lyon, S., Blossey, B.,
Hoddle M. and Reardon, R. (eds) Biological Control of
Invasive Plants in the Eastern United States.USDA Forest
Service, Morgantown, WV, pp. 311321.
Julien, M. and White, G. (1997) Biological Control of Weeds:
Theory and Practical Application. ACIAR, Canberra,
Australia, 190 pp.
Kirk, P.M., Cannon, P.F., David, J.C. and Stalpers, J.A. (2001)
Dictionary of the Fungi, 9th ed. CAB Publishing, Waling
ford, UK, 655 pp.
Lorenzi, H. and Matos, F.J.A. (2002) Plantas Medicinais no
Brasil Nativas e Exticas. Plantarum, Nova Odessa, So
Paulo, Brazil, 544 pp.
Lorenzi, H. (1992) rvores Brasileiras: Manual de Identificao e Cultivo de Plantas Arbreas Nativas do Brasil, vol 1. Plantarum, Nova Odessa, So Paulo, Brazil,
368 pp.
Mack, R.N. (1991) The commercial seed trade: an early dis
perser of weeds in the United States. Economic Botany
45, 257273.
Monteiro, F.T., Vieira, B.S. and Barreto, R.W. (2003) Curvularia lunata and Phyllachora sp.: two pathogens of
the grassy weed Hymenachne amplexicaulis from Brazil.
Australasian Plant Pathology 32, 449453.
Morgan, E.C. and Overholt, W.A. (2005) Potential allelo
pathic effects of Brazilian pepper (Schinus terebinthifolius Raddi, Anacardiaceae) aquaeous extract on germina
tion and growth of selected Florida native plants. Journal
of the Torrey Botanical Society 132, 1115.
Morton, J.F. (1978) Brazilian pepper its impact on people,
animals and the environment. Economy Botany 32, 353
359.
Nag Raj, T.R. (1992) Coelomycetous anamorphs with appendage-bearing conidia. Mycologue Publications, On
tario, Canada, 1101 pp.
Norman, D.J. (1995) Development of Colletotrichum gloeosporioides f. sp. clidemiae and Septoria passiflorae into
two mycoherbicides with extended viability. Plant Disease 79, 10291032.
276
Fungal pathogens of Schinus terebinthifolius from Brazil as potential classical biological control agents
Palm, M.E. (1991) Taxonomy and morphology of the syn
anamorphs Pilidium concavum and Hainesia lythri (coe
lomycetes). Mycologia 83, 787796.
Panetta, F.D. and McKee, J. (1997) Recruitment of the in
vasive ornamental, Schinus terebinthifolius, is dependent
upon frugivores. Australian Journal of Ecology 22, 432
438.
Pereira, O.L. and Barreto, R.W. (2005) The mycobiota of the
weed Mitracarpus hirtus in Minas Gerais (Brazil), with
particular reference to fungal pathogens for biological
control. Australasian Plant Pathology 34, 4150.
Pereira, J.M., Barreto, R.W., Ellison, C.A. and Maffia, L.A.
(2003) Corynespora cassiicola f.sp. lantanae: a potential
biocontrol agent from Brazil for Lantana cmara. Biological Control 26, 2131.
Pereira, O.L., Barreto, R.W., Cavalazzi, J.R.P. and Braun, U.
(2007) The mycobiota of the cactus weed Pereskia aculeata in Brazil, with comments on the life-cycle of Uromyces pereskiae. Fungal Diversity 25, 167180.
Priest, M.J. (2006) Fungi of Australia: Septoria. Melbourne,
Australia, 259 pp.
Saccardo, P.A. (1884) Sylloge fungorum 3, 1767.
Saccardo, P.A. (1902) Sylloge Fungorum 16, 11291.
Seixas, C.D.S., Barreto, R.W. and Killgore, E. (2007) Fun
gal pathogens of Miconia calvescens (Melastomataceae)
from Brazil, with reference to classical biological control.
Mycologia 99, 99111.
Shear, C.L. and Dodge, B.O. (1921) The life history and
identity of Patellina fragariae, Leptothyrium macro-
thecium and Peziza oenotherae. Mycologia 13, 135
170.
Taylor, L. (1998) Brazilian peppertree. In: Herbal Secrets of
the Rainforest. http://www.rain-tree.com/peppertree.htm
(accessed 7 December 2001).
Trujillo, E.E. (2005) History and success of plant pathogens
for biological control of introduced weeds in Hawaii. Biological Control 33, 113122.
Trujillo, E.E. and Norman, D.E. (1995) Septoria leaf spot
of lantana from Ecuador: a potential biological control
for bush lantana in forests of Hawaii. Plant Disease 79,
819821.
Vigas, A.P. (1961) ndice de Fungos da Amrica do Sul. Insti
tuto Agronmico de Campinas, Sao Paulo, Brazil, 921 pp.
Watson, A.K. (1991) The classical approach with plant patho
gens. In: Tebeest, D.O. (ed.) Microbial Control of Weeds.
Chapman & Hall, New York, pp. 323.
Williams, D.A., Overholt, W.A., Cuda, J. P. and Hughes,
C.R. (2005) Chloroplast and microsatellite DNA diver
sity reveal the introduction history of Brazilian peppertree
(Schinus terebinthifolius) in Florida. Molecular Ecology
14, 36433656.
Yoshioka, E.R., and Markin, G.P. (1991) Efforts of biologi
cal control of Christmas berry Schinus terebinthifolius in
Hawaii. In: Center, T.D., Doren, R.F., Hofstetter, R.L.,
Myers, R.L. and Whiteaker, L.D. (eds) Proceedings of
the Symposium of Exotic Pest Plants. US Department of
the Interior, National Park Service, Washington, DC, pp.
377385.
277
Testing the efficacy of specialist herbivores
to control Lepidium draba in combination
with different management practices
H.L. Hinz,1 A. Diaconu,2 M. Talmaciu,3 V. Nastasa4 and M. Grecu2
Summary
Lepidium draba L. [=Cardaria draba (L.) Desv.; Brassicaceae] is a perennial mustard, indigenous to
Eurasia. In the 18th century, L. draba was introduced into North America, where it is now listed as
noxious in 16 states and three Canadian provinces. In 2001, a biological control program was initiated investigating the possibilities for biological control of L. draba in North America. To determine
whether specialist herbivores are actually limiting population growth of L. draba in its area of origin
and how their effect might interact with soil nutrients and management regimes, we established manipulative field experiments in Spring 2006 in eastern Romania, where five of the currently studied
biological control agents are present. Plots (3 3 m) were established in already existing L. draba
stands, and four treatments applied in a split-plot design: (1) grazing (yes, no); (2) cultivation (none,
shallow, shallow + sowing of grasses/legume mix; (3) pesticide application to exclude herbivores
and/or augmentation of specialists (yes, no) and (4) carbon addition in the form of sawdust to reduce
plant available soil nitrogen (yes, no). First results indicate that cultivation and cultivation + sowing
reduced the number of L. draba plants, however only at the beginning of the field season. As expected,
neither pesticide nor sawdust applications had an effect on plant numbers or plant vigor in the first year;
they will probably need more time to become apparent. We expect that grazing treatments, which will
start in May 2007, will probably have the largest effect on L. draba vigor and densities. Final results
should allow us to develop recommendations for an integrated management strategy for L. draba.
Keywords: integrated weed management, topdown effects, hoary cress, grazing,
carbon addition, insect exclusion.
Introduction
Hoary cresses or whitetops, Lepidium spp. (=Cardaria
spp.), are perennial mustards of Eurasian origin (Hegi,
1987), which were introduced to the USA in the late
19th century and have since then spread throughout the
western and the northeastern states. They are aggressive invaders of crops, rangelands and riparian areas,
but they grow particularly well in disturbed and/or irrigated areas (Lyons, 1998). Because they are difficult
CABI EuropeSwitzerland, Rue des Grillons 1, 2800 Delmont,
Switzerland.
2
Institute of Biological Research, Bd. Carol I, 20-A, 700505 Iasi,
Romania.
3
University of Agronomy Sciences and Veterinary Medicine, M. Sadoveanu Alley, 700490, Iasi, Romania.
4
Central Research Station for Soil Erosion Control, 737405 Perieni,
Romania.
Corresponding author: H.L. Hinz <[email protected]>.
CAB International 2008
to control sustainably using mechanical or chemical
methods, a consortium was established in Spring 2001
to investigate the scope for classical biological control.
As a result of literature and field surveys conducted at
CABI EuropeSwitzerland between 2001 and 2003,
seven phytophagous insect species were prioritized as
potential biological control agents based on records of
their restricted host range, and five species (four weevils and one flea beetle) are currently being investigated
(Cripps et al., 2005). In addition, one gall-forming
weevil and an eriophyid mite are being studied by the
US Department of Agriculture Agricultural Research
Service and European Biological Control Laboratory
in Montpellier, France, and Montana State University,
Montana, respectively.
One of the assumptions in biological weed control
is that, in their area of origin, invasive plants are regulated by natural enemies. However, evidence for the
topdown regulative ability of herbivores is equivocal
(Crawley, 1989; Price, 1992; Maron and Vil, 2001 and
references therein). The outcome of studies depended
278
Testing the efficacy of specialist herbivores to control Lepidium draba
on the life history of the plant species investigated, the
herbivore community and on environmental conditions
(Brown and Gange, 1989; Louda and Potvin, 1995;
Maron et al., 2002; DeWalt et al., 2004).
To investigate whether herbivory, particularly by
specialists, has an effect on Lepidium draba at the
population level, we established two identical largescale experiments at two natural field sites of L. draba
in eastern Romania, where five of the currently studied
potential biological control agents occur. These include
the gall-inducing weevils, Ceutorhynchus cardariae
Korotyaev (Col., Curculionidae) and Ceutorhynchus
assimilis (Paykull) [=C. pleurostigma (Marsham)], the
seed-feeding weevil, Ceutorhynchus turbatus Schultze,
the stem-mining flea beetle, Psylliodes wrasei Leonardi & Arnold (Col., Chrysomelidae) and the gall mite,
Aceria drabae (Nal.) (Acari, Eriophyidae). Results
will indicate whether herbivores are actually limiting
population growth of L. draba in its area of origin and
how their effect might interact with soil nutrients and
management regimes. This should ultimately help us
to develop an integrated management strategy for L.
draba.
Materials and methods
The study species
L. draba is a herbaceous, long-lived perennial that
reproduces vegetatively and by seed (Lyons, 1998). In
the introduced range, established plants bolt in early
spring, flower in May and June and set seed in July.
A combination of several control practices, including
herbicide application and physical removal by hoeing
or tilling, followed by plantings of competitive species
is considered the most effective control strategy for L.
draba (Lyons, 1998). Sheep will graze L. draba and
especially like seedlings (Lyons, 1998). However, a
complete grazing management program has not been
developed yet (Sheley and Stivers, 1999).
Females of C. cardariae lay their eggs in the leaf
midribs, petioles and developing shoots of L. draba
from early spring until mid-June (Hinz et al., 2007).
Female oviposition induces the galls, in which the larvae mine and develop. The root-gall-forming weevil,
C. assimilis, has several generations per year (Jourdheuil, 1963) and appears to have at least two different host races, one of which may be host-specific to
L. draba (Fumanal et al., 2004). The seed-feedingweevil, C. turbatus, lays its eggs into the developing
fruits of hoary cress and its larvae destroy one or both
seeds during their development. P. wrasei lays its eggs
between the end of August and the end of October in
the soil close to shoots of L. draba. After an obligate
diapause, larvae hatch from eggs the following spring
and feed in the developing shoots and vegetative points
of L. draba (Hinz et al., 2007). C. cardariae, C. assimilis and P. wrasei can kill whole ramets or devel-
oping shoots of L. draba or at least stunt their growth
(Fumanal et al., 2004; Hinz et al., 2006). The gall mite,
Aceria drabae (Nal.) (Acari, Eriophyidae), overwinters
in dormant shoot buds of L. draba. In spring, the mite
is passively carried up in the developing shoot where
it feeds on meristematic tissue (J. Littlefield, personal
communication). The wind-dispersed mites have several generations per year and can reduce or completely
prevent seed production (Lipa, 1978). Extensive hostspecificity tests have shown that both C. cardariae and
A. drabae have a narrow host range, and petitions for
field release of both species will start to be prepared in
Autumn 2007.
The field sites
Experiments were established at two natural field
sites of L. draba situated in Eastern Romania, which
forms part of the area of origin of L. draba (Hegi,
1987). One is near Iai (4710N, 2727E) and is about
100 ha; it is the largest site of L. draba we found during our surveys. The site is used as a pasture for peasants cattle and horses from the neighbouring villages
and is heavily overgrazed. We believe that cattle avoid
grazing L. draba and, in combination with the disturbance caused by the livestock, this is assumed to favor
L. draba and may have facilitated its increase to the
exceptional current population.
The second site is a sheep pasture of about 70 ha and
belongs to the Central Research Station for Soil Erosion Control in Perieni (4816N, 2738E), about 100
km south of Iai. L. draba occurs in localized areas,
about 0.51 ha in size.
The field experiment
At the end of March 2006, six blocks (18 22 m)
were established at each of the two field sites. Twelve
plots (3 3 m), separated by 2-m buffer zones, were
set up within each block (i.e. 72 in total per site), and
the following four treatments were applied in a splitplot design: (1) grazing (yes, no); (2) cultivation (none,
shallow, shallow + sowing of grass/legume mix); (3)
pesticide application to exclude herbivores and/or augmentation of specialists (yes, no) and (4) carbon ad
dition in the form of sawdust to reduce plant available
soil nitrogen (yes, no). On 29 March 2006, plots were
harrowed, and on 1 April, a grass/legume mixture was
sown according to treatments. At the beginning of May,
all blocks were fenced to protect plots against grazing
animals to allow the newly sown grass and disturbed
plants in cultivated plots to recover. From the beginning of May 2007 onwards, one side of half of the plots
is opened for 1 to 2 weeks per month to either cattle and
horses (Iasi) or sheep (Perieni).
To exclude phytophagous arthropods, the systemic
insecticide imidacloprid (Confidor 200; Bayer AG,
279
XII International Symposium on Biological Control of Weeds
May
60
60
40
40
20
20
0
20
15
10
cult+sown
10
cultivated
cultivated
15
none
20
June
80
none
Number of L. draba plants
80
Figure 1.
treated with pesticides. In addition, about 50 individuals of four of the potential agents (i.e. C. assimilis, C.
cardariae, C. turbatus, P. wrasei) and other weevils
and flea beetles associated with L. draba were released
on each non-treated plot in Perieni at the end of June.
In Iasi, nearly all plants were naturally attacked, so no
artificial infestations were made.
In each 3 3 m plot, smaller 0.5 0.5 m subplots
were established during April, and the number of L.
draba plants (ramets) was recorded four to five times
between mid-April and mid-July 2006. During May,
plant traits (i.e. phenological stage, number of shoots
and height of each shoot) were recorded for a maximum of 20 plants per subplot, chosen along two diagonal lines. In addition, any foliage damage (visible from
the outside) was noted and, as far as possible, attributed
to specific herbivore species. To record potential differences in species composition, visual ground cover
estimates were noted in three 0.5 0.5 m subplots during July, and the following categories were recorded:
percent cover of L. draba, forbs, legumes, grasses and
bare ground.
cult+sown
Leverkusen, Germany) and the miticide abamectin
(Vermitec 18; Agri-Mek, 0.15 EC, Syngenta SA, Basel, Switzerland) were applied as a soil drench in the
second half of April 2006 at a rate of 13.5 and 1 ml,
respectively, per plot (3 3 m). Thereafter, plots were
sprayed in 2-week intervals until July with a mixture
of abamectin and the pyrethroid cypermethrin (Cyperguard 25 EC; Gharda Chemicals Ltd., India) at a rate of
1 and 0.2 ml, respectively, per plot. Because abamectin did not work satisfactorily, it was later replaced by
hexythiazox (Nissorun 10 WP; Nippon Soda Co., Ltd.,
Japan) at a rate of 0.5 g per plot. All pesticides applied
were tested for their potential direct effect on growth of
L. draba. Plots not receiving pesticides were sprayed
with an equal volume of water. For plots with carbon
addition, sawdust was applied once a month between
April and June by hand to the soil surface at a rate of
1.5 kg/m2 per year.
On 19 May and on 7 July, experimental plots in Perieni were infested with A. draba by collecting Aceria-
damaged L. draba plants in the vicinity of Iasi and
placing them close to L. draba plants on the plots not
Mean number (SE) of Lepidium draba plants in 0.25 m2 quadrats in May and June 2006 at Iasi (a and b) and
Perieni (c and d) after different cultivation regimes.
280
Testing the efficacy of specialist herbivores to control Lepidium draba
Plant traits were analysed using a hierarchical design
with pesticide application (yes, no) and sawdust addition (yes, no) nested within cultivation treatments. Because grazing treatments will only start in Spring 2007,
they were not included in the analyses of the 2006 data.
Plant numbers were analysed using repeated measures
analysis of variance. Data on percent cover and percent plants reproducing were arcsin-transformed. All
analyses were conducted with Statistical Package for
the Social Sciences 14.0.
Results and discussion
The density of L. draba was four to five times higher in
Iasi than in Perieni (Fig. 1), presumably because of the
long history of stronger grazing pressure on the associated vegetation in Iasi. Cultivation and cultivation +
sowing of a grass/legume mixture significantly reduced
the number of L. draba plants at both sites, however
only at the beginning of the field season (Iasi: F2,10 =
18.23; Perieni: F2,10 = 13.92; P < 0.001; Fig. 1). On
non-cultivated plots, plant numbers declined by over
50% between May and June, which was presumably
due to interspecific competition (Iasi: F2,47 = 103.88;
Leaves
% L. draba plants attacked
100
Shoots/root-crowns
10.0
75
7.5
50
5.0
25
2.5
0.0
100
7.5
50
5.0
25
2.5
P-/C-
P+/C-
0.0
P-/C+ P+/C+
10.0
75
0
Figure 2.
Perieni: F2,47 = 23.42; P < 0.001). The exclusion of
grazing in all blocks in 2006 considerably reduced disturbance in uncultivated plots and, in addition, allowed
the associated vegetation to compete more effectively
with L. draba.
Neither the number of shoots nor shoot height was
significantly affected by any of the treatments. Insect
attack on leaves was much higher in Iasi than in Perieni
(Fig. 2), and pesticide applications only reduced leaf attack in Perieni (F2,47 = 5.94; P = 0.019). Overall, attack
of leaves was very high, probably because pesticides
were applied too late and in a too-low dosage. Attack
of shoots and root crowns, which were visible from the
outside, was low at both sites (Fig. 2). However, attack by C. assimilis and P. wrasei were likely underestimated because plants were only visually inspected
for attack. In Iasi, sowing of a grass/legume mixture
increased percent cover of legumes in July (29% vs.
39%; F2,71 = 3.79; P = 0.027). In both Iasi and Perieni,
percent cover of grasses was highest in uncultivated
plots, 25% and 56%, respectively (Iasi: F2,71 = 3.47;
P = 0.037; Perieni: F2,71 = 3.85; P = 0.026), and cover
of bare ground was lowest (Iasi: F2,71 = 22.85, Perieni:
F2,71 = 23.32; P < 0.001).
P-/C-
P+/C-
P-/C+ P+/C+
Effect of regular application of pesticides (P +/-) and carbon in the form of sawdust (C +/-) on the mean proportion (SE) of Lepidium draba plants attacked in May 2006 at Iasi (a and b) and Perieni (c and d). Attack is based
on visual examination of plants.
281
XII International Symposium on Biological Control of Weeds
Conclusions and outlook
As expected, neither pesticide nor sawdust applications
had a large effect in the first year. To increase the effectiveness of pesticide applications, dosages are increased and treatments are started earlier in 2007. To
obtain a better estimate of endophagous herbivory, we
are also planning to dig up plants from the periphery
of plots and dissect them. This will also allow us to
correlate attack with individual plant traits. Cultivation
treatments only had a temporary effect on L. draba.
In a review of studies on the impact of herbivores
on plant population dynamics, Crawley (1989) found
that vertebrate herbivory generally had a stronger impact on plant population dynamics than invertebrate
herbivores. Only 6 weeks after fencing the blocks, we
observed large differences between the areas inside
and outside ungulate exclosures, i.e. vegetation outside
ungulate exclosures was heavily overgrazed. We therefore established eight additional 3 3 m plots outside
of exclosures in July 2006 as comparison.
In conclusion, we expect that grazing treatments
will have a larger effect on L. draba growth and densities than phytophagous arthropods and that the effect
of the latter and carbon addition will need more time to
become apparent. The project will continue until Autumn 2008, when final conclusions will be drawn.
Acknowledgements
We thank Valentin Cozma, Madalin Parepa, Cornelia
Closca and Dragos Filote (all Institute of Biological
Research, Iasi, Romania) for technical assistance, Nela
Talmaciu, Vasile Vintu and Costel Samuil (University
of Agronomy Sciences and Veterinary Medicine, Iasi,
Romania) for facilitating the establishment of field plots
in Iasi and discussions and Dumitru Nistor and Lucian
Stanescu (Central Research Station for Soil Erosion
Control, Perieni, Romania) for facilitating the field experiment in Perieni. We would also like to thank Ren
Eschen and Urs Schaffner for advice in experimental
design and data analyses. This project is financed by
the Swiss National Science Foundation in the framework of the SCOPES program (IB73AO-110772).
References
Brown, V.K. and Gange, A.C. (1989) Differential effects of
above- and below-ground insect herbivory during early
plants succession. Oikos 54, 6776.
Crawley, M.J. (1989) Insect herbivores and plant population
dynamics. Annual Review of Entomology 34, 531564.
Cripps, M.G., Hinz, H.L., McKenney, J.L., Harmon, B.L.,
Merickel, F.W. and Schwarzlaender, M. (2005) Comparative survey of the phytophagous arthropod faunas associ-
ated with Lepidium draba in Europe and the western United States, and the potential for biological weed control.
Biocontrol Science and Technology 16, 10071030.
DeWalt, S.J., Denslow, J.S. and Ickes, K. (2004) Natural-
enemy release facilitates habitat expansion of the invasive
tropical shrub Clidemia hirta. Ecology 85, 471483.
Fumanal, B., Martin, J., Sobhian, R., Blanchet, A. and Bon,
M. (2004) Host range of Ceutorhynchus assimilis (Coleoptera: Curculionidae), a candidate for biological control
of Lepidium draba (Brassicaceae) in the USA. Biological
Control 30, 598607.
Hegi, G. (1987) Illustrierte Flora von Mitteleuropa. In: Conert, H.J., Hamann, U., Schultze-Motel, W. and Wagenitz,
G. (eds) Spermatophyta, Band IV Teil 1. Angiospermae,
Dicotyledones 2. Paul Parey, Berlin, Germany, 598 pp.
Hinz, H.L., Borowiec, N., Coromoto Colmenarez, Y., Cortat,
G., Cuenot, M., Grecu, M. and Szucs, M. (2006) Biological control of whitetops, Lepidium draba and L. appelianum. Annual report 2005. Unpublished report, CABI
EuropeSwitzerland, Delmont, Switzerland, 33 pp.
Hinz, H.L., Cortat, G., Muffley, B. and Tostado, C. (2007) Biological control of whitetops, Lepidium draba and L. appelianum. Annual report 2006. Unpublished report, CABI
EuropeSwitzerland, Delmont, Switzerland, 32 pp.
Jourdheuil, P. (1963) Ceutorhynchus pleurostigma Marsham.
In: Balachowsky, A.S. (ed.) Entomologie applique
lagriculture (Coloptres). Masson et Cie., Paris, France,
pp. 10211028.
Lipa, J.J. (1978). Preliminary studies on the species Aceria
drabae (Nal.) (Acarina, Eriophyiidae) and its potential
for the biological control of the weed Cardaria draba L.
(Cruciferae). Prace Naukowe Instytutu Ochrony Roslin
20, 139155.
Louda, S.M. and Potvin, M.A. (1995) Effect of inflorescencefeeding insects on the demography and lifetime fitness of
a native plant. Ecology 76, 229245.
Lyons, K.E. (1998) Cardaria draba (L.) Desv. heart-podded
hoary cress, Cardaria chalepensis (L.) Hand-Maz. lenspodded hoary cress and Cardaria pubescens (C.A. Meyer)
Jarmolenko globe-podded hoary cress. In: Meyers-Rice,
B. (ed.) Elemental Stewardship Abstract. The Nature
Conservancy, Artlington, VA.
Maron, J.L. and Vil, M. ( 2001) When do herbivores affect
plant invasion? Evidence for the natural enemies and biotic resistance hypotheses. Oikos 95, 361373.
Maron, J.L., Combs, J.K. and Louda, S.M. (2002) Convergent
demographic effects of insect attack on related thistles in
coastal vs. continental dunes. Ecology 83, 33823392.
McInnis, M.L., Larson, L.L. and Miller, R.F. (1990) Firstyear defoliation effects on whitetop (Cardaria draba (L.)
Desv.). Northwest Science 64, 107.
Price, P.W. (1992) Plant resource as the mechanistic basis for
insect herbivore population dynamics. In: Hunter, M.D.,
Ohgushi, T. and Price, P.W. (eds) Effects of Resource Distribution on AnimalPlant Interactions. Academic, London, pp. 139173.
Sheley, L. and Stivers, J.I. (1999) Whitetop. In: Sheley, L.
and Petroff, K. (eds) Biology and Management of Noxious Rangeland Weeds. OSU Press, Corvallis, OR, pp.
401407.
282
Assessing herbivore impact on a highly
plastic annual vine
J.A. Hough-Goldstein1
Summary
Effective biological control of a weed population requires an understanding of the impact of herbivory on the target host plant. This can be complicated by phenotypic plasticity in response to environmental heterogeneity. Polygonum perfoliatum L. [also known as Persicaria perfoliata (L.) H. Gross],
an invasive annual vine accidentally imported from Asia to the Mid-Atlantic region of the USA in
the 1930s, has been the target of a biological control program since 1996. In 2004, a Chinese weevil,
Rhinoncomimus latipes Korotyaev, was approved for release in North America. Cage studies in 2005
showed that P. perfoliatum is highly plastic in its response to light: Individual plants grown in cages
produced over 2000 seeds per plant in full sun but fewer than 400 in the shade. In 2006, with all plants
in full sun, weevils that were introduced into cages early in the season suppressed seed production for
about 9 weeks, but plants were able to produce substantial numbers of seeds late in the year. Further
studies with plants under conditions more closely approximating the field situation (e.g. with competition from other plants) are likely to show a greater impact of weevil herbivory on the plants.
Keywords: Polygonum perfoliatum, Persicaria perfoliata, Rhinoncomimus latipes.
Introduction
Mile-a-minute weed (MAM), Polygonum perfoliatum
L. [also known as Persicaria perfoliata (L.) H. Gross]
is an alien invasive weed from Asia that infests natural areas in a variety of habitats in its imported range.
This annual vine is a prolific seed producer and has become a serious problem in the Mid-Atlantic region of
the USA. The North American population is thought
to have originated near York, PA in the 1930s, probably introduced as a seed contaminant with holly seed
imported from Japan (Moul, 1948). Although it was
recognized as a potentially dangerous weed that should
be eradicated, no action was taken, and the weed can
now be found from Delaware west to Ohio, south to
West Virginia and north to Massachusetts. A biological
control program was initiated by the US Forest Service
in 1996. Over 100 insect species were identified on
MAM in China, including several that appeared to have
a narrow host range (Ding et al., 2004). One of these,
Rhinoncomimus latipes Korotyaev (Coleoptera: Curculionidae), was tested on plant species in China and
in quarantine in Delaware and found to be extremely
University of Delaware, Department of Entomology and Wildlife Ecology, Newark, DE, USA <[email protected]>.
CAB International 2008
1
host-specific (Price et al., 2003; Colpetzer et al., 2004).
This insect was approved for release in the USA by
the US Department of Agricultures Animal and Plant
Health Inspection Service in 2004.
Eggs of R. latipes are laid on MAM plants and hatch
in about 5 days. Neonates crawl along stems and enter a
node, where they feed internally for 1 to 2 weeks, after
which they drop out of the stem and pupate in the soil.
Adults emerge about 1 week later and feed on MAM
leaves and terminals. The weevils go through multiple,
overlapping generations until early to mid-September,
when egg laying ceases (unpublished data). Adult weevils overwinter in the soil or leaf litter.
The weevils have been reared at the New Jersey
Department of Agriculture Phillip Alampi Beneficial
Insect Laboratory in Trenton, NJ, since autumn 2004,
and in 2006, more than 20,000 weevils were reared and
released. Most releases occurred in New Jersey, but
insects have also been released and have established
at sites in Delaware, Pennsylvania, West Virginia and
Maryland. Although it is too soon to assess their impact
in the field, plant mortality has been observed in some
areas where weevils have heavily defoliated MAM
plants in New Jersey.
To gain a better understanding of the potential impact of the weevil on P. perfoliatum, experiments were
conducted in 2005 and 2006 using isolated MAM plants
enclosed in weevil-proof cages with various levels of
283
XII International Symposium on Biological Control of Weeds
weevils applied at different times. The 2005 experiment
also addressed the question of whether seed-cluster
consumption by birds or deer may have a significant
impact on P. perfoliatum seed production.
Methods and materials
2005 Experiment
Controls consisted of single MAM plants completely enclosed in cages approximately 2 m tall and
0.9 m square, while treatments with weevils were the
same but with 20 weevils per cage added on 21 July.
Cages open to birds were identical except with open
tops, and cages available to deer browse were ~1 m tall
with open tops. No R. latipes were present in this area
in 2005, and therefore, open cages were not exposed
to weevils. All cages were made from white polyester
netting material with a mesh of approximately 10 8
cm (BioQuip Products, Inc., Gardena, CA, USA), with
a Velcro opening sewn into one corner and supported
by frames constructed of 1.9 cm diameter polyvinyl
chloride conduit pipe.
Five replicates were set up for each of the four treatments, assigned at random to 20 plants, approximately
4 m apart, growing naturally at a site in White Clay
Creek State Park near Newark, DE. Other plants within
~ 0.5 m of the plants were removed on 6 June and again
on 21 July when cages were installed. Each plant was
provided with a tomato support extended with three
bamboo poles wrapped in wire to support growth of the
vine. The plants were about 1 m tall when treatments
Figure 1.
were applied. Nylon window screening was placed at
the bottom of each cage to collect seeds as they fell
from the plants.
The number of seed clusters per plant was recorded,
and fallen seeds (achenes) were collected from the
screening each week and counted. Relative plant size
and relative light exposure (full sun, partial sun or
shade) were also noted weekly. On 29 November, after
all plants were dead, the remaining plant material was
collected into large paper bags, left to dry in a greenhouse for 2 weeks and then weighed.
2006 Experiment
Thirty cages identical to the tall closed cages used
in 2005 were placed over isolated P. perfoliatum plants
at a different site in White Clay Creek State Park, approximately 1000 m away from the 2005 site, on 19
May 2006, when plants were about 30 cm tall. Cages
were at least 4 m apart and were arrayed along the edge
of a meadow in a randomized complete block design,
so that plants in the same block had similar exposure to
sun and most plants were exposed to full sun for much
of the day. There were six replicate cages each of five
treatments: early high, 20 weevils per cage added on 26
May; early low, five weevils per cage added on 26 May;
late high, 20 weevils per cage added on 23 June; late
low, five weevils per cage added on 23 June and control, no weevils added. Weevils were obtained from the
Phillip Alampi Beneficial Insect Laboratory, Trenton,
N.J. Although they were not sexed, they were assigned
randomly to the different treatments, and spot checks at
Effect of sun exposure on total numbers of seeds produced and plant dry weights (means
SE) for single Polygonum perfoliatum plants enclosed in cages in 2005. Means with the same
lowercase or uppercase letter are not significantly different (Tukeys test on square-root
transformed data; untransformed means and standard errors are shown).
284
Assessing herbivore impact on a highly plastic annual vine
the New Jersey laboratory indicated a 1:1 sex ratio in
reared weevils (D. Palmer, personal communication).
Cages were checked weekly for presence of defoliation, number of weevils that could be observed, presence of node damage and weevil eggs and number of
seed clusters per plant. Seeds were collected from the
screen at the base of the cages each week beginning
in late July and counted in the laboratory. Plants were
cut off at the base on 15 November and left to dry in
paper bags for several weeks as in 2005, after which
they were weighed.
Statistical analyses
Data were transformed by square root (x + 0.5) to reduced heteroskedasticity of variance residuals. Trans-
Figure 2.
formed data were analyzed using two-way analysis of
variance, by treatment and sun exposure in 2005 and by
treatment and block in 2006. Tukeys test was used for
mean separation. Non-transformed means and standard
errors are presented in figures.
Results
2005 Experiment
The total number of seeds produced by individual
plants varied from a low of 39 for a small control plant
(enclosed in a closed cage, without weevils) growing in
the shade, to a maximum of 3172 for a bird-exposed
plant (i.e. in a tall cage with an open top) growing in
the full sun. There were no significant differences by
Effect of weevil treatments on cumulative total number of seeds produced
(means SE) during A the first 9 weeks, and B the last 9 weeks of seed collection, for single Polygonum perfoliatum plants enclosed in cages in 2006.
Means with the same letter are not significantly different (Tukeys test on
square-root transformed data; untransformed means and standard errors are
shown). Late treatments had weevils added on 23 June, and Early weevils
were added on 26 May; Low treatments received five weevils per cage and
High treatments received 20 weevils per cage.
285
XII International Symposium on Biological Control of Weeds
cage treatment for total number of seeds (F3,2 = 2.90,
P = 0.0754), plant dry weights (F3,2 = 1.11, P = 0.3809)
or maximum number of seed clusters counted on plants
(on 29 September: F3,2 = 2.74, P = 0.0858). However,
differences in all of these parameters were highly significant by sun exposure (seed production, F3,2 = 41.10,
P < 0.0001; plant dry weights, F3,2 = 46.92, P < 0.0001
and seed clusters, F3,2 = 11.41, P = 0.0014). Plants
grown in full sun averaged over 2200 seeds per plant,
while those in full or partial shade produced an average
of 300600 seeds per plant (Fig. 1). Plant dry weight
differences were even more dramatic and differed
among all three levels of sun exposure (Fig. 1).
2006 Experiment
Total seed production did not differ by treatment in
2006 (F4,5 = 1.03, P = 0.4175). However, there was a
significant difference in the cumulative total number of
seeds produced during the first 9 weeks that seeds were
collected (24 July to 17 September: F4,5 = 3.77, P =
0.0213), with significantly more seeds produced on the
control plants and the late low plants than on the early
high plants (Fig. 2A). Differences by treatment were
not significant for total numbers of seeds produced
during the last 9 weeks (F4,5 = 1.08, P = 0.3938; Fig.
2B) or plant dry weights (F4,5 = 0.18, P = 0.9455). All
treatments in 2006 averaged more than 2600 seeds per
plant.
Discussion
The 2005 study revealed the extreme plasticity of P.
perfoliatum under different conditions of light exposure. Plants grown in full sun were more than ten times
larger and produced more than six times as many seeds
as plants grown in shade. Similar results were obtained
by Sultan and Bazzaz (1993), who found very large
differences in fruit and plant biomass produced by Polygonum persicaria L. under different light regimes in
the greenhouse. These differences apparently swamped
any that may have occurred due to minor feeding on
seed clusters by deer or birds in the open cages or by
weevils added to the cages in July in 2005.
In 2006, seed production was almost completely
suppressed between late July and mid-September in
plants with early application of weevils at the high level
(20 weevils per plant). However, all plants produced
numerous seeds in October, resulting in no significant
difference by treatment in total seed production over
the season. In this experiment, the caged plants were
unusually large and robust, as most were growing in
full sun and had minimal competition with other plants.
Further studies on MAM plants under conditions more
closely approximating the field situation, especially
with competition from other plants, are likely to show
a greater impact of weevil herbivory on the plants.
Based on observations in the field, where large weevil
populations have developed and plants are subject to
normal competitive stress, complete seed suppression
and plant mortality can occur.
As with many invasive plant species, MAM plants
showed an impressive ability to recover from severe
insect damage. Nevertheless, under conditions where
weevil populations grow exponentially over several
years, the insect may be able to act as an effective biological control agent.
Acknowledgements
Megan Schiff, Ellen Lake, Brian Butterworth, Jamie
Pool, Jason Graham, Matt Frye and Louisa Harding all
contributed greatly to this project. Daniel Palmer and
Amy Diercks, New Jersey Department of Agriculture,
developed rearing methods and shared their knowledge
and weevils. I also thank Richard Reardon for support
through the Forest Health Technology Enterprise Team,
USDA Forest Service, Morgantown, WV.
References
Colpetzer, K., Hough-Goldstein, J., Ding, J. and Fu, W.
(2004) Host specificity of the Asian weevil, Rhinoncomimus latipes Korotyaev (Coleoptera: Curculionidae), a
potential biological control agent of mile-a-minute weed,
Polygonum perfoliatum L. (Polygonales: Polygonaceae).
Biological Control 30, 511522.
Ding, J., Fu, W., Reardon, R., Wu, Y. and Zhang, G. (2004)
Exploratory survey in China for potential insect biocontrol
agents of mile-a-minute weed, Polygonum perfoliatum L.,
in Eastern USA. Biological Control 30, 487495.
Moul, E.T. (1948) A dangerous weedy Polygonum in Pennsylvania. Rhodora 50, 6466.
Price, D.L., Hough-Goldstein, J. and Smith, M.T. (2003) Biology, rearing, and preliminary evaluation of host range of
two potential biological control agents for mile-a-minute
weed, Polygonum perfoliatum L. Environmental Entomology 32, 229236.
Sultan, S.E. and Bazzaz, F.A. (1993) Phenotypic plasticity
in Polygonum persicaria. I. Diversity and uniformity in
genotypic norms of reaction to light. Evolution 47, 1009
1031.
286
The disintegration of the
Scrophulariaceae and the biological
control of Buddleja davidii
M.K. Kay,1 B. Gresham,1 R.L. Hill2 and X. Zhang3
Summary
The woody shrub buddleia, Buddleja davidii Franchet, is an escalating weed problem for a number
of resource managers in temperate regions. The plants taxonomic isolation within the Buddlejaceae
was seen as beneficial for its biological control in both Europe and New Zealand. However, the recent revision of the Scrophulariaceae has returned Buddleja L. to the Scrophulariaceae sensu stricto.
Although this proved of little consequence to the New Zealand situation, it may well compromise European biocontrol considerations. Host-specificity tests concluded that the biocontrol agent, Cleopus
japonicus Wingelmller (Coleoptera, Curculionidae), was safe to release in New Zealand. This leaffeeding weevil proved capable of utilising a few non-target plants within the same clade as Buddleja
but exhibited increased mortality and development times. The recent release of the weevil in New
Zealand offers an opportunity to safely assess the risk of this agent to European species belonging to
the Scrophulariaceae.
Keywords: Cleopus, Buddleja, taxonomic revision, phylogeny.
Introduction
There are approximately 90 species of Buddleja L.
indigenous to the Americas, Asia and Africa (Leeuwenberg, 1979), and a number have become naturalized outside their native ranges (Holm et al., 1979).
Buddleia, Buddleja davidii Franchet, in particular, is an
escalating problem for resource managers in temperate
regions and has been identified as a target for classical biological control in New Zealand (Kay and Smale,
1990) and Europe (Sheppard et al., 2006).
Buddleia is a large woody shrub of Asian origin that
was introduced to the rest of the world as an ornamental
species in the 1890s. It was considered naturalized in
the UK in the 1930s and in New Zealand in the 1940s
(Esler, 1988). It has many of the features that characterize successful weed species, and it is ranked in the
top ten invasive plants of Britain (Crawley, 1987). It
matures quickly, is capable of flowering in its first year
of life and produces an extraordinary number of small
seeds that are efficiently dispersed by wind. However,
1
Ensis, Private Bag 3020, Rotorua, New Zealand.
Hill & Associates, Christchurch, New Zealand.
3
NAU, Plant Protection, Nanjing, China.
Corresponding author: M.K. Kay <[email protected]>.
CAB International 2008
2
there is no significant soil seed bank. The seed germinates almost immediately, and the density and rapid
early growth of buddleia seedlings suppresses other
pioneer species (Smale, 1990).
As a naturalized species, buddleia is a shade-intolerant
colonizer of urban wastelands, riparian margins and
other disturbed sites, where it may displace indigenous
species, alter nutrient dynamics and impede access
(Smale, 1990; Bellingham et al., 2005). In New Zealand, on sites prepared for exotic forest plantations, the
rapid growth of buddleia causes the suppression and a
quantifiable loss of growth in newly planted Pinus radiata Don. (Richardson et al., 1996). The inefficiencies
of conventional controls prompted the investigation of
classical biological control (Kay and Smale, 1990).
The taxonomic isolation of a target weed from indigenous and other valued non-target plant species reduces the risk posed by introduced biological control
agents. However, taxonomy is far from an exact science, and the taxonomy of the paraphyletic Buddleja
has had a chequered history. Buddleja has variously
been placed within the families, Scrophulariaceae, Loganiaceae, the conveniently promoted Buddlejaceae
and, most recently, returned to the Scrophulariaceae,
which has been a recognized repository for undefined
Lamiales (Tank et al., 2006). The on-going reconstructing of the Scrophulariaceae combines morphological,
287
XII International Symposium on Biological Control of Weeds
embryological, molecular and chemical parameters, as
well as the host preferences of specialist invertebrates
(Stevens, 2001).
Fortunately, there are few close relatives of buddleia
within the New Zealand indigenous Scrophulariaceae
s.s., although the indigenous shrub, Myoporum laetum
G. Forst., Myoporaceae, has now been relegated to the
tribe Myoporeae, within the same clade as Buddlejeae
(Tank et al., 2006). Most other New Zealand genera
previously placed in the Scrophulariaceae, including
the manifold Hebe Comm. ex Juss., are now better defined in other clades within the Scrophulariaceae sensu
lato.
During a survey of insects and pathogens associated
with B. davidii in China, Cleopus japonicus Wingelmller (Coleoptera, Curculionidae) appeared to be a
potential biological control agent because of its apparent host specificity and ubiquity. The adults and larvae
feed externally on leaves. Eggs are oviposited singly
within excavated leaf cavities, and the emergent sluglike larvae remain attached to the plant by secreting a
coating of viscous fluid.
The host-specificity studies reported in this paper
evaluated whether C. japonicus is a safe biological
control agent for buddleia in New Zealand.
Methods and materials
Preliminary trials conducted in China tested species
from 16 plant families. C. japonicus was then imported into quarantine in New Zealand. The 76 plant
taxa tested in New Zealand were selected following the
internationally accepted centrifugal phylogenetic system of Wapshere (1974). The relative susceptibility of
14 Buddleja taxa was tested. Thirty-five New Zealand
indigenous plant species were tested, including Geniostoma rupestre J.R. Forst.& G. Forst., the only endemic
representative of the Loganiaceae, and 21 species from
the Scrophulariaceae s.l. Given the uncertain nature of
Buddleja taxonomy, it was considered prudent to give
extensive coverage of New Zealand scrophularia species, particularly the many species of Hebe. A further
11 species of Scrophulariaceae s.l. that are exotic to
New Zealand were tested, along with 16 exotic species
from other families that commonly grow in association
with buddleia in New Zealand. The New Zealand trials
were conducted in a quarantine insectary maintained at
20C 2 and 70% 10 RH, 14-h photoperiod.
Tests were run with nave and pre-fed adults, in
both choice and no-choice trials. The degree of feeding, oviposition and mortality was scored against that
of insects placed on concurrent buddleia controls. Nochoice larval trials utilized both pre-fed and nave first
instar larvae. To obtain nave larvae, eggs of known age
were monitored closely for larval eclosion. Emerging
larvae were transferred to the test plant material before
feeding.
Results
A full account of trial results is available on the Environmental Risk Management Authority website (www.
ermanz.org). Adult C. japonicus did not oviposit, or
feed, on any of the 35 species belonging to 24 plant
families outside of the Scrophulariaceae in either of
the preliminary trials in China or the trials conducted
in New Zealand. However, the weevil did lay a very
small number of eggs on a few of the 21 New Zealand indigenous species within the family Scrophulariaceae s.l. These eggs were laid externally, rather than
in purposefully excavated sites, and failed to produce
larvae. Larvae transferred to these plants also developed poorly. Within the genus Buddleja, C. japonicus
could complete development on all, except Buddleja
salviifolia (L.) Lam and Buddleja auriculata Benth. but
performed best and had a significant preference for B.
davidii (Table 1).
Newly emerged larvae transferred to the foliage of
17 New Zealand indigenous Hebe species died quickly
without completing development. One anomaly occurred when one larva of one replicate completed
development to adult on the foliage of an ornamental
specimen of Hebe speciosa (A.Cunn.) Ckn. & Allan.
One larva also completed development on each of
the indigenous Limosella lineata Glck, [Limosellae
Table 1.
288
Summary of feeding and oviposition trials
for Cleopus japonicus presented as a ranking
of host suitability within the genus Buddleja.
Ranking was determined from the combined
rankings of larval and adult feeding and oviposition for each species.
Buddleja species
B. davidii Franch.
var. lochinch
var. weyeriana
B. madagascariensis
Lam.
B. japonica Hemsl.
B. alterniflora
Maxim.
B. globosa Hope
B. lindleyana
Fortune
B. parviflora
H. B. K.
B. asiatica Lour.
B. colvillei
Hook. f. et Thoms
B. dysophylla
(Benth.) Radlk.
B. auriculata
Benth.
B. salviifolia
(L.) Lam.
Rank
1
2
3
3
Section
Neemda
Nicodemia
Origin
SE Asia
Madagascar
Neemda
Neemda
SE Asia
SE Asia
5
6
Neemda
Neemda
S America
SE Asia
7
8
Neemda
N America
Neemda
Neemda
SE Asia
India
10
10
Chilianthus
S Africa
12
Neemda
S Africa
13
Neemda
S Africa
14
The disintegration of the Scrophulariaceae and the biological control of Buddleja davidii
(Scroph I.)] and Glossostigma elatinoides Benth. ex
Hook. f. [Phrymaceae (Scroph IV)], but adult weevils
did not oviposit on these species.
Most exotic scrophularia appeared to be immune to
attack by the weevils, but larval and adult feeding and
oviposition occurred on the weedy European species
Verbascum thapsus L., Verbascum virgatum Stokes and
Scrophularia auriculata L.
Discussion
Cleopus Dejean, belonging to a tribe (Cionini) of hostspecific figwort weevils and the European representatives (Cionus Clairville and Cleopus species), feed
on Scrophularia, Verbascum and occasionally on adventive Buddleja (Walker, 1914, Hoffman, 1958; Cunningham, 1974, 1975; Williams, 1974; Read, 1976,
1978; Bullock, 1987; Smith, 1992). Conversely, the
Asian species, C. japonicus, has only been recorded
from B. davidii (Zhang et al., 1993), and this study
found that it could only complete its life cycle on a few
Buddleja taxa, but could feed on Scrophularia and Verbascum. The host associations of these species appear
to support the recent revision of the Scrophulariaceae
(Fig. 1).
Other invertebrates are also known to feed exclusively on these plant species, which have been recognized as a distinct clade, Scrophulariaceae s.s. [Scroph
I of Olmstead and Reeves (1995) and Olmstead et al.,
(2001)] within the Scrophulariaceae s.l. Allen (1960)
noted that weevils of the Gymnetrini distinguished between the Plantaginaceae (Scroph II) and the Scroph I
clade and that they fed indifferently upon Scrophularia
and Verbascum.
Westwood (1849 in Scott 1937) remarked that Cionus scrophulariae L.;
long ago discovered the Natural System, and
proved by the fact of their sometimes indiscriminately
feeding on mulleins [Verbascum] and figworts [Scrophularia] that these plants were in truth closely allied in
Nature.
Scott (1937) records the same weevil feeding on
the introduced Cape figwort Phygelius capensis Benth.
(Scroph I) and notes the observations, by others, of Cionus and Cleopus occasionally feeding on introduced
Buddleja in the UK. In the UK, the only specialist
Lepidopteran to occasionally feed on buddleia is the
mullein moth, Cucullia verbasci L., which normally
has the same host range as the figwort weevils (Owen
and Whiteway, 1980). The flea beetles, Longitarsus
spp. Latreille, also only have the hosts Scrophularia
and Verbascum, and in summary of these observations,
Allen (1960) stated: Yet none of these insects, apparently, is known ever to attack Lanaria or Antirrhinum
[both of Antirrhineae, Scroph II ] in a state of nature,
and I am aware of no instance of a Linaria feeder (of
which there are many) having Scrophularia as a host.
Elements of the distinctive iridoid and terpenoid
phytochemistry of Buddleja have been shown to be
Buddlejeae (Af., Asia, N.Am. Buddleja***)
Teedieae (S.Af.)
(Phygelius*)
Scrophularieae (NH, Scrophularia** & Verbascum**)
Limoselleae (S.Af.)
Leucophylleae (C.Am.)
Myoporeae (Aust.)
Aptosimeae (Af.)
Hemimerideae (S.Af.)
Figure 1.
Summary of the phylogenetic relationships among the tribes and the unresolved genus, Phygelius, of
the Scrophulariaceae sensu stricto (after Tank et al., 2006). Low (single asterisk) to high level (triple
asterisk) of feeding by Cleopus japonicus.
289
XII International Symposium on Biological Control of Weeds
biologically active (Yoshida et al., 1976; Houghton
et al., 2003) and may well influence the invertebrate
feeding guild associated with this and closely related
genera. Iridoids are known to be feeding stimulants for
specialist Lepidoptera (Bowers, 1988) and deterrents
for generalists (Stephenson, 1982).
The study reported in this paper not only confirms
the restricted host preference of the Cionini and demonstrates the low risk of C. japonicus to the New Zealand
flora but also supports the current position of Buddleja
within the new taxonomy. Bearing in mind that laboratory studies are thought to overestimate the host range of
potential biological control agents (Hill, 1999), the host
range of C. japonicus in the field may be more limited.
However, these results cannot preclude the possibility that, if released, C. japonicus could produce self-
sustaining populations on Verbascum, Scrophularia and
Buddleja species. Four other species of Buddleja [B.
salviifolia L., Buddleja madagascariensis Lam., Buddleja globosa Hope and B. dysophylla (Benth.) Radlk.]
have already partially naturalized in New Zealand. The
early flowering B. salviifolia is valued as a spring nectar source for bees (Kay and Smale, 1990). However,
C. japonicus adults fed poorly, larvae failed to feed and
no eggs were laid on B. salviifolia. It is unlikely that
this species would be colonized by C. japonicus. In
contrast, B. madagascariensis is one of a number of
Buddleja species to be considered strongly invasive on
the west coast of USA and Hawaii (Randall and Marinelli, 1996). It ranked highly as a host of C. japonicus.
B. globosa is not considered to be invasive, and any attack by C. japonicus may limit the potential ornamental value of this species. B. dysophylla appeared to be a
poor host for C. japonicus, but it is rarely cultivated in
New Zealand and is not considered invasive.
C. japonicus appears to have a restricted host range
and could be expected to be at least as host-specific as
its European congeners. We conclude that it is possible
that C. japonicus may cause minor damage to some
non-target scrophulareous species, particularly within
Buddleja, Verbascum and Scrophularia. These are all
weedy species or ornamentals of minor importance in
New Zealand, but in Europe, the tribe Scrophularieae
exhibits considerable radiation, resulting in 400500
species of Scrophularia and Verbascum. A number of
these are rare or endangered (Wigginton, 1999), and
rigorous testing would be advisable.
References
Allen, A. A. (1960) Foodplants of Gymnetrini (Col., Curculionidae), etc., as an indication of botanical affinities.
Entomologists Monthly Magazine 96, 48.
Bellingham, P.J., Peltzer, D.A. and Walker, L.R. (2005) Contrasting impacts of a native and an invasive exotic shrub
on floodplain succession. Journal of Vegetation Science
16, 135142.
Bowers, M.D. (1988) Chemistry and coevolution: iridoid
glycosides, plants and herbivorous insects. In: Spencer,
K.D. (ed.) Chemical Mediation of Coevolution. Academic
Press, New York, pp. 133165.
Bullock, J.A. (1987) Cionus scrophulariae (L.), (Col., Curculionidae) feeding on Buddleja globosa Hope. Entomologists Monthly Magazine 123, 190.
Crawley, M.J. (1987) What makes a community invasible?
In: Gray, A.J., Crawley M.J. and Edwards, P.J. (eds) Colonisation, Succession and Stability. Blackwell Scientific,
London, UK, p. 429.
Cunningham, P. (1974) Studies on the occurrence and distribution of the genera Cionus and Cleopus (Col.: Curculionidae) in South Hampshire, 1973. Entomologists
Record 86, 184188.
Cunningham, P. (1975) A convenience food for weevils of
the genera Cionus and Cleopus (Col., Curculionidae).
Entomologists Monthly Magazine 112, 13401343.
Esler, A.E. (1988) The naturalisation of plants in urban Auckland, New Zealand. Success of the alien species. New
Zealand Journal of Botany 26, 565584.
Hill, R.L. (1999) Minimising uncertainty in support of
no-choice tests. In: Withers, T.M., Barton-Browne, L.,
Stanley, J. (eds) Host-Specificity Testing in Australasia:
Towards Improved Assays for Biological Control. Scientific Publishing, Indooroopilly, p. 1.
Hoffman, A. (1958) Faune de France, vol. 62. Coloptres
Curculionides (Troisime Partie). Fdration Franaise
des Socits de Sciences Naturelles, ditions Paul Lechevalier, Paris, pp. 12111233.
Holm, L., Pancho, J.V., Hergerger, J.P. and Plucknett, D.L.
(1979) A Geographical Atlas of World Weeds. Wiley, New
York, 391 pp.
Houghton, P.J., Mensah, A.Y., Iessa, N., and Hong, L.Y.
(2003) Terpenoids in Buddleja: Relevance to chemosystematics, chemical ecology and biological activity. Phytochemistry 64, 385393.
Kay, M.K. and Smale, M.C. (1990) The potential for biological control of Buddleia davidii Franchet in New Zealand.
In: Bassett, C., Whitehouse, L.J. and Zabkiewicz, J.A.
(eds) Alternatives to the Chemical Control of Weeds. Ministry of Forestry, FRI Bulletin 155, pp 2933.
Leeuwenberg, A.J.M. (1979) The Loganiaceae of Africa
XVIII Buddleja L.II: Revision of the African and Asiatic
species. Mededelingen.Landbouwhogeschool Wageningen 6, 1163.
Olmstead, G.R. and Reeves, P.A. (1995) Evidence for the
polyphyly of the Scrophulariaceae based on chloroplast
rbcL and ndhF sequences. Annals of the Missouri Botanical Garden 82, 176193.
Olmstead, G.R, de Pamphilis, C.W., Wolfe, A.D., Young,
N.D., Elisons, W.J and Reeves, P.A. (2001) Disintegration
of the Scrophulariaceae. American Journal of Botany 88,
348361.
Owen, D.F. and Whiteway, W.R. (1980_ Buddleja davidii in
Britain: History and development of an associated fauna.
Biological Conservation 17, 149155.
Randall, J.M. and Marinelli, J. (1996) Invasive Plants: Weeds
of the Global Garden. Brooklyn Botanic Gardens, Brooklyn, NY.
Read, R.W.J. (1976) Notes on the biology of Cleopus pulchellus Herbst (Coleoptera: Curculionidae). Entomologists Gazette 27, 11981220.
290
The disintegration of the Scrophulariaceae and the biological control of Buddleja davidii
Read, R.W.J. (1978) Notes on the biology of Cionus scrophulariae (L), together with preliminary observations on C.
tuberculosus (Scopli) and C. alauda (Herbst) (Col., Curculionidae). Entomologists Gazette 28, 183202.
Richardson, B., Vanner, A., Ray, J., Davehill, N. and Coker,
G. (1996) Mechanisms of Pinus radiata growth suppression by some common forest weed species. New Zealand
Journal of Forestry Science 26, 421437.
Scott, H. (1937) Notes on Cionus scrophulariae (L.) infesting
a South African plant, Phygelius capensis E. Entomologists Monthly Magazine 73, 2934.
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006) Top 20
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 93117.
Smale, M.C. (1990) Buddleia a growing weed problem
in protected areas. Whats New in Forest Research 185,
14.
Smith, K.G.V. (1992) Cionus scrophulariae (L.), (Col., Curculionidae) feeding on Buddleja globosa Lam. Entomologists Monthly Magazine 128, 254.
Stephenson, A.G. (1982) Iridoid glycosides in the nectar of
Catalpa speciosa are unpalatable to nectar thieves. Journal of Chemical Ecology 8, 10251034
Stevens, P.F. (2001). Angiosperm Phylogeny Website, Version 6. Available at: http://www.mobot.org/MOBOT/
research/APweb/ (accessed May 2005).
Tank, D.C., Beardsley, P.M., Kelcher, S.A. and Olmstead,
R.G. (2006) Review of the systematics of Scrophulariaceae s.l. and their current disposition. Australian Systematic
Botany 19, 289307.
Walker, J.J. (1914) Species of Cionus on Buddleia globosa.
Entomologists Monthly Magazine 50, 248.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
Wigginton, M.J. (ed) (1999) British Red Data Books I. Vascular Plants. JNCC, Peterborough, UK, 200 pp.
Williams, S.A. (1974) Two species of Cionus (Col., Curculionidae) on Buddleja davidii. Entomologists Monthly
Magazine 110, 63.
Yoshida, T., Nobuhara, J., Uchida, M. and Okuda, T. (1976)
Buddledin A, B and C, Piscidal sesquiterpenes from
Buddleja davidii Franchet. Tetrahedron Letters 41, 3717
3720.
Zhang, X., Xi, Y., Zhou, W. and Kay, M. (1993) Cleopus japonicus, a potential biocontrol agent for Buddleja davidii in
NZ. New Zealand Journal of Forestry Science 23, 7883.
291
Quarantine evaluation of
Eucryptorrhynchus brandti (Harold)
(Coleoptera: Curculionidae), a potential
biological control agent of tree of
heaven, Ailanthus altissima,
in Virginia, USA
L.T. Kok, S.M. Salom, S. Yan, N.J. Herrick and T.J. McAvoy1
Summary
Tree of heaven, Ailanthus altissima (Mill.) Swingle, is an imported invasive weed tree from China
that has become established throughout much of the continental USA. It colonizes disturbed forest
sites and often out-competes native vegetation. Short-term cultural and chemical controls of this
weed are expensive and have limited efficacy. Eucryptorrhynchus brandti (Harold) and Eucryptorrhynchus chinensis (Olivier), two curculionid species, are pests of A. altissima in China and have no
other known hosts. The objectives of our project are to (1) assess the pest status of A. altissima in
Virginia and (2) evaluate E. brandti for its potential as a biological control agent. A statewide survey
showed significant presence of tree of heaven but no native herbivores with potential of controlling
it, suggesting biological control to be an attractive method of management. As E. brandti requires
live trees for development, quarantine studies have focused on developing a rearing technique and
testing the host specificity on native plants approved by the Technical Advisory Group for Biological
Control Agents of Weeds. Preliminary results indicate that E. brandti feeds only on tree of heaven,
with greatly reduced feeding observed on corkwood, Leitneria floridana Chapman, and paradise tree,
Simarouba glauca DC.
Keywords: rearing, Eucryptorrhynchus brandti, host specificity testing, natural enemy.
Introduction
Ailanthus altissima, tree of heaven, is an introduced
species in Europe (Ballero et al., 2003; Lenzin et al.,
2004), Africa, South America and North America (Ding
et al., 2006). Seeds were introduced from China to Paris
between 1740 and 1750 (Hu, 1979; Tellman, 2002) and
in North America as an ornamental shade tree during
the late 18th century from Europe into Philadelphia,
Pennsylvania (Feret, 1985; Tellman, 1997). Multiple
introductions into New York occurred during the early
19th century (Davies, 1942; Dame and Brooks, 1972;
Hu, 1979).
Virginia Polytechnic Institute and State University, Blacksburg, Department of Entomology, VA 24061-0319, USA.
Corresponding author: L.T. Kok <[email protected]>.
CAB International 2008
1
Currently, tree of heaven is found in 41 of the lower
48 continental USA from Washington to New England
and south to Florida, Texas and Southern California
(USDA-NRCS Plants Database, 2007). It is often used
as an ornamental adjacent to sidewalks, streets and in
parking lots. In Virginia, tree of heaven is a dominant
species along roadsides and occupies hundreds of acres
in the Shenandoah National Park (Marler, 2000).
Tree of heaven has many beneficial attributes and
often is regarded as an important ornamental species
in the countries of origin because of its aesthetic value
and ability to withstand environmental pollutants and
water stress in an environment caused by human activities (Ding et al., 2006). Traditional Chinese culture has
used tree of heaven for its anti-tumour properties (Ammirante et al., 2006). In its native range, tree of heaven
is fed upon by more than 40 phytophagous arthropod
species and is susceptible to nearly 20 pathogens (Ding
292
Quarantine evaluation of Eucryptorrhynchus brandti
et al., 2006). Few pathogens occur in the USA, and
only one has caused isolated fatality of an individual
tree of heaven plant (Ding et al., 2006). None of the
pathogens occurring in the USA appears to be specific
to tree of heaven (Ding et al., 2006). Tree of heaven
produces allelopathic compounds capable of inhibiting
the growth of nearly 90 tree species (Mergen, 1959;
Heisey, 1990a,b; Lawrence et al., 1991; Heisey and
Heisey, 2003). The lack of natural enemies of tree of
heaven and its ability to suppress plant growth over a
wide range of habitats allow it to out-compete native
flora in North America.
Chemical control of tree of heaven is the most common control method. Herbicides registered for tree of
heaven control are dicamba, glyphosate, imazapyr,
metasulfuron methyl and triclopyr. They are applied
as foliar sprays, basal-bark treatments, injection or applied to cut stumps. However, chemicals often cause
treated sites to be barren of any plant life resulting in
the re-intrusion of invasive species like tree of heaven.
There is a concern that continued large-scale herbicide
applications may become detrimental to the environment in addition to being labor intensive and costly. For
example, at a heavily infested 4-ha site on the median
of interstate 81 in Virginia, the Virginia Department of
Transportation estimated that the cost to control tree of
heaven was $8750 per ha and produced only reasonable control.
The lack of natural enemies of tree of heaven in the
USA and the potential for biological control led to foreign exploration to identify potential biocontrol agents
in China. A survey of the Chinese literature suggested
that Eucryptorrhynchus brandti Harold and Eucryptorrhynchus chinensis (Olivier) (Coleoptera: Curculionidae) would be potential agents to investigate (Ding et
al., 2006). E. brandti is a univoltine beetle species native to China where it is considered a pest. In some
areas of China, 80% to 100% of tree of heaven trees
surveyed were attacked by E. brandti and E. chinensis causing 12% to 37% mortality (Ge, 2000; Ding et
al., 2006). Chinese people make considerable efforts
to control E. brandti with chemicals and have found a
nematode (Steinernema feltiae) that can produce 70%
mortality in the field (Dong et al., 1993; Jianguang et
al., 2004). The general biology of E. brandti is not well
known. However, its development is probably similar to
that of other curculionid species associated with woody
trees (Barrett, 1967). Adults feed on leaves, buds and
petioles (Ding et al., 2006). Larvae develop under the
bark and emerge as adults, leaving round emergence
holes approximately 4 mm in diameter.
Our goal was to identify insect herbivores that can
reduce the spread of tree of heaven. Specific objectives
were to (1) survey for native insect herbivory on A. altissima in different regions of Virginia to identify any
species with potential to impact tree of heaven and (2)
import selected herbivores from the native habitat of A.
altissima for host-specificity evaluation under quaran-
tine to determine their potential for survival and development in Virginia.
Materials and methods
Tree of heaven in Virginia
Locations distributed throughout three different
regions of Virginia were identified to determine the
extent of A. altissima colonization. The regions were
Ridge and Valley (Appalachian Mountains), Piedmont
and Coastal Plain (Fig. 1). Within each region, at least
two sites each along highways, rights-of-way or natural forest disturbance areas were selected. At each site,
all tree species and their percent cover within the tree
of heaven infestation were recorded. Observations of
herbivores feeding on tree-of-heaven were done by visually examining at least 200 leaflets for herbivores.
Herbivores were also collected by beating leaves over
1 m2 beat sheets. Any herbivores found were collected
and identified. Insect sampling also consisted of whole
tree observation for activity and damage from insects.
Whole tree observation included detailed examination
of foliage, stems, buds and seeds (when available).
Where obvious insect damage was detected, insects
were collected if located. Our objective was not to
carry out a biodiversity study or complete census of
insects found on tree of heaven but to focus on insects
colonizing or feeding on tree tissue causing observable
damage. The survey began in the summer of 2004 and
continued in 2005. Each site was visited monthly during the growing season (May to October).
Concurrent with the survey effort was the assessment of impact of identified herbivores on the growth,
reproduction and survival of tree of heaven. When damage associated with insect feeding or colonization was
found, it was followed by more intensive sampling of
the causal agents. The intention was to determine, for
each identified herbivore, the level and timing of activity across all regions and site types. We recognize that
native herbivores are not likely to effectively control
tree of heaven, as the weed has been highly successful
thus far. However, this information will be helpful, as
we evaluate exotic biological control agents and their
potential interaction with our existing fauna.
Quarantine testing of an exotic weevil
imported from China
As part of a collaborative project with Dr. Ding
Jianqing, Biological Control Institute, Chinese Academy of Agricultural Sciences, two weevil species, E.
brandti and E. chinensis, identified as important pests
of A. altissima, were studied in China. E. brandti was
imported into the US Department of Agricultures
(USDA) Animal and Plant Health Inspection Service
(APHIS) approved Quarantine facility at Virginia Tech
beginning in 2004.
293
XII International Symposium on Biological Control of Weeds
= Survey site
Rocky Bar
20 0 20 40 Kilometers
New Post
Dayton
Mountains
Milton
Piedmont
Blacksburg
Vera
Radford
Nassawadox
York River
State Park
Tidewater
98
99
Figure 1.
Ailanthus altissima (Mill.) Swingle survey sites in Virginia, USA: (1) Mountain, including Radford (forest),
Blacksburg (forest), Dayton (forest) and Rockbar (roadside); (2) Piedmont, including Vera (forest), Milton
(roadside) and New Post (roadside); (3) Coastal plain, including York River State Park (roadside) and Nassawadox (forest).
Survival and development: Adult E. brandti is reared
on tree of heaven foliage and stems in containers at several constant temperatures (20C to 30C) within the
range that the species will encounter in the US release
areas. Biological data recorded included survival and
development times of the egg, larval and adult stages,
as well as fecundity of the females and egg hatch rates.
Two colonies were established; one colony was maintained for production and the second used for biological and host specificity studies.
Production colony: This colony has been maintained
and caged in screened plastic boxes (30 15 165 cm)
at 22C. Groups of five males and five females were
placed in the cages. In each cage, tree of heaven billets,
with the upper ends sealed with heated paraffin to reduce desiccation, were provided for oviposition and foliage added for adult feeding. Vermiculite was added to
the bottom of the cage and kept moist. Weekly, billets
were removed and placed in separate cages for larval
development. Newly emerged adults were transferred
to new oviposition cages.
Fecundity: Single pairs of male and female were caged
in screened plastic boxes, as in the production colony
cages, and replicated. Tree of heaven billets were added
and checked for oviposition daily by removing the
bark and examining the cambium for eggs. After ini-
tiation of oviposition, the bark was checked daily for
eggs.
Egg and larval development: Eggs recovered from the
billets were placed in Petri dishes with moistened filter
paper and reared at several constant temperatures and
checked daily for hatching. Newly hatched larvae were
inoculated in tree of heaven billets by drilling a 7-mm
diameter hole into the cambium and inserting one larva
into the hole. Billets were checked at an appropriate
time after inoculation to determine when development
to the second instar occurred. This procedure was repeated at least three times for subsequent instars until
pupation.
Physiological host specificity testing: Host-range
studies included a series of no-choice and choice feeding and oviposition tests. Test plant species were from
three groups, with a total of 30 species: 18 taxonomically related species (Simaroubaceae, Meliaceae and
Rutaceae), six ecologically related species, and six eco
nomically related species. The taxonomically related
species were chosen based on Wapsheres method of
centrifugal phylogentic testing (Wapshere, 1974). This
method involves exposing the biological control agent
to a sequence of plants from those most closely related
to the target species progressing to successively more
and more distantly related plants until the host range
294
Quarantine evaluation of Eucryptorrhynchus brandti
Table 1.
Plant species to be tested for their suitability as hosts of Eucryptorrhynchus brandti (Harold). Species are listed with the most closely
related listed first and the most distant last.
Family
Simaroubaceae
Picramniaceae
Meliaceae
Species
Chinese Ailanthus altissima (Mill.)
Swingle
Simarouba glauca DC
Simarouba tulae Urban
Leitneria floridana Chapman
Castela emoryi (Gray) Moran
& Felger
Castela erecta Turp.
Paradise tree
Aceitillo falso
Corkwood
Crucifixion thorn
Cockspur, goat-bush,
retama, rupagita
Holacantha stewartii C. H. Muell.
Stewart crucifixion thorn
Alvaradoa amorphoides Liebm.
Mexican alvaradoa
Picramnia pentandra Sw.
Florida bitterbush
Swietenia mahagoni (L.) Jacq.
West Indian mahogany
Citrus aurantifolia (Christm.) Swingle Lime
Citrus aurantium L.
Sour orange
Citrus limon (L.) Burm. F.
Lemon
Citrus paradisi Macfad.
Grapefruit
Citrus reticulate Blanco
Tangerine
Citrus sinensis Osbeck
Sweet orange
Ptelea trifoliate L.
Common hop tree
Northern prickly-ash
Zanthoxylum
americanum Mill.
is thoroughly determined. The ecologically associated
species were chosen based on our observations in Virginia. This species list is currently under review by the
Technical Advisory Group (TAG) USDA/APHIS. A
suitable host is defined as a plant species that will support feeding, oviposition and larval development of the
insect to the adult stage, and the latter is capable of producing viable progeny. If any of these four conditions
Table 2.
Common name
Tree of heaven
Economically important and ecologically associated plant species to be tested.
Family
Species
Economically important
Aceraceae
Acer rubrum L.
Fagaceae
Quercus alba L.
Quercus rubra L.
Juglandaceae
Carya glabra
(Mill.) Sweet
Juglans nigra L.
Magnoliaceae
Liriodendron
tulipifera L.
Ecologically associated
Anacardiaceae Rhus typhina L.
Cuppressaceae Juniperus
virginiana L.
Leguminosae
Robinia
pseudoacacia L.
Pinaceae
Pinus virginiana Mill.
Rosaceae
Crataegus spp.
Prunus serotina Ehrh.
Common name
Red maple
White oak
Red oak
Pignut hickory
Black walnut
Tulip poplar
Staghorn sumac
Eastern redcedar
Black locust
Virginia pine
Hawthorne
Black cherry
is not met, the insect will not be able to survive and is
therefore not able to sustain a population solely on the
plant species in question.
Feeding tests: Foliage and billets of the species listed
in Tables 1 and 2 were used in choice and no-choice
tests to date. Tests were conducted for 7 days at 20C.
No-choice. Leaf clusters of each test plant species
were caged with three adults. The amount of feeding
was determined by placing a transparent millimeter
square grid over the fed upon portion of the leaves. The
control cage maintained at the same time with tree of
heaven was without weevils.
Choice. Two series of tests were conducted, one with
tree of heaven in the cage and one without. In the first
test series, one cluster of tree of heaven leaves together
with leaf clusters of one or more non-target species was
placed in cages with adult beetles. In the second series,
two or more non-target test species based on availability were placed in the cage with no tree of heaven. The
available target species were selected randomly. Feeding was recorded as described above.
Oviposition tests: No-choice. Billets of one test plant
species were placed in cages with one gravid female
and compared with control billets of tree of heaven
without weevils. The cambium was examined for eggs
after 1 week. Foliage of the same species as the billet
was placed in the cage with the adults for food.
Choice. Two or more species of billets were placed
in cages with one gravid female. Two series of tests
were carried out, with tree of heaven plus one or more
other species in the cage and one with only a non-target
295
Table 3.
Tree species composition and their percent coverage (%) at nine survey sites in Virginia, USA.
296
Ailanthus altissima
(P. Mill.) Swingle
Quercus palustris Muenchh
Pinus virginiana P. Mill.
Juniperus virginiana L.
Pinus taeda L.
Robinia pseudoacacia L.
Liquidambar styraciflua L.
Pinus strobi L.
Rhus glabra L.
Prunus virginiana L.
Liriodendron tulipifera L.
Acer negundo L.
Ilex opaca Ait.
Juglans nigra L.
Acer rubrum L.
Ulmus americana L.
Acer saccharinum L.
Mountain
Piedmont
Coastal plain
Mean
Dayton
RockBar
Blacksburg
Radford
New Post
Milton
Vera
Nassawadox
15
85
80
30
35
70
15
20
York River
State Park
15
25
60
10
15
95
10
15
10
20
15
20
10
10
15
10
10
10
15
25
15
15
10
20
5
20
10
20
50
15
5
5
15
10
40
15
30
40
20
32
14
11
6
4
5
7
7
4
6
4
3
3
3
2
1
1
XII International Symposium on Biological Control of Weeds
Species
Quarantine evaluation of Eucryptorrhynchus brandti
test plant species. Foliage was added for food. After 1
week, the billets were checked for eggs.
Table 5.
Mean (SD) number of webworm, Atteva punctella Cramer, including eggs, larvae, pupae and
adults, in 2004 and 2005, at three survey regions: Piedmont, Mountain and Coastal Plain.
Month
Region
Results
Survey of VA and impact assessment
of native herbivores
2004
June
July
Aug.
Sept.
2005
May
June
July
Aug.
Sept.
Seventeen tree species were found to co-exist with
tree of heaven in Virginia, with coverage of tree of
heaven ranging from 15% to 85% (Table 3). The common associates found with tree of heaven were Quercus spp., observed in six of the nine survey sites, with
coverage ranging from 5% to 60%. Twenty insect herbivore species were collected from tree of heaven in
2004 and 2005 (Table 4). The 12 beetles (Coleoptera)
in Table 4 had little impact on tree of heaven. Their potential as biological control agents is minimal because
of their low abundance and broad host range. The only
Table 4.
Insect herbivores found on Ailanthus altissima
(Mill.) Swingle in 2004 and 2005 from nine survey sites in Virginia, USA.
Herbivores
Coleoptera
Odontota dorsalis
(Thunberg)
Chrysomelidae spp.
Bruchid spp.
Popillia japonica
Newman
Neotrichophorus spp.
Chrysolina
quadrigemina Suffrian
Apion spp.
Merhynchites spp.
Sphenophorus spp.
Orchestes spp.
Scolytinae spp.
Lepidoptera
Ectropis crepuscularia
D. and S.
Thyridopteryx
ephemeraeformis
(Haworth)
Atteva punctella (Cram.)
Saturniidae spp.
Hemiptera
Empoasca sp.
Anormenis sp.
Acanalonia sp.
Orthoptera
Scudderia furcata
Brunner
Common Name
Number
individuals/
site
Locust leaf
miner
Leaf beetle
Seed beetle
Japanese beetle
0.7
Click beetle
Flea beetle
5.2
0.3
Weevil
Leaf rolling
weevil
Snout beetle
Weevil
Ambrosia beetle
<0.1
<0.1
The small
engrailed
Bagworm
2.5
1.3
3.4
0.3
<0.1
<0.1
<0.1
<0.1
<0.1
Ailanthus
webworm moth
Silkworm moths
<0.1
Leaf hopper
Plant hopper
Plant hopper
0.7
2.3
1.2
Katydid
0.5
>30
Mountain
Piedmont
Coastal Plain
4.0 1.4
11.5 8.9
46.5 17.0
36.3 30.4
37
54.3 17.9
91.3 15.0
21.7 13.0
28
37.5 26.2
58.0 2.8
5.0 5.7
2.0 2.8
6.5 7.5
12.0 5.7
12.5 10.6
102.5 88.4
0.3 0.6
17.0 23.6
111.0 163.7
46.3 55.6
106.0 112.2
0
21.0 5.7
75.0 19.8
84.5 50.2
52.5 31.8
abundant Coleoptera herbivore that may be causing serious damage to tree of heaven are the ambrosia beetles
Euwallacea validus (Eichoff) and Xyleborus atratus
Eichoff. These emerged from dying tree of heaven.
We suspect that these species only attacked the dying or dead trees and had little effect on healthy tree
of heaven. Based on our observations of herbivores in
Virginia in 2004 and 2005, these herbivores had a negligible impact on tree of heaven.
Ailanthus webworm, Atteva punctella Cramer,
was the only herbivore consistently present in all sites
with a total of over 30 (eggs, larvae, pupae and adults)
per visit. A. punctella caused >50% defoliation for 1year-old seedlings. However, its effect on larger trees
(>3 cm diameter) was minimal, causing less than 5%
defoliation with no visible impact. The population of
this species peaked around August (Table 5) with no
significant difference among the three geographic regions [F(22,11) = 0.63, p = 0.83) Two other insect species
have been reported to feed on tree of heaven foliage:
Cynthia moth, Samia cynthia (Drury), and the Asiatic
garden weevil, Maladera castanea (Arrow) in eastern
USA (Drooz, 1985). However, their presence was not
identified in this survey, and it is unlikely that these
two insect species will have any impact on the tree of
heaven in Virginia.
Quarantine testing of E. brandti imported
from China: Development and rearing of
E. brandti in the laboratory
A total of 500 and 1200 E. brandti adults were received from our cooperator in China in 2005 and 2006,
respectively. In 2005, we initiated a study to evaluate
the optimum conditions for rearing E. brandti. Eighteen adults (nine females, seven males and two unsexed) emerged from tree of heaven billets in Spring
297
XII International Symposium on Biological Control of Weeds
Table 6.
No choice foliage feeding tests of Eucryptorrhynchus brandti (Harold) adults on
target and test species.
Family
Species
Common name
X SD
N
(mm2 per
adult per day)
Simaroubaceae
Tree of heaven
9
Ailanthus altissima (Mill.)
56.4 21.0aa
Swingle
8.5 0.4.8 b
Paradise tree
9
Simarouba glauca DC
9 21.0 9.2 b
Leitneria floridana Chapman Corkwood
Rutaceae
Lime
3
0c
Citrus aurantifolia
(Christm.) Swingle
Sour orange
3
0c
Citrus aurantium L.
Red maple
Aceraceae
4
0c
Acer rubrum L.
Staghorn sumac
Anacardiaceae
4
0c
Rhus typhina L.
Eastern redcedar
Cupressaceae
4
0c
Juniperus virginiana L.
White oak
Fagaceae
2
0c
Quercus alba L.
Red oak
2
0c
Quercus ruba L
Pignut hickory
Juglandaceae
4
0c
Carya glabra (Mill.) Sweet
Black walnut
2
0c
Juglans nigra L.
Black locust
Leguminosae
4
0c
Robinia pseudoacacia L.
Tulip poplar
Magnoliaceae
4
0c
Liriodendron tulipifera L.
Virginia pine
Pinaceae
2
0c
Pinus virginiana Mill.
Rosaceae
Hawthorne
4
0c
Crataegus spp.
Black cherry
2
0c
Prunus serotina Ehrh.
Means within a column followed by different letters are significantly different at P < 0.05, TukeyKramer
multiple comparison test.
2006, indicating that the weevil could complete its life
cycle in a cut tree of heaven log.
At approximately 20C, most E. brandti developed
from egg to adult in 3 months. However, some individuals did not complete within 9 months and were still
larvae after 9 months.
E. brandti did not expel frass from the billet. Frass
remained in the billet within the feeding galleries. This
made it difficult to know where the weevils were located and their life stage without dissecting the billet.
Only 15% (18/119) of the weevils completed their life
cycle and emerged. A few weevils that completed development failed to emerge, possibly due to poor food
quality and quantity. Three young larvae were removed
from one billet and transferred to another by inserting
Table 7.
them into a 7-mm diameter hole. Frass was observed 3,
6, 9 and 11 days after the transfer, and two larvae completed development to the adult stage. They developed
into adults but died inside the tunnel. This suggests that
transfer of larvae into an artificially drilled hole has the
potential to be used as a bioassay to test larval development on non-target species.
Based on the above observations, we developed a
rearing procedure in 2006. Live tree of heaven trees
were periodically cut into 1-m lengths, with diameter
ranging from 10 to 22 cm. One end was treated with
paraffin and the other end was placed in a 5-cm water
bath to help maintain viable phloem tissue as long as
possible. Groups of four billets were placed in a cage
together with up to 150 weevils (male and female) for
Two choice foliage feeding tests of Eucryptorrhynchus brandti (Harold) adults on target and test plant
species.
Family
Species
Common name
Test species
A. altissima
SD (mm per
X
adult per day)
SD (mm2 per
X
adult per day)
0.9 1.7a
2.7 1.9a
0a
0a
0a
0a
0a
40.4 15.0a
41.6 18.5a
26.6 1.6a
28.7 16.8a
22.3 16.8a
27.0 5.0a
26.0 8.4a
Simaroubaceae
Aceraceae
Magnoliaceae
Anacardiaceae
Leguminosae
Rosaceae
a
Simarouba glauca DC
Leitneria floridana Chapman
Acer rubrum L.
Liriodendron tulipifera L.
Rhus typhina L.
Robinia pseudoacacia L.
Prunus serotina Ehrh.
Paradise tree
Corkwood
Red maple
Tulip poplar
Staghorn sumac
Black locust
Black cherry
9
9
4
4
4
4
2
Denotes significant differences (P 0.05) between Ailanthus altissima and the test species (Students t test).
298
Quarantine evaluation of Eucryptorrhynchus brandti
2 to 3 weeks. The billets and developing larvae were
maintained at 22C. After 2 months, the billets were
checked daily to capture emerging adults. Through
2006, 523 F1 generation adults emerged. Generally,
their size was smaller than weevils shipped from China.
As we refine this rearing method, we will focus on using fewer weevils in the oviposition cages to reduce
competition among developing larvae in the billets.
Adult feeding test
Of the species tested, feeding by E. brandti adults
occurred only on species in Simaroubaceae. Tree of
heaven was highly favored for feeding over the nontarget species tested (Table 6). In no-choice tests, adults
consumed nearly seven and three times more of tree
of heaven foliage than S. glauca and L. floridana, respectively. No feeding occurred on 15 other non-target
species. In two-choice tests, an even less non-target
species foliage was consumed. When given a choice of
tree of heaven and S. glauca or L. floridana, E. brandti
adults consumed 48 and 15 times more tree of heaven
foliage than the two non-target species, respectively.
No feeding occurred on five species outside of Simaroubaceae (Table 7).
Figure 2.
Billet inoculation assay
Seven E. brandti larvae were inoculated into 7.5-cm
diameter 76-cm long billets of tree of heaven, Robinia pseudoacacia, Prunus serotina and Citrus aurantifolia to determine if young larvae (<30 days old,
approximately second instar) can survive on plants
other than tree of heaven. These preliminary assays
were replicated twice, and larvae remained in billets for
30 days. Our results indicate a high level of specificity
for tree of heaven. No feeding occurred in species other
than tree of heaven, resulting in 100% larval mortality (Fig. 2). This is compared with 36% larval survival
in A. altissima past the instar when the larvae were
inoculated.
Conclusions
The survey work in Virginia helped characterize tree
species associated with tree of heaven, with Quercus
spp. being the predominant associate regardless of region. The insects found feeding on tree of heaven were
of inconsequential value in terms of damaging the weed
tree and contributing to its overall control. Rearing
studies have improved to the point that a continuous
Example of Eucryptorrhynchus brandti (Harold) larval galleries in each of 4 species tested. Note that galleries
were created only in Ailanthus altissima (Mill.) Swingle (photo credit, N. Herrick).
299
XII International Symposium on Biological Control of Weeds
colony of E. brandti is possible. The use of billets with
one end sealed with paraffin and the other sitting in
water may be an adequate bioassay for larval survival
tests. Preliminary results from the quarantine studies
showed that the risk of this beetle attacking non-target
trees is minimal. Continued testing is ongoing.
References
Ammirante, M., Giacomo, R.D., Martino, L.D., Rosati,
A., Festa, M., Gentillela, A., Pascale, M.C., Belisario,
M.A., Leone, A., Turco, M.C. and Feo, M.C. (2006)
1-Methoxy-canthin-6-one induces c-jun NH2-terminal
kinase-dependent apoptosis and synergizes with tumor
necrosis factor-related apoptosis-inducing ligand activity
in human neoplastic cells of hematopoietic or endodermal
origin. Cancer Research 66, 43854393.
Ballero, M., Ariu, A. and Falagiani, P. (2003) Allergy to
Ailanthus altissima (tree of heaven) pollen. Allergy 58,
532533.
Barrett, J.H. (1967) The biology, ecology and control of
Vanapa oberthuri Pouill. (Coleoptera: Curculionidae) in
hoop pine Araucaria plantations in New Guinea. Papua
New Guinea Agricultural Journal 19, 4760.
Dame, L.L. and Brooks, H. (1972) Handbook of the trees of
New England. Dover Publications, New York.
Davies, P.A. (1942) The history, distribution, and value of
Ailanthus in North America. Transactions of the Kentucky
Academy of Science 9, 1214.
Ding, J., Wu, Y., Zheng, H., Fu, W., Reardon, R. and Liu,
M.(2006) Assessing potential biological control of the invasive plant, tree of heaven, Ailanthus altissima. Biocontrol Science and Technology 16, 547566.
Dong, Z.L., Gao, W.C., Cao, Q., Shan, J.G., Qi, Q.S., Wang,
W.X., Lei, J.W., Zheng, G. and Zhang, L.H. (1993) Control of weevils damaging Ailanthus trees in Beijing with
steinernematid nematodes. Chinese Journal of Biological
Control 9, 173175.
Drooz, A.T. (1985) Insects of eastern forests. USDA Forest
Service Miscellaneous Publication 1426, 608 pp.
Feret, P.P. (1985) Ailanthus: variation, cultivation, and frustration. Journal of Arboriculture 11, 361368.
Ge, T. (2000) Preliminary study on the biology of Eucryptorrhynchus brandti. Newsletter of Forest Pests 2, 1718.
Heisey, R.M. (1990a) Evidence for allelopathy by tree of
heaven (Ailanthus altissima). Journal of Chemical Ecology 16, 20392055.
Heisey, R.M. (1990b) Allelopathic and herbicidal effects of
extracts from tree of heaven (Ailanthus altissima). American Journal of Botany 77, 662670.
Heisey, R.M. and Heisey, T.K. (2003) Herbicidal effects under field conditions of Ailanthus altissima bark extract,
which contains ailanthone. Plant and Soil 256, 8599.
Hu, S.Y. (1979) Ailanthus. Arnoldia 39, 2950.
Jianguang, L., Zhao, H., and Jie, Y. (2004) Use of ZXX-65
vacuum circulatory fumigation equipment against Eucryptorrhynchus brandti (Harold). Forest Pests and Disease 1, 2004.
Lawrence, J.G., Colwell, A. and Sexton, O.J. (1991) The ecological impact of allelopathy in Ailanthus altissima (Simaroubaceae). American Journal of Botany 78, 948958.
Lenzin, H., Erismann, C., Kissling, M., Gilgen, A.K. and
Nagel, P. (2004) Abundance and ecology of selected
neotypes in the city of Basel (Switzerland). Tuexenia 24,
359371.
Marler, M. (2000) A survey of exotic plants in federal wilderness areas. In: Wilderness Science in a Time of Change
Conference: Widnerness Ecosystems, Threats and Management 1999. U.S. Department of Agriculture, Forest
Service, Rocky Mountain Research Station, pp. 318
327.
Mergen, F. (1959) A toxic principle in the leaves of Ailanthus.
Botanical Gazette 121, 3236.
Tellman, B. (1997) Exotic pest plant introduction in the
American Southwest. Desert Plants 13, 310.
Tellman, B. (2002) Human introduction of exotic species in
the Sonoran Region. In: Tellman, B. (ed.) Invasive exotic
species in the Sonoran Region. University of Arizona
Press, Tucson, Arizona, pp. 2546.
USDA-NRCS Plants Database (2007) Plants Profile. Available
at: http://plants.usda.gov/java/nameSearch?keywordquery=
Ailanthus+altissima&mode=sciname&submit.x=15&
submit.y=10
Wapshere, A. J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
300
The insect fauna of Chondrilla juncea
L. (Asteraceae) in Bulgaria and preliminary
studies of Schinia cognata (L.)
(Lepidoptera: Noctuidae) as a potential
biological control agent
I. Lecheva,1 A. Karova1 and G. Markin2
Summary
Between 2001 and 2005, a survey of rush skeletonweed, Chondrilla juncea L. (Asteraceae), and its
associated insect fauna was conducted in Bulgaria. The weed occurs from sea level to 1200 m mainly
on roadsides (47% of populations encountered), disturbed and abandoned farmlands (40%), as well as
in orchards, vineyards and fields of wheat, roses and lavender. For the first time in Bulgaria, the insect
species associated with the plant were inventoried. A total of 51 insect species were collected, but
only four appeared to be specific to the plant. The most dominant species and the one considered most
promising as a potential biological control agent was the moth, Schinia cognata Fr. (Lepidoptera:
Noctuidae). S. cognata larvae feed on the reproductive parts of the plant, and during development,
one larva can consume 61 to 62 flower buds or seed heads. In Bulgaria, the moth has two generations
that overlap, with maximum population densities in July and August. S. cognata is widely distributed
throughout Bulgaria and was found in high densities in most of the populations of C. juncea studied.
It was not observed attacking any other native plant or cultivated plants, and preliminary host-range
studies of four closely related species indicated that it could only feed and develop on C. juncea. S.
cognata has therefore been selected as a potential biological control agent for possible future introduction in North America.
Keywords: Schinia cognata, host-range testing, rush skeletonweed, distribution.
Introduction
Rush skeletonweed, Chondrilla juncea L. (Asteraceae),
has been accidentally introduced in northwest USA and
adjacent Canada, and in Argentina and Australia and
in all areas, it has become a major noxious weed. In
Australia, a program in the 1970s resulted in the successful introduction of three biological control agents
that soon controlled C. juncea over most of its range
(Cullen and Groves, 1977). A similar program by the
Agricultural Research Service of the US Department
of Agriculture introduced and established the same
three agents in North America and resulted in satisfactory control of the weed in the state of California and
1
Agricultural University of Plovdiv, Faculty of Plant Protection and
Agroecology, 12 Mendeleev Str., Plovdiv 4000, Bulgaria.
2
USDA Forest Service, Rocky Mountain Research Station Forestry Sciences Laboratory, Bozeman, MT, USA.
Corresponding author: I. Lecheva <[email protected]>.
CAB International 2008
some areas of Washington (Piper and Andres, 1995).
However, in the cooler interior states of Oregon, Idaho
and Montana, the three agents have not given effective
control. A new program was implemented to determine
if new and more effective biological control agents
could be found in Eurasia (Markin and Quimby, 1997).
As part of this program, a study on C. juncea-associated
phytophagous insects to identify potential biocontrol
agents was conducted in Bulgaria in 2000 to 2005
(Karova and Lecheva, 2005, Karova, 2006).
Methods and materials
From 2000 to 2005, field visits were made to all regions
of Bulgaria to determine the distribution, main habitats, population densities and phenology of C. juncea.
When stands of C. juncea were encountered, the plants
were visually examined, and all associated insects collected. Samples of flower buds, stems and roots were
also collected and dissected. Individuals collected as
301
XII International Symposium on Biological Control of Weeds
larvae were reared in the laboratory using potted C.
juncea plants or using the appropriate plant part in
plastic cages or Petri dishes, and adults that emerged
were identified.
From this survey, the most common and destructive
insect encountered was the moth, Schinia cognata Fr.
(Lepidoptera: Noctuidae), which was selected for more
detailed studies. Laboratory studies of the development
and feeding behaviour of S. cognata were conducted
at the facility of the Department of Agri-Ecology and
Entomology of the Agricultural University of Plovdiv
(Bulgaria). Observations on the seasonal population
dynamics of S. cognata were conducted on two populations near Plovdiv in central Bulgaria in 2003 and
2004, using sweep netting. The sites were visited eight
times during each summer, and five batches of 100
plants swept with a 37-cm diameter canvas sweep net.
The sampling was carried out primarily to determine
the abundance of S. cognata adults and larvae, but all
other insects collected were also counted to provide
an estimate of their relative abundance (percent of all
insects collected over the 2 years). To determine the
potential host range of S. cognata, other plant species
of Asteraceae growing adjacent to attacked C. juncea
plants were searched for S. cognata larvae.
Laboratory feeding tests were conducted using field
collected early instar larvae or larvae reared from eggs.
We determined the development and feeding impact
(number of reproductive structures consumed) of larvae held on potted plants and bouquets of plants, covered by transparent plastic screen cages. Choice feeding
tests used bouquets of C. juncea intermixed with a test
plant with their bases wrapped in cotton, placed in small
tubes of water and held in plastic screen cages. For
no-choice feeding tests, the reproductive parts of the
plant being tested were offered to larvae held in a 20
1.5 cm glass Petri dish. Reproductive parts were replaced when consumed or wilted until either the larvae
pupated or died. All choice and no-choice tests were
replicated ten times using one larva each. For these
preliminary host tests, four locally available species of
closely related Asteraceae were used.
Results and discussion
Distribution, habitats, population density
and phenology of C. juncea
C. juncea is widespread throughout Bulgaria and was
found in 23 of the 26 regions (=states; Karova and Lecheva, 2005). It occurs mainly on roadsides and disturbed
lands and also in orchards, grape vineyards and fields of
wheat, lavender and roses (Table 1). It was found from 0
to 1200 m above sea level. The densest populations (50
plants/m2) were observed on abandoned farmland around
the cities of Plovdiv, Varna, Ihtiman and Dupnitza and
nearly as dense populations (25 to 50 plants per square
meter) in Blagoevgrad. Other plants occurring with C.
Table 1.
Frequency distribution by habitat of 84 stands of
Chondrilla juncea found in a survey of Bulgaria
between 2001 and 2005 (Karova and Lecheva,
2005; Karova, 2006).
Habitats in which
Chondrilla juncea
was found
Road sides
Abandon farmland
Fallow wheat fields
Grape vineyards
Fields of rosesa
Fields of lavendera
Other locations
Percent of stands
46.75
39.4
4.25
4.25
1.06
1.06
4.27
Fields of ornamental roses (Rosa spp.) (Rosaceae) and lavender
(Lavandula officinallis L.) (Labiatae) are extensively grown in
Bulgaria as a source of fragrance.
juncea were usually other weed species such as Cichorium intybus L. (Asteraceae), Chamomilla recutita (L.)
Rauschert (Asteraceae), Avena fatua L. (Poaceae), Cuscuta spp. (Convolvulaceae), Cirsium arvense (L.) Scop.
(Asteraceae), Centaurea cyanus L. (Asteraceae) and
Verbascum thapsus L. (Scrophulariaceae).
The vegetative growth of rush skeletonweed in Bulgaria begins at the end of March and the first weeks of
April depending on the local climatic conditions and
altitude. Flower buds are formed at the beginning of
June, and flowering was observed at the end of the same
month; by November, the flowering stem had died.
Phytophagous insects feeding
on C. juncea
During the 5 years of the survey, a total of over
51 insect species from Coleoptera, Lepidoptera, Heteroptera, Homoptera and Diptera were found associated with C. juncea. The species found, their density
and seasonal occurrence has been presented elsewhere
(Karova and Lecheva, 2005; Karova, 2006). Most of the
identified species were moth larvae, leaf beetles, plant
bugs and other sucking insects. To a lesser extent, the
flower buds, flowers and seed heads of C. juncea were
damaged by plant bugs, Lepidoptera larvae and beetles.
Only two species were recovered from the roots: larvae
of the moth, Bradyrrhoa gilveolella Tri. (Lepidoptera:
Pyralidae), which feeds in a sand-encrusted silken case
on the outside of the root, and the beetle Mordelistena
micans Germ. (Coleoptera: Mordellidae), the larva of
which mines down the central core of the root. Most of
the species are generalists and thus cannot be considered as potential biocontrol agents. Only four species
appeared to have a restricted host range: B. gilveolella,
M. micans, the midge Cystiphora schmidti Rub. (Diptera: Cocidomidae) and the flower-feeding moth, S.
cognata Fr. (Lepidoptera: Noctuidae). B. gilveolella
and C. schmidti have already been studied and introduced as biological control agents in Australia and the
USA (Julien and Griffin, 1998; Piper et al., 2004). Min-
302
The insect fauna of Chondrilla juncea L. (Asteraceae) in Bulgaria and preliminary studies of Schinia cognata
ing by M. micans appears to have no effect on the plant,
so we concentrated most of our effort on studying S.
cognata.
S. cognata
Among all recovered insect species in Bulgaria,
the seed-head-feeding moth, S. cognate, was the most
abundant and damaging and seemed the most promising candidate for biological control of C. juncea. It
represented 20% to 34% of all the insects collected
in sweeping the plants in 2003 and 2004 (Table 2).
Table 2.
Field collected larvae of S. cognata in flower buds of
C. juncea were most abundant not only in the vicinity
of Plovdiv but were also recovered in most regions of
the country. Very little previous information exists on
S. cognata. According to Nowacki and Fibiger (1996)
and Rakosy (1996), it occurs in the European part of
the former USSR, Czech Republic, Austria, Hungary,
Former Yugoslavia, Greece, Albania and Bulgaria. The
first record of S. cognata in Bulgaria was from Bahmetev (1902) who found only single individuals. Until
now, the moth was considered a rare species. The results of our study show that it has two generations per
Relative abundance of insects collected on Chondrilla juncea by sweep netting
over the summer of 2004 at two locations near Plovdiv, Bulgaria. Over 50 species
were collected (Karova and Lecheva, 2005; Karova, 2006), but only these were
observed or suspected of actually feeding on the plant.
Genus and species
Coleoptera
Gastrophysa polygoni L.
Cassida nebulosa L.
Clytra novempunctata L.
Coptocephala rubicunda Laich
Coptocephala unifasciata Scop.
Cryptocephalus connexus Ol.
Cryptocephalus sericeus L.
Epicometis hirta Poda.
M. micans Germ.b
Mylalris polymorpha Pall.
Zonabris polymorpha Pall.
Lepidoptera
Bradyrrhoa gilveolella Tr.b
Emmelia trabealis Scop.
Idaea muricata Hufn.
Melitaea didyma Esp.
Plusia gamma L.
S. cognata Fr.
Simira nervosa (Den. & Schiff.)
Heteroptera
Adelphocorvis lineolatus Goez.
Brachycoleus decolar Renter
Dolycoris baccarum L.
Eurybema oleraceae L.
Eurydema ornata L.
Lygus equestris L.
Lygus pratensis L.
Lygus rugulipennis Popp.
Syramastes rhombeus L.
Homoptera
Aphis spp.
Philaenus spumarius L.
Diptera
C. schmidti Rub.b
Dasyneura sp.
Family
Feeding site
Percent
abundance
Chrysomelidae
Chrysomelidae
Chrysomelidae
Chrysomelidae
Chrysomelidae
Chrysomelidae
Chrysomelidae
Scorabaeidae
Mordellidae
Meloidae
Meloidae
Foliage
Foliage
Flower
Foliage
Foliage
Foliage/Flower
Flower
Flower
Roots
Pollen
Pollen
0.95
1.09
Occ.a
Occ.
Occ.
2.45
3.59
1.63
0.85
10.17
10.76
Pyralidae
Noctuidae
Geometridae
Nymphalidae
Noctuidae
Noctuidae
Noctuidae
Roots
Foliage
Foliage
Flower
Flower
Flower bud
Foliage
Occ.
10
1.24
10.26
0.67
27.15
0.81
Miridae
Miridae
Pentatomidae
Pentatomidae
Pentatomidae
Lygaeidae
Lygaeidae
Lygaeidae
Lygaeidae
Sap feeding
Sap feeding
Sap feeding
Sap feeding
Sap feeding
Sap feeding
Sap feeding
Sap feeding
Sap feeding
4.29
1.25
3.41
Occ.
Occ.
Occ.
Occ.
Occ.
1.46
Aphididae
Cereopidae
Sap feeding
Sap feeding
2.45
5.45
Cecidomiidae
Cecidomiidae
Under cuticle
Flower
Occ.
Occ.
Occ. = Occasionally found, but were generally rare or found for only a very short period during the
summer
b
These insects were rarely collected as adults during sweep netting, but their larvae were found on or in
the plants.
a
303
XII International Symposium on Biological Control of Weeds
Figure 1.
eason population density (determined by 5 100 sweeps with a collection net) of
S
larvae and adults of Schinia cognata near Plovdiv, Bulgaria in 2003 and 2004.
year that overlap, with maximum population densities
in July and August (Fig. 1). The moth overwinters as a
pupa, and emergence begins in early June. The females
lay their eggs singly on the upper parts of the stems,
flower buds and flowers of the host plant beginning in
mid-June. The eggs begin to hatch in the first weeks of
July. In the two populations examined near Plovdiv, the
highest density of larvae was observed in mid-July and
mid-August. The larvae feed on the reproductive parts
of C. juncea, i.e. flower buds, flowers and seed heads.
Observations under laboratory conditions showed that,
for full development, one larva consumes 61 to 62
flower buds, flower heads or seed heads.
During our investigations, the moth was not encountered on other related weed species or cultivated plants
growing in the vicinity of attacked C. juncea plants.
Laboratory feeding tests confirmed its host specificity.
Four closely related species of Asteraceae were used
for both choice and no-choice trials: C. intybus L.,
Lactuca serriola L., Lactuca sativa L. (domestic let-
tuce), and Sonchus oleraceus L. In all choice trials, the
larvae fed only on C. juncea, and there was no feeding on the four tested plants. In the no-choice trials,
where the larvae were forced to feed on plants different
from C. juncea, some larvae attempted to feed but none
survived to pupate. On C. juncea, larvae development
was normal; most of them pupated and adults emerged
successfully.
Conclusion
Among the insect species found feeding on C. juncea
in Bulgaria, the dominant species was the moth S. cognata. The larvae feed on the reproductive parts of the
plant and were observed to cause extensive damage
in the field. S. cognata is widely distributed, occurs at
high densities in all C. juncea population studies and
was not recorded or observed feeding on other, local or
cultivated plants. It could not be reared on four closely
related plant species and is therefore considered a po-
304
The insect fauna of Chondrilla juncea L. (Asteraceae) in Bulgaria and preliminary studies of Schinia cognata
tential biocontrol agent. Arrangements have been made
to ship colonies of S. cognata to a plant containment
facility in Bozeman, Montana in the USA where it can
undergo more extensive host testing on North American crops and native species.
Acknowledgements
We would like to thank the USDA-ARS, European
Biological Control Laboratory, Montpellier, France,
and USDA Forest Service, Rocky Mountain Experiment Station, Fort Collins, Colorado, U.S.A., for the
successful collaboration and financial support.
References
Bahmetev, P. (1902) Babochki Bolgarii. Trudui Rossijskoi
Entomologicheskoi Obshtestva Sankt-Petersburg, 35,
356466.
Cullen, J.M. and Groves, R.H. (1977) The population biology
of Chondrilla juncea L. in south-eastern Australia. Journal of Ecology 54, 345365.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control of
Weeds: A world catalogue of agents and their target weeds,
4th edn. CABI Publishing, London, England, 223 pp.
Karova, A. (2006) Study on Chondrilla juncea L. (Asteraceae), associated phytophagous insects and their potential
as agents for biological control. PhD Dissertation. Agricultural University of Plovdiv, Plovdiv, Bulgaria, 137 pp.
(in Bulgarian).
Karova A. and Lecheva, I. (2005) Study on the habitats of
Chondrilla juncea L. (Asteraceae) and the diversity of its
natural enemies in Bulgaria. Plant Sciences 42, 456460
(in Bulgarian).
Markin G. and Quimby, P. Jr. (1997) Report of work on
biological control of rush skeletonweed (Chondrilla juncea). Unpublished report on file at USDA Forest Service,
Forestry Sciences Laboratory, MSU Campus, Bozeman,
Montana, p. 37.
Nowacki, J. and Fibiger, M. (1996) Noctuidae. In: Karsholt,
O. and Razowski, J (eds) The Lepidoptera of Europe.
Apollo, Stenstrup, Denmark, pp. 251293.
Piper, G.L. and Andres, L.A. (1995) Rush skeletonweed. In:
Nechols, J.R., Andres, L.A., Beardsley, J.W., Goeden,
R.D. and Jackson, C.G. (eds) Biological Control in the
Western United States. University of California. Division
of Agriculture and Natural Resources. Publication 3361,
pp. 252255.
Piper, G.L., Coombs, E.M., Markin, G.P. and Joley, D.B.
(2004) Rush skeletonweed. In: Coombs, E.M., Clark,
J.K., Piper, G.L. and Confrancesco, Jr., A.F. (eds) Biological Control of Invasive Plants in the United States.
Oregon State University Press, Corvallis, OR, pp. 293
303.
Rakosy, L. (1996) Die Noctuidae Rumaniens (Lepidoptera,
Noctuidae). Druckerei Gutenberg, LinzDornach, 648 pp.
305
Biological control of aquatic weeds by
Plectosporium alismatis, a potential
mycoherbicide in Australian rice crops:
comparison of liquid culture media for
their ability to produce high yields of
desiccation-tolerant propagules
C. Moulay,1 S. Cliquet,1 K. Zeeshan,1 G.J. Ash2 and E.J. Cother3
Summary
A methodology to develop a stable, effective Plectosporium alismatis (Oudem.) Pitt, Gams and Braun
[syn. Rhynchosporium alismatis (Oudem) J.J. Davis] mycoherbicide is currently being investigated.
We compared a nitratemalt extract medium liquid culture medium to other liquid media for their
ability to support high conidial and chlamydospore yields and subsequent tolerance of conidia and
chlamydospores to air-drying. When grown in a casamino acidsglucose-based liquid medium, P.
alismatis developed hyphae and produced high yields of conidia (1 107 conidia per millilitre) and
dry weights (220 mg dry weight per erlen), while no chlamydospore was formed. In a nitrateglucosebased medium, growth was poor, P. alismatis producing aggregated hyphae that contained 6.5 104
chlamydospores per millilitre. The addition of nitrate in the casamino acidsglucose-based medium
restored partially chlamydospore formation (1 104 chlamydospores per millilitre). Chlamydospores
and conidia were air-dried and stored at 25C. No conidia germinated after 40 days storage, while
50% to 20% chlamydospores, respectively, produced in a nitratemalt extract medium or in nitrate
glucose medium, remained viable after 120 days storage.
Keywords: fermentation, air-drying, storage, chlamydospores, conidia.
Introduction
The endemic fungus, Plectosporium alismatis (Oudem.) Pitt, Gams and Braun [syn. Rhynchosporium
alismatis (Oudem) J.J. Davis] (Pitt et al., 2004) is being
developed as a mycoherbicide (Crump et al., 1999) for
the control of starfruit and other closely related weed
species (Jahromi et al., 2001). The fungus sporulates
1
Universit de Bretagne Occidentale, Laboratoire de Microbiologie
Applique de Quimper (LUMAQ), Biopesticide Research, 2, rue de
lUniversit, Quimper 29000 France.
2
Charles Sturt University, E.H.H Graham Centre for Innovative Agriculture, School of Agricultural and Veterinary Sciences, Boorooma Street,
PO Box 588, Wagga Wagga, NSW 2678, Australia.
3
NSW Department of Primary Industries, Agricultural Institute, Orange,
New South Wales 2800, Australia.
Corresponding author: S. Cliquet <[email protected]>.
CAB International 2008
abundantly on solid media (Jahromi et al., 1998) and
is able to infect host species (Lanoiselet et al., 2001),
leading to reduced biomass of the weed or to reduced
seed set (Fox et al., 1999).
A culture production method for the development of
P. alismatis mycoherbicide is currently being investigated. P. alismatis produces high numbers of conidia in
most liquid media; chlamydospore formation also occurs in a liquid standard medium based on the Czapex
Dox composition, in which nitrogen and carbon are
provided by sodium nitrate (3 g l-1) and malt extract
(2.2 g l-1; Cliquet et al., 2004).
We modified the carbon and the nitrogen sources
and concentrations of the nitratemalt extract medium to investigate the impact of nitrogen and carbon
sources on both conidia and chlamydospore inductions.
Shelf-life of air-dried conidia and of air-dried chlamydospores produced in modified carbon and nitrogen
306
Biological control of aquatic weeds by Plectosporium alismatis
sources was investigated and compared to shelf-life of
air-dried propagules produced in the nitratemalt extract medium.
fragmented in a Potter homogenizer (Fisher). Spore
counts were performed using a haemocytometer.
Chlamydospore and conidial germination
Materials and methods
Isolate
P. alismatis RH 145 (DAR 73154) was isolated from
Damasonium minus (R.Br) Buch., which was obtained
from the culture collection of the New South Wales
Department of Primary Industries. Stock cultures were
maintained in a soil/sand mixture.
Inoculum production
Sub-cultures on potato dextrose agar (PDA, Difco,
Detroit, MI, USA) were sampled from the soil and
sand mixture and renewed every year. From these sub-
cultures, conidia were inoculated on PDA plates and incubated at 25C. Four-day-old Petri dishes of the fungus
were washed with distilled water to produce conidial
suspensions for liquid culture as described hereafter.
Media composition
The basal mineral composition of nitratemalt extract medium and defined medium was derived from a
CzapexDox composition containing: 1.0 g K2HPO4;
1.0 g MgSO4.7H20; 0.5 g KCl; 0.018g Fe2SO4.7H2O in
1 l deionized water. The nitratemalt extract medium
contained 3 g NaNO3 (Sigma Chemicals, St. Louis,
MO, USA ) and 2.2 g malt extract (Amyl Media).
In the defined medium, malt extract was replaced by
2.4 g l-1 glucose at the carbon concentration (920 mg
C l-1) provided by malt extract. The 3.12 g l-1 NaNO3
provided the same nitrogen content (0.5 g N l-1) than in
the nitratemalt extract medium. Bacto yeast nitrogen
base without amino acids and (NH4)2SO4 (Difco) was
provided as a nitrogen-free vitamin source (0.17 g l-1).
As the organic nitrogen source, 0.5 g N l-1 technical
casamino acids (Difco) was used. In shelf-life experiments, propagules were produced in media containing
3.68 g C l-1 and 1 g N -1.
Growth and harvest
All cultures of 100 ml were placed in 250-ml flasks
(Bellco Glass, Inc, Vineland, NJ, USA), inoculated with
4 105/ml conidial suspension and incubated at 25C
on a rotary shaker incubator (Certomat BS-1 Braun
Biotech International, Germany) at 100 rpm for 7 days
for growth experiments or for 4 days for drying and
storage experiments. Cultures were vacuum-filtered
on 110 mm cellulose filter papers (Whatman plc,
Brentford, UK). Filtered cultures were rinsed with 50
ml deionized water and allowed to dry on the bench top
until constant weight. Dry mats were weighed and resuspended in 50 ml distilled water. The suspension was
Drops of the propagule suspension were placed on
four 2-cm2 pieces of cellophane on the surface of water
agar plates. Cellophane pieces were removed after 12 h
at 25C and germination evaluated microscopically
using lactophenol cotton blue as previously described
(Cliquet et al., 2004).
Statistical analysis
All growth experiments were performed using duplicate or triplicate flasks, and all experiments were
repeated at least once. Statistical analysis of variance
was performed. For data not suitable for analysis of
variance, standard errors values were estimated as a
measure of variance.
Results and discussion
P. alismatis produced significantly higher yields of conidia when grown in casamino acids (1 107conidia
per millilitre) compared to conidial yields produced in
sodium nitrate (2 105conidia per millilitre; Fig. 1).
When nitrate was the sole nitrogen source, 6.5 104
chlamydospores per millilitre were observed; however,
the addition of sodium nitrate to the medium containing casamino acids resulted in the production of less
chlamydospores (1 104 chlamydospores per millilitre; Fig. 1).
When grown in a medium containing casamino acids, P. alismatis produced numerous hyphae resulting
in high dry weights with a maximum of 220 mg dry
weight per erlen, compared to low dry weights (80 mg
per erlen) obtained when sodium nitrate was the sole
nitrogen source (Fig. 2). Homogenization of cultures
showed that chlamydospores were mainly formed inside these aggregates.
In our culture conditions, replacing sodium nitrate
by casamino acids as the sole nitrogen source and at the
same nitrogen content improved growth as expressed
by the high conidial yields and dry weights. Organic
nitrogen provided by casamino acids was probably utilized preferentially to inorganic nitrogen by P. alismatis, as previously reported for a majority of filamentous
fungi (Garraway and Evans, 1984). The absence of
chlamydospore formation in these conditions is likely
the consequence of the addition of organic nitrogen.
In filamentous fungi, chlamydospore production
may vary considerably depending on the nutritional
environment, i.e. a nutrient excess or starvation conditions (Gardner et al., 2000). In our culture conditions,
starvation due to a lack of organic nitrogen is likely
responsible for chlamydospore formation.
307
XII International Symposium on Biological Control of Weeds
160
10
Chlamydospores
Conidia
140
100
80
60
40
Chlamydospores x 104 ml-1
Conidia x 105 ml-1
120
20
0
NO3+malt extract
Figure 1.
160
80
60
40
NO3+casa+glc
10
240
Chlamydospores
Dry weights
Conidia
Chlamydospores x 104 ml-1
Conidia x 105 ml-1
100
NO3+glc
Impact of the carbon and nitrogen sources on Plectosporium alismatis conidial and chlamydospore production. NaNO3 + malt extract: sodium nitrate: 3 g l-1; malt extract: 2.2 g l-1; NaNO3 + GLC: sodium
nitrate + glucose (1); CA+GLC: casamino acids + glucose (2); NaNO3+CA+GLC: sodium nitrate + casamino acids + glucose (3); In (1), (2) and (3), glucose provides 0.92 g C l-1; nitrogen sources provide each
0.5 g N l-1.
140
120
casa+glc
220
200
180
160
140
120
100
20
0
80
0
60
0
Figure 2.
Dry weights (mg erlen-1)
-1
Casamino acids (gL )
Influence of casamino acids concentration on production of chlamydospores and conidia by Plectosporium alismatis (Na nitrate: 0.5 g N l-1; casamino acids 4.7g l-1 = 0, 5g N l-1).
308
Biological control of aquatic weeds by Plectosporium alismatis
Conidia and chlamydospores produced in nitrate
malt extract or nitrateglucose were air-dried, and germination was estimated during storage at 25C (Fig. 3).
The type of propagules, i.e. conidia or chlamydospores, had a significant impact on survival rate (Fig. 3).
No conidia germinated after 40 days storage, while 50
to 20% chlamydospores, respectively, produced in a
nitratemalt extract medium or in nitrateglucose medium remained viable after 120 days storage (Fig. 3).
Microscopic observation showed that some chlamydospores produced in nitrateglucose medium sporulated through a microcycle conidiation.
malt extract nitrate
120
chlamydospores
conidia
100
% germination
80
60
40
20
100
0
0
20
40
80
glucose nitrate
80
%germination
60
100
120
chlamydospores
conidia
60
40
20
0
0
20
40
60
80
100
120
day of storage (25C)
Figure 3.
Shelf-life at 25C of air-dried Plectosporium alismatis conidia and chlamydospores. Plectosporium
alismatis was grown in a nitratemalt extract medium (8.8 g l-1 malt extract, 5.74 g l-1 sodium nitrate) or in a nitrateglucose medium (6 g l-1 sodium nitrate, 10 g l-1 glucose) and harvested after 7
days growth.
309
XII International Symposium on Biological Control of Weeds
Conclusion
This work shows that numerous factors are to be investigated to develop a stable mycoherbicide. Survival
during storage depends upon the type of propagules produced and upon the culture conditions during growth.
Moreover, the microcycle conidiation observed during
germination experiments may allow the fungus to extend rapidly and colonize aquatic weeds effectively. As
a conclusion, chlamydospores may be promising stable
propagules compared to conidia, although the nutritional conditions impact these qualities.
More work needs to be done to consider as many
parameters (physical, chemical and morphological) as
possible in an experimental design for the selection of
factors that impact chlamydospore formation and tolerance to drying.
References
Cliquet, S., Ash G. and Cother, E. (2004) Conidial and chlamydospore production of Rhynchosporium alismatis in
submerged culture. Biocontrol Science and Technology
14, 801810.
Crump, N. S., Cother E.J. and Ash, G.H. (1999) Clarifying
the nomenclature in microbial weed control. Biocontrol
Science and Technology 9, 8997.
Fox, K.M., Cother, E.J. and Ash, G.J. (1999) Influence of infection of Rhyncosporium alismatis on seed production
by rice paddy weed Damasonium minus (starfruit). Australasian Plant Pathology 28, 197199.
Gardner, K., Wiebe, M.G., Gillespie A.T. and Trinci, A.P.J.
(2000) Production of chlamydospores of the nematodetrapping Duddingtonia flagrans in shake flask culture.
Mycological Research 104, 205209.
Garraway, M.O. and Evans, R.C. (1984) Fungal nutrition and
physiology. Wiley, New York, 8 pp.
Jahromi, F.G., Ash, G.J. and Cother, E.J. (1998) Influence of
cultural and environmental conditions on conidial production, conidial germination and infectivity of Rhynchosporium alismatis, a candidate mycoherbicide. Australasian
Plant Pathology 27, 180185
Jahromi, F.G., Cother, E.J. and Ash, G.J. (2001) The use of a
fungal pathogen to reduce weed competition in Australian
rice. 13th Biennial Conference of the Australasian Plant
Pathology Society, Cairns, Australia.
Lanoiselet, V., Cother, E.J., Ash, G.J. and van de Ven, R.
(2001) Production, germination and infectivity of chlamydospores of Rhynchosporium alismatis. Mycological
Research 105, 441446.
Pitt, W.M., Cother, N.J., Cother, E.J. and Ash, G.J. (2004)
Infection process of Plectosporium alismatis on host and
non-host species in the Alismataceae. Mycological Research 108, 837845.
310
Feeding and oviposition tests refute
hostherbivore relationship between
Fragaria spp. and Abia sericea, a candidate
for biological control of Dipsacus spp.
B.G. Rector,1 V. Harizanova2 and A. Stoeva2
Summary
Two species of teasels, Dipsacus fullonum L. and Dipsacus laciniatus L. (Dipsacales: Dipsacaceae),
have become invasive in the USA and are targets of a classical biological control programme. The
sawfly, Abia sericea (L.) (Hymenoptera: Cimbicidae), was identified as a teasel biological control
candidate, but reports in the literature raised concerns about the possibility of non-target effects on
members of the genus Fragaria L. Oviposition tests and multiple- and no-choice feeding tests were
conducted on A. sericea to test the suitability of three Fragaria spp. (Fragaria ananassa Duchesne,
Fragaria vesca L. and Fragaria viridis Duchesne) as host plants. In oviposition tests, males were
paired with females in cages containing a D. laciniatus plant and plants of the three Fragaria spp. In
16 replicates, all eggs were laid in D. laciniatus leaves, except in one replication in which two eggs
were laid in a leaf of F. viridis (vs >100 eggs in D. laciniatus leaves by the same female and >1000 eggs
by all females). From these two eggs, one larva hatched, which fed only on D. laciniatus in a choice
test and died before pupating. In no-choice feeding tests on the three Fragaria spp., larvae of first,
second, third and fourth instars were infested in separate tests, and no feeding, tasting or damage was
observed on any Fragaria spp. in any replicate; all larvae died. In multiple-choice feeding tests, first,
second, third and fourth instar larvae were reared to pupation. Larvae of all instars fed exclusively on
D. laciniatus, while no feeding attempts, tastes or feeding damage was observed on any Fragaria spp.
plant. Entomology and agricultural literature were thoroughly reviewed, and the connection between
Fragaria and A. sericea was traced to two brief anecdotal mentions that were widely repeated but
never supported by collection or bioassay data. The tests and other evidence presented in this paper
refute the idea that Fragaria spp. are suitable hosts for A. sericea.
Keywords: teasel, invasive plants, strawberry.
Introduction
Two invasive teasels of European origin, Dipsacus
fullonum L. and Dipsacus laciniatus L., are emerging
as problem weeds in various parts of North America,
particularly in non-agricultural habitats (Sforza, 2004).
Either or both species occur in 43 states (Singhurst and
Holmes, 2001; USDA, 2005; Rector et al., 2006) and in
several Canadian provinces (Werner, 1975). Five states
(Colorado, Iowa, Missouri, New Mexico and Oregon)
have declared teasels noxious (USDA-NRCS, 2005).
They are also listed as invasive by 11 other states and
1
USDA-ARS, European Biological Control Laboratory, Campus International de Baillarguet, Montferrier-sur-Lez, France.
2
Agricultural University, Faculty of Plant Protection, Department of Entomology, 12 Mendeleev St., 4000 Plovdiv, Bulgaria.
Corresponding author: B.G. Rector <[email protected]>.
CAB International 2008
as affecting natural areas in 14 states and four national
parks (USDI-NPS, 2005). This combined status and
other factors led to the initiation of a US-governmentsponsored biological control programme targeting Dipsacus spp. in 2003.
The genus Dipsacus L. is in the family Dipsacaceae,
an exclusively Old World family with no species native
to the Western Hemisphere (Verlaque, 1985) and no
members of significant economic importance (Bailey,
1951). Thus, in selecting biological control candidates
(BCCs) to combat invasive teasels in North America,
those restricted to feeding on hosts within the family
Dipsacaceae should be specific enough to avoid nontarget concerns (Wapshere, 1974; Rector et al., 2006).
However, for any teasel BCC, the list of plants tested
will include many species outside the family Dipsacaceae, including economic and rare or threatened
species, as well as those occupying similar ecological
311
XII International Symposium on Biological Control of Weeds
niches, those sharing similar life histories, plant architecture and secondary chemistry, and those with published associations with the BCC (Rector et al., 2006).
Abia sericea (L.) (Hymenoptera: Cimbicidae) is a
sawfly that is native to Europe and is a candidate for
biological control of invasive teasels in the USA. The
most recent, comprehensive host-range records for A.
sericea (Liston, 1995, 1997; Taeger et al., 1998) list
two host species in the family Dipsacaceae [Knautia arvensis (L.) Coult. and Succisa pratensis Moench], plus
the genus Fragaria L., which is in the family Rosaceae
and includes cultivated and wild strawberries.
In assessing A. sericea as a BCC of teasels, it was necessary to start by testing the suitability of Fragaria as a
host, to either confirm or refute the existing herbivore
host records. Any amount of A. sericea larval feeding on
Fragaria plants, particularly on cultivated strawberry
(Fragaria ananassa Duchesne), would immediately
disqualify it as a BCC of teasel. It was also necessary
from the start to confirm the use of D. laciniatus as
a host plant suitable for complete development by A.
sericea, as this relationship had only previously been
observed in the field without further testing (Rector et
al., 2006; Rector et al., unpublished data).
The purpose of this paper is to report the successful
rearing of A. sericea on D. laciniatus and to refute the
published herbivorehost relationship between A. sericea and Fragaria spp.
Materials and methods
Literature survey
An intensive literature search was conducted to
determine the origin(s) of reports of a herbivorehost
relationship between A. sericea and Fragaria spp. An
attempt was made to identify the original source literature behind this alleged association and to gauge
its plausibility. Literature surveyed included various
entomological texts, articles and monographs. Most of
these were on the subject of basic sawfly (Hymenoptera: Symphyta) biology and taxonomy (e.g. Andr,
1879; Stein, 1883; Cameron, 1890; Dalla Torre, 1894;
Konow, 1901; Enslin, 1917; Berland, 1947; Lorenz and
Kraus, 1957; Ermolenko, 1972; Vasilev, 1978; Wright,
1990; Magis, 2001). Other sources included sawfly
host-range compendia (Liston, 1995, 1997; Taeger et
al., 1998), work specific to Abia spp. (Kangas, 1946)
and a general entomology text (Krishtal, 1959). In addition, pest management literature for cultivated strawberries was surveyed to determine whether A. sericea is
considered a pest to commercial strawberry growers in
its native range in Europe.
Insect material
On 4 September 2005, 29 late-instar larvae of A.
sericea were collected from the wild on D. laciniatus
in northern Bulgaria at two locations in the vicinity
of Pleven (4324.26N, 2428.76E and 4314.16N,
2418.24E) and one location near Lovech (4304.12N,
1444.09E). All larvae from these three sites were
grouped together and reared to pupation under ambient
climatic conditions near Plovdiv, Bulgaria (4208.64N,
2443.81E). On 29 June 2006, adult females were collected from D. laciniatus plants in the field near Sofia
(423757 N, 23303427 E). Oviposition tests and feeding tests were conducted using all of these insects and
their progeny.
Test plants
The D. laciniatus plants used in host-specificity
testing were either grown from seed gathered from
wild plants in Bulgaria or were transplanted from fallow fields near Plovdiv (4208.64N, 2443.81E).
The test plants of the wild Fragaria spp. (Fragaria vesca
L. and Fragaria viridis Duchesne) were dug up from
the wild on a mountainside near Plovdiv (4153.80N,
2520.07E), transferred to pots and identified to species with a key (Markova, 1973). Cultivated strawberry
(F. ananassa) plants were vegetatively propagated
from 6-year-old plants.
Oviposition tests
Test 1: On 11 Oct 2005, a male and three female A.
sericea adults were put in a large cage (40 20 40 cm,
made from clear, plastic panels with fine nylon mesh
tops) with two D. laciniatus and two F. ananassa
plants. At 8:30 a.m. on 12 October, the teasel plants
were removed from the cages, leaving only the strawberry plants. At 4:30 p.m. on the same day, the teasel
plants were returned to the cages. The insects were then
left in the cage with both plant species until they died.
A 5% sugar solution was provided for the insects with
a cotton wick from which to feed.
Test 2: Pairs of newly emerged adults were released
into small plastic cages (20 20 30cm) with one potted plant each of D. laciniatus, F. ananassa, F. vesca
and F. viridis. Cages were kept in an insectary at approximately 22/15C, day/night, and 16 h daylight.
The adults were kept in the cages until they died, after which the plants were removed from the cages and
examined with a magnifying glass for eggs laid, and
this number was recorded for each plant species. A total
of 16 replications were conducted during the period 1
April to 1 Aug 2006.
No-choice larval feeding tests
No-choice feeding tests were conducted in small
plastic cages. In each cage, one plant of each of the
three Fragaria spp.F. vesca, F. viridis and F. ananassawere exposed to six larvae of first instar of A.
sericea. Four replications were carried out, while one
cage with only D. laciniatus plants and six larvae of the
312
Feeding and oviposition tests refute hostherbivore relationship between Fragaria spp. and Abia sericea
same instar from the same cohort was set up as a control. Larvae were checked every 48 h whereupon the
number of dead larvae was recorded. The same procedure was followed with four replicates each of second,
third and fourth instar larvae that had been reared on D.
laciniatus to that stage.
Multiple-choice larval feeding tests
The tests were carried out by placing six A. sericea
larvae of the same instar into a small cage with one
potted plant each of D. laciniatus, F. vesca, F. viridis
and F. ananassa whose leaves were all touching. To
begin the test, two larvae were placed on each of the
three Fragaria spp. plants, after which they were free
to move to other plants. The experiment was repeated
three times: from 24 April to 16 May 2006 with seven
replications each of larvae from second, third and fourth
instars; from 17 May to 15 June with seven replications
each of larvae from first to fourth instars and from 7
July to 5 August with eight replications each of larvae
from first to fourth instars. The larvae were monitored
daily, and the number of dead larvae was recorded. The
experiment was terminated after the pupation of the last
larva. D. laciniatus plants were replaced as needed.
hostplants list for A. sericea. The Krishtal (1959) information in particular has been characterized as unreliable (A. Taeger, personal communication).
In addition, if A. sericea were in fact a herbivore of
Fragaria spp., it is likely that it would be of concern
to commercial strawberry (F. ananassa) production
within its native range. However, A. sericea is not listed
as a pest of strawberry in current pest management literature in Bulgaria (Harizanov and Harizanova, 1998),
England (Marks, 2008) or France (Lamarque and Bossennec, 2001), three strawberry-producing nations
within the native range of A. sericea. Taken together,
these various lines of evidence provide no support
whatsoever to the records by Konow (1901) and Krishtal (1959) suggesting that Fragaria spp. are host plants
for A. sericea.
Oviposition tests
Test 1: A total of 64 eggs were laid on D. laciniatus
plants by the three A. sericea females before the teasel
plants were removed from the cage (Figs. 1 and 2). No
eggs were laid on the strawberry plants, neither in the
presence of the teasel plants nor in their absence. After
the teasel plants were returned to the cage, an additional
26 eggs were laid on them.
Results
Literature survey
There appear to have been only two direct reports
of Fragaria spp. recorded as a host plant for A. sericea,
with all other such reports in the literature referring either to one of these two original reports or to others
that have, in turn, cited the original two. The first report (Konow, 1901) simply lists A. sericea as occurring
on S. pratensis and F. vesca while including neither
a reference nor any specific collection data nor any information regarding feeding nor rearing experiments.
The second independent report alleging an association
between A. sericea and Fragaria comes from a Ukrainian general entomology text by Krishtal (1959) that
covers all insects in all orders occurring in that country.
This book declares that A. sericea is polyphagous,
although the only food plants listed are two Fragaria
spp. that are native to Ukraine: F. vesca and F. viridis,
and there is no collection or feeding/rearing information provided. This alleged hostherbivore connection
has been repeated many times throughout the entomological literature, particularly in literature pertaining to
Symphyta in general (e.g. Enslin, 1917; Ermolenko,
1972; Liston, 1995, 1997; Taeger et al., 1998) or the
genus Abia in particular (e.g. Kangas, 1946). However,
recent reports have cast doubt on this connection. Indeed, Taeger et al. (1998) stated explicitly Fragaria
is surely not a food plant [for Abia sericea] under field
conditions. However, they presented no reference or
data to support this statement and left Fragaria on their
Figure 1.
313
Abia sericea female ovipositing into the leaf
margin of a D. laciniatus plant.
XII International Symposium on Biological Control of Weeds
was observed on D. laciniatus during daylight hours.
There appeared to be a stimulus to this oviposition-
associated behaviour that was linked to the presence of
D. laciniatus plants and a lack of such stimulus from
the strawberry plants.
Oviposition occurred along the margins of the D.
laciniatus leaves, with the eggs deposited just beneath
the epidermal layer of either the upper or lower leaf
surface (Figs. 1 and 2). None of the adults were observed feeding on the sugar solution, consistent with
the observation by Vasilev (1978) that most Cimbicidae adults do not require such food.
Test 2: The presence of D. laciniatus plants did not
stimulate oviposition on adjacent strawberry plants or
on other surfaces such as the walls of the cage, with
only one exception. A total of 16 females laid 936 eggs,
934 of which were laid in D. laciniatus leaves (99.8%),
while two were laid by one female in a F. viridis leaf
(0.2%). This same female also laid >100 eggs in D. laciniatus leaves during the same trial. Only one of the
eggs laid in the F. viridis leaf hatched, and the larva died
before pupation after feeding exclusively on D. laciniatus leaves despite hatching from a F. viridis leaf and
always having all three Fragaria spp. available as food
options. There were no eggs laid on any of the F. vesca
or F. ananassa leaves in any of the replications.
Figure 2.
Abia sericea eggs (along leaf margins) and neonate larvae on a leaf of Dipsacus laciniatus.
A difference in female A. sericea behaviour was
observed that coincided with the presence or absence
of the teasel plants in the cage. The females moved
restlessly over the surface of the D. laciniatus leaves,
raising and lowering their abdomens and periodically
piercing the leaf epidermis with their ovipositors.
This behaviour was not observed when they were on
strawberry leaves. When the teasel plants were absent,
leaving only the three Fragaria spp. plants, the females
were no longer active at all. This indifferent behav
iour did not appear to correlate with their only spend
ing daylight hours alone with the strawberry plants, as
oviposition by A. sericea and its associated behaviour
Table 1.
No-choice larval feeding tests
No feeding, feeding attempts or evidence of feeding was observed on any of the F. vesca, F. viridis or
F. ananassa leaves in any of the cages (Table 1). Most
of the first-instar larvae died after 1 or 2 days, apparently of starvation. The second-instar larvae died 1 to
4 days after their release into the cage. As the instar
of the larvae at the beginning of the test increased, the
survival period to death also increased. The third-instar
larvae died within 5 days (most after 3 to 4 days), and
the fourth-instar larvae were all dead by the eighth day
(most after 4 or 5 days) without access to D. laciniatus plants as food. In contrast, only one second-instar
larva died feeding on D. laciniatus in a control cage.
All other control larvae of all instars (23 in total) fed
normally and survived to pupation.
Summary of oviposition and choice and no-choice feeding experiments of Abia A.
sericea on Dipsacus laciniatus, Fragaria ananassa, F. vesca and F. viridis.
Test type
Oviposition, Test 1
Oviposition, Test 2
Larval feeding,
no-choice
Larval feeding,
free-choice
Number of
replicates
No. eggs laid/larvae feeding on each host plant
D. laciniatus
1
16
16
90
934
96
F. ananassa
0
0
0
81
486
314
F. vesca
F. viridis
0
0
0
0
2
0
Feeding and oviposition tests refute hostherbivore relationship between Fragaria spp. and Abia sericea
Multiple-choice larval feeding
A total of 486 A. sericea larvae fed exclusively on
the teasel plants until they either pupated or died. They
did not feed at all on the F. viridis, F. vesca, or F.
ananassa plants, and no tasting attempts were observed.
Most of the larvae feeding on D. laciniatus developed
normally and pupated (88.6%). Of the larvae that died
while feeding on D. laciniatus, some showed symptoms of virus infection (Lacey and Brooks, 1997). A
virus was identified from some of these A. sericea cadavers as belonging to the family Iridoviridae. These
same symptoms were not observed in any of the larvae
that died on Fragaria spp. in the no-choice tests.
Discussion
Fragaria vesca and F. viridis, two species of wild
strawberry, have been recorded in the entomological
literature as hosts for the European sawfly A. sericea
(Konow, 1901; Krishtal, 1959). Doubt has been cast
on this insectplant association (Taeger et al., 1998; A.
Taeger, personal communication), but to date, there have
been no data to either confirm or refute these reports.
The phylogenetic evidence suggests that such a connection is unlikely. If the inclusion of Fragaria as a
host of A. sericea were accurate, this would represent
the only known host plants outside the closely related
families Caprifoliaceae and Dipsacaceae for any species in the genus Abia (Taeger et al., 1998). Fragaria
is in the family Rosaceae, which is not at all closely
related to Dipsacaceae (APG II, 2003). Although host
relationships among herbivores do not always follow
plant phylogeny, this evidence is not trivial.
Cultivated strawberry, F. ananassa, is a hybrid of
two New World species, Fragaria chiloensis (L.) P.
Mill. and Fragaria virginiana Duschene (Hokanson et
al., 2006), and thus neither the hybrid plant nor its progenitors would have had any exposure to A. sericea before the introduction of the two progenitor species into
Europe in the 19th century and the subsequent creation
of the hybrid (Hokanson et al., 2006).
In an oviposition test with three gravid females,
eggs were laid on only D. laciniatus. In a second oviposition test with 16 females, a total of 934 eggs (99.8%)
were laid in the leaves of D. laciniatus plants, while
two eggs were laid in a F. viridis leaf (0.2%). Possible
explanations for the laying of these two eggs on a nonhost plant are central nervous system excitation or sensitation, as described by Marohasy (1998), or the lack
of appropriate leaf-edge oviposition sites available on
the D. laciniatus plant in the cage (whose leaves were
more or less lined with eggs by the end of the trial). In
no-choice feeding studies presented in this paper, all
insects died without leaving any trace of feeding or attempted feeding. In control cages in which larvae from
the same cohorts as the insects tested on Fragaria spp.
were feeding exclusively on D. laciniatus, 23 of 24 developed to pupation.
Given that the purported associations between A.
sericea and Fragaria spp. have been shown in this study
to be invalid and given that all confirmed hosts of A.
sericea belong to the family Dipsacaceae, there is no
reason to suspect that A. sericea will be a threat to F.
ananassa. The absence of concern for A. sericea in
European strawberry pest management literature supports this assertion.
Conclusions
There was no evidence to support a herbivorehost
relationship between A. sericea and any Fragaria spp.
These results, combined with evidence from the entomological and strawberry pest management literature
and the lack of any comparable data or evidence in
the literature supporting such an insectplant association, strongly suggest that existing records of such a
relationship are erroneous. At the least, it can be confidently stated that the populations of A. sericea inhabiting the region of northern Bulgaria from whence the
test insects for these studies came do not attack the
two species implicated by Konow (1901) and Krishtal
(1959), F. vesca and F. viridis, nor do they pose any
threat to the cultivated strawberry, F. ananassa. For
the purpose of ensuring host fidelity in a weed BCC
such as A. sericea, this is a satisfactory conclusion, as
biological control agents are selected on a populationspecific basis due to the need for absolute certainty in
avoidance of non-target effects. The studies presented
in this paper also confirm the use of D. laciniatus as
food plant for A. sericea.
Acknowledgements
The authors would like to thank Dr. D. Smith of USDAARL-SEL, Washington, DC and Dr. A. Taeger of DEI,
Mnester, Germany for their assistance in locating literature and for specimen identification. Thanks also to
Dr. Taeger, Dr. R. Sobhian of USDA-ARS-EBCL,
France and Dr. M. Volkovich of the Zoological Institute
of St. Petersburg, Russia for translation of literature.
Thanks to Dr. W.G. Meikle of USDA-ARS-EBCL for
his comments on the manuscript. Blagodarya to Dr. K.
Kojuharova of the Dept. of Botany, Agricultural University, Plovdiv, Bulgaria who provided identification of
wild strawberry species, Dr. M. Velichkova-Kojuharova
of The Plant Protection Institute, Kostinbrod, Bulgaria
for virus identification, and to O. Todorov for technical
assistance.
References
Andr, E. (1879) Species des Hymnoptres dEurope et
dAlgrie, vol. 1. Beaune, France, pp. 2932.
315
XII International Symposium on Biological Control of Weeds
APG II (2003) An update of the angiosperm phylogeny
group classification for the orders and families of flowering plants. Botanical Journal of the Linnean Society 141,
399436.
Bailey, L.H. (1951) Manual of Cultivated Plants. Macmillan,
New York, pp. 731733.
Berland, L. (1947) Faune de France, vol. 47: Hymnoptres
Tenthredinoides. Paul Lechevalier, Paris, pp. 422423.
Cameron, P. (1890) A Monograph of the British Phytophagous Hymenoptera. (Tenthredo, Sirex and Cynips, Linn),
vol. 3. Ray Society, London, UK.
Dalla Torre, C.G. (1894) Tenthredinidae incl. Uroceridae
(Phyllophaga & Xyllophaga) Catalogus Hymenopterorum hucusque descriptorum systematicus et synonymicus.
Lipsiae 1, 359360.
Enslin, E. (1917) Die Tenthredinoidea Mitteleuropas VI. Deutsche Entomologische Zeitschrift (Beiheft 6), 585586.
Ermolenko, V.M. (1972) Rogochvosti ta Pillshchiki. Cimbicidi. Blasticotomidi. Fauna Ukrainii. Kiev 10, 159164.
Harizanov, A., and Harizanova, V. (1998) Pests on cultivated plants and their identification. Zemizdat, Sofia, pp.
295138.
Hokanson, K.E., Smith, M.J., Connor, A.M., Luby, J.J., and
Hancock, J.F. (2006) Relationships among subspecies of
New World octoploid strawberry species, Fragaria virginiana and Fragaria chiloensis, based on simple sequence
repeat marker analysis. Canadian Journal of Botany 84,
18291841.
Kangas, E. (1946) ber die Gattung Abia Leach (Hym., Tenthredinidae) im Lichte ihrer europischen Arten. Annales
Entomologici Fennici 12, 77122.
Konow, F.W. (1901) Systematische Zusammenstellung der
bisher bekannt gewordenen Chalastogastra (Hymenopterorum subordo tersius). Zeitschrift fr systematische Hymenopterologie und Dipterologie 1, 161176.
Krishtal, O.P. (1959) Komachi Shkidniki silskogopodarskich
roslin v umovach lisostepu ta polissja Ukraini. Vyl-vo
Kyivsk, Kyiv, Ukraine, p. 90.
Lacey, L.A. and Brooks, W.M. (1997) Initial handling and
diagnosis of diseased inscets. In: Lacey, L.A (ed.) Manual
of Techniques in Insect Pathology. Academic, London,
pp. 115.
Lamarque, C. and Bossennec, J.-M. (2001). Hypermdia
en procetion des plantes: Fraisier. Unit de phytopatologie et methodologies de la detection. INRA, Versailles-
Grignon, France. Available at: http://www.inra.fr/Internet/
Produits/HYPPZ/Cultures/3c---028.htm(accessed22January
2008).
Liston, A.D. (1995) Compendium of European Sawflies.
Chalastos Forestry. Dailbersdorf 6, Gottfieding, Germany,
190 pp.
Liston, A.D. (1997) Hostplant list for European and North
African Megalodontoidea and Tenthredinoidea (Hymenoptera). Sawfly News 1, 3058.
Lorenz, H. and Kraus, M. (1957) Die Larvalsystematik der
Blattwespen (Tenthredinoidea und Megalodontoidea).
Abhandlungen zur Larvalsystematik der Insekten, Berlin
1, 1389.
Magis, N. (2001) Apports la chorologie des Hymnoptres
Symphytes de Belgique et du Grand-Duch de Luxem-
bourg, XXIII. Notes Fauniques de Gembloux 43, 39
46.
Markova, M. (1973) Fragaria spp. In: Jordanov, D. (ed) Flora
of Bulgaria, vol. 5. Bulgarian Academy of Sciences, Sofia, 269 pp.
Marks, D. (2008) Garden action: Garden pest and disease con
trol and protection. Available at: http://www.gardenaction.
co.uk/techniques/pests/plant_pest_disease_centre.htm
(accessed 22 January 2008).
Marohasy, J. (1998) The design and interpretation of hostspecificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News and
Information 19, 13N20N.
Rector, B.G., Harizanova, V., Sforza, R., Widmer, T., and Wiedenmanna, R.N. (2006) Prospects for biological control of
teasels, Dipsacus spp., a new target in the United States.
Biological Control 36,114.
Sforza, R. (2004) Candidates for biological control of teasel,
Dipsacus spp. In: Cullen, J.M., Briese, D.T., Kriticos,
D.J., Lonsdale, W.M., Morin, L. and.Scott, J.K. (eds)
Proceedings of the XI International Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra,
Australia, pp. 155161.
Singhurst, J.R., and Holmes, W.C. (2001) Dipsacus fullonum
(Dipsacaceae) and Verbesina walteri (Compositae):New
to Texas. Sida: Contributions to Botany 19, 723725.
Stein, R. von (1883) Tenthredinologische Stidien. 4. Neue
oder wenig bekannte Afterraupen. Entomologie Nachrichten 9 (1718), 206213.
Taeger, A., Altenhofer, E., Blank, S.M., Jansen, E., Kraus, M.,
Pschorn-Walcher, H., and Riztau, C. (1998) Kommentare
Pflanzenwespen Deutschlands (Hymenoptera, Symphyta).
In: Taeger, A. and Blank, S. (eds) Pflanzenwespen
Deutschlands. Deutsches Entomologisches Institut, Verlag Goecke and Evers, Keltern, pp. 49135.
USDA-NRCS (2008) The PLANTS Database, Version 3.5.
Available at: http://plants.usda.gov (accessed 22 January 2008) (Data compiled from various sources by
M.W.Skinner. National Plants Data Center, Baton Rouge,
LA 70874-4490 USA).
USDI-NPS (US Department of Interior, National Park Service) (2005) Alien plant invaders of natural areas. Available
at: http://www.nps.gov/plants/alien/list/d.htm (accessed
22 January 2008).
Vasilev, I.B. (1978) Fauna Bulgarica, vol. 8: Hymenoptera,
Symphyta. Aedibus Academiae Scientiarum Bulgaricae,
Sofia, Bulgaria, pp. 2224.
Verlaque, R. (1985) Etude biosistematique et phylogenetique
des Dipsacaceae. III. Tribus des Knautiae et des Dipsacaceae. Revue de Cytolologie et de Biolologie Vegetale Le
Botaniste 8, 171243.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
Werner, P.A. (1975) The biology of Canadian weeds: 12. Dipsacus sylvestris Huds. Canadian Journal of Plant Science
55, 783794.
Wright, A. (1990) British sawflies: a key to the adults of
genera occurring in Britain. AIDGAP. Field Studies 7,
531593.
316
The cereal rust mite, Abacarus hystrix,
cannot be used for biological
control of quackgrass
A. Skoracka1 and B.G. Rector2
Summary
Quackgrass, Elymus repens (L.) Gould, is a perennial grass spreading by vigorous underground rhizomes. Because of its capacity for rapid spread and persistence, it is considered as a common weed
in many settings worldwide. The cereal rust mite, Abacarus hystrix (Nalepa), is a polyphagous, phytophagous mite attacking quackgrass and many other grasses including wheat. Its feeding causes leaf
discoloration and inhibition of seed production. This mite can also transmit plant pathogens. Its role
as a potential biological control agent for quackgrass control was considered since previous work had
suggested that populations of this mite colonizing quackgrass may be specifically adapted to that host.
The ability to colonize wheat by these quackgrass population should, however, be first excluded. The
aim of this study was to estimate whether the cereal rust mite quackgrass population can colonize
wheat. For this purpose, female mites from quackgrass were transferred and subsequently reared on
quackgrass (control, n = 132) and wheat (n = 125). Colonization ability was assessed by comparing
the mean oviposition rate, mean female survival and mean number of progeny on each host. Mites
had similar success in the colonization of both quackgrass and wheat. The conclusion is that the
quackgrass population of cereal rust mite is well adapted to wheat and thus cannot be considered as a
potential agent against quackgrass.
Keywords: Acari, Eriophyidae, wheat, quackgrass.
Introduction
Quackgrass, Elymus repens (L.) Gould [ex. Agropyron
repens (L.) P. Beauv.], is a common and cosmopolitan species, occurring almost worldwide. It reproduces vegetatively by vigorous underground rhizomes,
which, in turn, develop axillary buds that are capable
of developing into new rhizomes and daughter shoots.
Quackgrass is a highly aggressive, sod-forming, perennial grass native to Eurasia. Because of its invasiveness and persistence, it is considered as a serious weed
of agronomic crops, turfgrass, landscapes, grasslands,
gardens, lawns and nurseries in many parts of the
world. In the USA, it is listed as a noxious and invasive weed introduced from Europe (Palmer and Sagar,
1963; Hultn and Fries, 1986).
1
Adam Mickiewicz University, Department of Animal Taxonomy and
Ecology, Institute of Environmental Biology, Faculty of Biology,
Umultowska 89, 61-614 Pozna, Poland.
2
USDA-ARS, European Biological Control Laboratory, Montpellier,
France.
Corresponding author: A. Skoracka <[email protected]>.
CAB International 2008
Quackgrass is most effectively and commonly controlled by a combination of chemical and cultural methods. Herbicides for its control are available for most
crops (e.g. Kells and Wanamarta, 1987; Ivany, 2002;
Ivany and Sanderson, 2003). A few arthropods that
live and feed on quackgrass are known, including Hydraecia spp. (Lepidoptera: Noctuidae) (Giebink et al.,
2000), Delia coarctata Fallen (Diptera: Anthomyiidae)
(Marriot and Evans, 2003) and Abacarus hystrix (Nalepa) (Acari: Eriophyidae) (Frost and Ridland, 1996).
However, no information on the biological control of
this species was found in the literature.
The cereal rust mite (CRM), A. hystrix (Nalepa), is
a phytophagous mite belonging to the family Eriophyidae. Eriophyid mites are often considered to be promising biological control agents for weeds because they
debilitate their hosts by their feeding, they can transmit
diseases to their hosts in certain cases and they tend
to be highly host-specific (Rosenthal, 1996). Feeding
on A. hystrix causes leaf discoloration and inhibition
of host seed production. The mite is known to transmit
plant pathogens, including ryegrass mosaic virus, a serious disease of temperate grasslands, and Agropyron
317
XII International Symposium on Biological Control of Weeds
mosaic virus, a minor disease of quackgrass (Oldfield
and Proeseler, 1996). Thus, the mite is able to seriously
damage its host plants. However, up to now, A. hystrix was not considered as a possible biological control
agent due to its wide host range. The host range of the
CRM includes records from at least 60 grass species,
many of which are of economic importance (E. de Lillo
and J. Amrine, 2007, personal communication). Among
the hosts, the mite is dispersed passively by wind
currents.
High host specificity is an extremely important characteristic in a biological control candidate. Above all,
a weed biological control agent should not attack nontarget plants, especially important crops. Despite many
grass species having been recorded as hosts of A. hystrix, its role as an agent of quackgrass control was considered since previous work has shown that some CRM
populations colonizing quackgrass may be specifically
adapted to that host. Previous observations have shown
that certain quackgrass and ryegrass populations of the
CRM neither accept nor survive on each others hosts
(Skoracka and Kuczyski, 2006; Skoracka et al., 2007).
Moreover, the existence of reproductive barriers among
these populations has been found (A. Skoracka, 2007,
personal communication), suggesting that host populations of A. hystrix may represent a complex of species.
If a quackgrass-associated population of the CRM is
highly specific and adapted only to this host, it should
be considered as a potential biological control agent of
quackgrass control. Because quackgrass often grows
near wheat, the potential for wheat colonization by a
quackgrass-associated population of A. hystrix should
be first excluded. Specifically, the aim of this study
was to estimate whether the quackgrass-specific CRM
population can colonize wheat.
Methods and materials
Two grass species were used as host plants in this study:
quackgrass and wheat Triticum aestivum L. Quackgrass rhizomes were collected in September 2006 from
a study plot in Pozna, Poland (5224.5N, 1653.0E;
elevation 87 m) and put in boxes with sandy soil. The
plants were kept at room temperature and exposed to
artificial light for 19 h per day. To protect the plants
from infestation by mites, insects or fungi, the boxes
were covered with nylon taffeta fastened to a wooden
frame. When sufficiently grown, plants were used for
the preparation of the stock mite colony and the experiment. A stock colony of mites was established with
individuals collected from quackgrass from the same
study plot and from two other study plots (5225.0N,
1655.0E, elevation 63 m, and 5223.0N, 1652.0E,
elevation 79 m.) in October 2006. Females from fieldcollected plants were randomly selected and transferred
to the un-infested plants in the laboratory. A detailed
description of the stock colony preparation can be
found elsewhere (Skoracka and Kuczyski, 2004). The
colony was maintained in the laboratory (20C, 40%
1 RH, 1617:78 L/D) for 6 weeks. Afterwards, mites
were used for the experiment.
Plants for the experimental were prepared as follows. Laboratory-grown quackgrass shoots were transplanted to pots, and wheat shoots were grown from
seed. There was one grass shoot in each pot. Ten to
fifteen females were transferred from the quackgrass
stock colony to grasses grown in pots. Females were
transferred under a stereomicroscope using an eyelash
glued to a preparatory needle and put to the leaf of the
grass. Two combinations were tested: (1) QQ females
from quackgrass transferred to quackgrass (control),
n = 134 and (2) QW females from quackgrass transferred to wheat, n = 125. Ten replicates were carried
out for each combination. A replicate was defined as
a single plant with 10 to15 mites transferred onto it.
Each plot was covered on with nylon taffeta and maintained at 20C, 40% 1 RH for 14 days. Afterwards,
the plants were checked, and mites (number of experimental females and their progeny: eggs, larvae, nymphs
and adults) were counted using a stereomicroscope.
Three components of colonization ability were
measured for each treatment: (1) mean oviposition rate
(total number of eggs oviposited by all females within
the trial/total number of females tested within the trial),
(2) mean female survival rate (total number of females
survived within the trial/total number of females tested
within the trial) and (3) mean number of progeny (eggs,
larvae, nymphs and adults tallied separately) within
each trial. Bootstrap (Efron and Tibshirani, 1993) was
used to compute 95% confidence intervals (CI). Differences between means were tested using the criterion
of CI overlapping (i.e. means were regarded as significantly different when their CI did not overlap) and
using a Hotellings T2-test and t-test.
Results
No significant differences were shown between the
ability to colonize quackgrass and wheat by the quackgrass-associated mites (T2: F3,16 = 0.49, p < 0.6946).
Specifically, there were no significant differences in
mean oviposition rate (t = 0.56, df = 18, p = 0.5865),
mean female survival (t = 0.36, df = 18, p = 0.7215) or
mean number of progeny between mites developing on
quackgrass and those developing on wheat (Table 1,
Fig. 1).
Table 1.
318
Trials
QQ
QW
Mean oviposition rate and mean female survival
with 95% confidence limits (in parentheses) of
Abacarus hystrix that developed on quackgrass
(QQ) and wheat (QW).
Oviposition
13.9 (11.7-16.9)
13.1 (11.8-14.3)
Survival
0.5 (0.40.7)
0.5 (0.30.6)
The cereal rust mite, Abacarus hystrix
80
QQ
QW
Number
60
40
20
0
Figure 1.
egg
larva
nymph
adult
Mean number of progeny of Abacarus hystrix obtained on quackgrass (QQ) and wheat
(QW). Bars represent 95% confident limits around means.
Discussion
The idea of using eriophyoid mites for biological control of weeds has been of great interest since the 1970s.
The characteristics that make eriophyids promising
candidates for biological control are their frequent monophagy (or frequent specificity directed to one host),
ability to suppress plant growth and reproduction, ability to destroy whole plant populations under favourable conditions, attack on all plant organs, particularly
the inflorescences, and seed suppression (Rosenthal,
1996). A few eriophyid species have been used in biological control programs targeting weeds, e.g. Aceria
chondrillae (Canestrini) to control Chondrilla juncea
introduced to Australia and USA (Anders, 1983) and
Aceria malherbae Nuzzaci to control Convolvulus arvensis in USA (Boldt and Sobhian, 1993). Many other
species have been investigated and recommended for
biological control, e.g. Aceria tamaricis (Trotter) (De
Lillo and Sobhian, 1994), Aceria centaureae (Nalepa)
(Sobhian et al., 1989), A. salsolae De Lillo & Sobhian
(Sobhian et al., 1999), Cecidophyes rouhollahi Craemer (Sobhian et al., 2004), Epitrimerus taraxaci Liro
(Petanovic, 1990), Phyllocoptes nevadensis Roivainen
(Littlefield and Sobhian, 2000) and Floracarus perrepae Khinicki et Boczek (Freeman et al., 2005). Several European eriophyoids were suggested by Boczek
and Petanovic (1996) for the control of weeds in North
America in the genera Cirsium, Lythrum, Convolvulus and Galium. In addition, Boczek and Chyczewski
(1977) found eriophyid mites in Poland associated with
56 weedy species.
Among the eriophyid mites causing notable damage to their grass hosts, A. hystrix is regarded as one of
the most common and important. The mite has a great
capacity for rapid population increase achieving high
population densities (Skoracka and Kuczyski, 2004).
At very high densities, the feeding of this mite causes
plant wilting and delayed growth (Frost and Ridland,
1996). However, this study shows that the quackgrass-associated CRM had high fecundity, survival
and development on wheat. It clearly appears that this
quackgrass-associated population has great colonization ability and is well adapted to wheat and cannot
be considered as a potential biological control agent
against quackgrass.
Acknowledgements
We thank Dr Lechosaw Kuczyski (AMU, Pozna) for
remarks regarding the manuscript. The study was financially supported by Polish MNiSW grant 3P04C03825.
References
Anders, L.A. (1983) Considerations on the use of phytophagous mites for the biological control of weeds. In: Hoy,
M.A., Cunningham, G.L. and Knutson, L. (eds) Biological control of pests by mites. University of California,
Berkeley, CA, pp. 5356.
Boczek, J. and Chyczewski, J. (1977) Eriophyid mites (Acarina: Eriophyoidea) occurring on weedy plants in Poland.
Roczniki Nauk Rolniczych 7, 109113.
319
XII International Symposium on Biological Control of Weeds
Boczek, J. and Petanovic, R. (1996) Eriophyid mites as agents
for the biological control of weeds. In: Moran, V.C. and
Hoffmann, J.F. (eds) Proceedings of the IX International
Symposium on Biological Control of Weeds. University of
Cape Town, South Africa, pp. 127131.
Boldt, P.E. and Sobhian, R. (1993) Release and establishment
of Aceria malherbae (Acari: Eriophyidae) for control of
field bindweed in Texas. Environmental Entomology 22,
234237.
De Lillo, E. and Sobhian, R. (1994) Taxonomy, distribution, and host specificity of a gall-making mite, Aceria
tamaricis (Trotter) (Acari: Eriophyoidea), associated with
Tamarix gallica L. (Parietales: Tamaricaceae) in southern
France. Entomologica (Bari) 28, 516.
Efron, B. and Tibshirani, R.J. (1993) An Introduction to the
Bootstrap. Chapman and Hall, London.
Freeman, T.P., Goolsby, J.A., Ozman, S.K. and Nelson, D.R.
(2005) An ultrastructural study of the relationship between the mite Floracarus perrepae Knihinicki & Boczek
(Acariformes: Eriophyidae) and the fern Lygodium microphyllum (Lygodiaceae). Australian Journal of Entomology 44, 5761.
Frost, W.E. and Ridland, P.M. (1996) Grasses. In: Lindquist,
E.E., Sabelis, M.W. and Bruin J.W. (eds) Eriophyoid Mites
Their Biology, Natural Enemies and Control. Elsevier
Science, Amsterdam, The Netherlands, pp. 619629.
Giebink, B.L., Scriber, J.M. and Wedberg, J. (2000) Survival
and growth of two Hydraecia species (Noctuidae: Lepidoptera) on eight midwest grass species. Great Lakes Entomologist 32, 247256.
Hultn, E. and Fries, M. (1986) Atlas of North European Vascular Plants North of the Tropic of Cancer. Koelz Scientific Books, Knigstein, 498 pp.
Ivany, J.A. (2002) Banded herbicides and cultivation for
weed control in potatoes (Solanum tuberosum L.). Canadian Journal of Plant Science 82, 617620.
Ivany, J.A. and Sanderson, J.B. (2003) Quackgrass (Elytrigia
repens) control in potatoes (Solanum tuberosum) with
clethodim. Phytoprotection 84, 2735.
Kells, J. and Wanamarta, G. (1987) Agropyron repens control
with selective postemergence herbicides. Weed Technology 2, 129132.
Littlefield, J.L. and Sobhian, R. (2000) The host specificity of
Phyllocoptes nevadensis Roivainen (Acari: Eriophyidae),
a candidate for the biological control of leafy and cypress
spurges. In: Spencer, N.R. (ed.) Proceedings of the X International Symposium on Biological Control of Weeds.
Montana State University, Bozeman, MT, pp. 621626.
Marriot, C. and Evans, K.A. (2003) Host plant choice and
location by larvae of the wheat bulb fly (Delia coarctata).
Entomologia Experimentalis et Applicata 106, 16.
Oldfield, G.N. and Proeseler, G. (1996) Eriophyoid mites as
vectors of plant pathogens. In: Lindquist, E.E., Sabelis,
M.W. and Bruin, J. (eds) Eriophyoid Mites Their Biology, Natural Enemies and Control. Elsevier Science, Amsterdam, The Netherlands, pp. 259273.
Palmer, J.H. and Sagar, G.R. (1963) Agropyron repens (L.)
Beauv. (Triticum repens L.; Elytrigia repens (L.) Nevski).
Journal of Ecology 51, 783794.
Petanovic, R.U. (1990) Host specificity and morphological
variation in Epitrimerus taraxaci Liro (Acarida: Eriophyoidea). Zastita Bilja 41, 387394.
Rosenthal, S.S. (1996) Aceria, Epitrimerus and Aculus species and biological control of weeds. In: Lindquist, E.E.,
Sabelis, M.W. and Bruin, J.W. (eds) Eriophyoid Mites
Their Biology, Natural Enemies and Control. Elsevier
Science, Amsterdam, The Netherlands, pp. 729739.
Skoracka, A. and Kuczyski, L. (2004) Demography of the
cereal rust mite Abacarus hystrix (Acari: Eriophyoidea)
on quackgrass. Experimental and Applied Acarology 32,
231242.
Skoracka, A. and Kuczyski, L. (2006) Is the cereal rust mite,
Abacarus hystrix really a generalist? testing colonization performance on novel hosts. Experimental and Applied Acarology 38, 113.
Skoracka, A., Kuczyski, L. and Rector, B.G. (2007) Divergent host-acceptance behavior suggests host specialization in populations of the polyphagous mite Abacarus
hystrix (Nalepa) (Acari: Prostigmata: Eriophyidae). Environmental Entomology 36, 899909.
Sobhian, R., Katsoyannos, B.I. and Kashefi, J. (1989) Host
specificity of Aceria centaureae (Nalepa), a candidate for
biological control of Centaurea diffusa De Lamarck. Entomologica Hellenica 7, 2730.
Sobhian, R., McClay, A., Hasan, S., Peterschmitt, R. and
Hughes, R.B. (2004) Safety assessment and potential of
Cecidophyes rouhollahi (Acari, Eriophyidae) for biological control of Galium spurium (Rubiaceae) in North
America. Journal of Applied Entomology 128, 258
266.
Sobhian, R., Tunc, I. and Erler, F. (1999) Preliminary studies on the biology and host specificity of Aceria salsolae DeLillo and Sobhian (Acari: Eriophyidae) and Lixus
salsolae Becker (Col., Curculionidae), two candidates for
biological control of Salsola kali. Journal of Applied Entomology 123, 205209.
320
Refining methods to improve pre-release
risk assessment of prospective agents:
the case of Ceratapion basicorne
L. Smith,1 M. Cristofaro,2 C. Tronci3 and R. Hayat4
Summary
Ceratapion basicorne (Illiger) (Coleoptera: Apionidae) is a univoltine weevil native to Eurasia whose
larvae develop in root crowns of yellow starthistle, Centaurea solstitialis L. (Asteraceae). This insect
was rejected as a prospective biological control agent about 15 years ago after preliminary evaluation of its host plant specificity showed that it could develop on safflower (Clement et al., 1989).
However, it is known to attack very few plant species in the field and has never been reported from
safflower. We conducted a series of no-choice, choice and field experiments to measure the risk that
this insect would pose to non-target plants. Larval development occurred on nine species, including
Carthamus tinctorius (safflower) and Centaurea cyanus (bachelors button and cornflower). These
host plants are within a small monophyletic clade within the Centaureinae. Three years of field studies conducted in eastern Turkey, at three sites with natural populations of the insect, demonstrated
that the weevil does not damage safflower plants despite attack rates of 4898% on C. solstitialis. A
combination of taxonomic analyses and a hierarchy of host-specificity experiments were required to
determine that this insect would be safe to introduce.
Keywords: host plant specificity, risk assessment, field experiments, phylogeny.
Introduction
Determination of the host plant specificity of a prospective biological control agent plays a key role in the process of selecting and approving new biological control
agents of weeds. Today, many consider host-specificity
evaluation to be so routine that some professional journals refuse to publish such work. However, there still
remains a great deal of art to this work, which suggests room for improving the science. The goal of hostspecificity evaluation is to predict the behaviour of
agents after they are released into a new environment.
To do this accurately is no small feat. Such predictions
are usually based on the results of highly artificial laboratory experiments conducted under the constraints of
working inside quarantine space. Furthermore, the tolerance of the public and regulatory agencies for risk of
1
USDAARS, 800 Buchanan Street, Albany, CA 94710, USA.
ENEA C.R. Casaccia, Via Anguillarese 301, 00060 S. Maria di Galeria,
Rome, Italy.
3
Biotechnology and Biological Control Agency, Via del Bosco, 10,
00060, Sacrofano, Rome, Italy.
4
Ataturk University, Faculty of Agriculture, Plant Protection Department, 25240 Erzurum, Turkey.
Corresponding author: L. Smith <[email protected]>.
CAB International 2008
2
injury to non-target species has been decreasing. The
combined effect is to reduce the probability of finding
agents that can be approved for release. Can improving
our methods help counter this trend?
There is a substantial literature reviewing methods
of host-specificity evaluation, e.g. Clement and Cristofaro (1995), Withers et al. (1999), van Driesche (2000),
Spafford Jacob and Briese (2005) and Sheppard et al.
(2005). These reviews discuss many of the parameters
that should be considered in the design of experiments.
However, the choice of design should always reflect
knowledge of the life history, ecology and behaviour
of the agent. Thus, accurate testing requires customizing the experimental design for a particular agent on
a particular target weed. A flow chart (Fig. 1) developed by Sheppard (1999) provides a starting point for
choosing the most appropriate types of host-specificity
experiments for a particular agent. However, given
the multitude of experimental parameters that must be
chosen for any experiment, these guidelines are only a
beginning (Marohasy, 1998).
In this paper, we present an example of a prospective
biological control agent that was once rejected based
on preliminary experimental observations. However,
because natural history observations suggested that
the agent was more specific than what was indicated
321
XII International Symposium on Biological Control of Weeds
Figure 1.
Decision tree for choosing general type of host specificity test (Sheppard, 1999). When adults of prospective agents are difficult to obtain or rear, the tendency may be to conduct larval transfer experiments.
However, if the agent is naturally selective at the oviposition stage, then oviposition or field tests will give
more realistic results.
by the experimental results, we conducted more thorough experiments, which have shown that the insect
is indeed suitable for introduction. Nevertheless, preliminary studies are a critical early step in the process
of focusing research efforts to efficiently develop new
biological control agents. We will discuss what kinds
of experiments are most likely to improve the reliability of such preliminary studies.
Yellow starthistle, Centaurea solstitialis L. (Asteraceae: Cardueae), is an herbaceous winter annual that is
adapted to a Mediterranean climate: mild wet winters
and dry hot summers (Keil and Turner, 1993; Roch
and Roch, 2000). It is native to Eurasia and was introduced to the west coast of the USA over 100 years
ago (Maddox, 1981). Seeds usually germinate in the
autumn after the onset of winter precipitation. Rosettes
grow during winter and spring, bolt in May to June
and flower continually until frost or lack of moisture
kills the plant. The plant has been the target of classical biological control in the USA since the late 1960s,
but despite the introduction of six seed-head-attacking
insects, it is not yet under control over most of its range
(Turner et al., 1995; Piper, 2001; Pitcairn et al., 2004,
2006). This suggests the need for agents that attack vegetative parts of the plant and Ceratapion basicorne (Illiger) (Coleoptera: Apionidae) was considered a likely
prospect (Zwlfer, 1965; Rosenthal et al., 1994).
C. basicorne is distributed throughout Europe and
southwestern Asia, from Spain to Azerbaijan (AlonsoZarazaga, 1990; Wanat, 1994). It commonly infests
yellow starthistle in Turkey and Greece (Rosenthal et
al., 1994). This species has primarily been reared from
yellow starthistle, but there are also reports of rearing it
from Centaurea cyanus L. (bachelors button and cornflower), Centaurea depressa Bieb. (which is very similar to C. cyanus), and in one case, Cnicus benedictus
L. C. benedictus has recently been placed in the Jacea
group of Centaurea (which includes yellow starthistle),
based on phylogenetic analysis of DNA (Garcia-Jacas
et al., 2000). Thus, the insect has only been reared from
a few species of plants in the Jacea and Cyanus groups,
within the genus Centaurea. Adults have been found
resting on plants only in the Cardueae tribe, which includes Centaurea.
C. basicorne adults emerge from hibernation in
the early spring and feed on yellow starthistle leaves
(Clement et al., 1989; Smith and Drew, 2006). Females
lay eggs in the leaves of rosettes from late March to
early May (as observed in central Italy). Eggs hatch after about 10 days at room temperature, and first instar
larvae mine in the leaf blade and down the petiole to the
root crown. Larvae feed primarily in the root crown,
complete development in about 2 months and pupate
inside the plant. Adults emerge in June, feed on yellow
322
Refining methods to improve pre-release risk assessment of prospective agents
starthistle leaves for a few days and then disappear.
They are thought to aestivate and hibernate in secluded
places, and adults have been found under tree bark in
July (Hayat et al., 2002). Newly emerged females are
in reproductive diapause, and although they mate, they
are not able to lay eggs until completion of hibernation.
In the spring, after feeding for 12 weeks, females lay
a few eggs per day for 12 months before dying (Smith
and Drew, 2006).
Clement et al. (1989) studied a population of C.
basicorne in Italy. Because the species is univoltine
and newly emerging adults do not oviposit, they used
adults collected in the field on wild yellow starthistle
plants. This greatly limited the number of adults that
they could test (nine females). These adults were used
in a few choice and no-choice feeding and oviposition experiments. Adults fed and oviposited on safflower under no-choice conditions but preferred yellow
starthistle under choice conditions. Neonate larvae
from these females were transferred to holes made in
the central meristem of test plants. Larval development
was observed on three non-target plants, including safflower. Because of these results, further evaluation of
the insect was abandoned.
Considering that C. basicorne has never been reported as a pest of safflower and that field records indicate that it develops in only a few species of Centaurea,
we decided to conduct a more thorough study of its
host specificity.
Methods
No-choice oviposition
Individual mated females that had completed reproductive diapause were placed in a clear plastic tube (3.5
11 cm) mounted on an intact rosette leaf of a nontarget plant species for 4 to 5 days (Smith, 2007). Each
trial was preceded and followed by a positive control:
placing the female with a cut yellow starthistle leaf for
2 to 3 days to determine if she could still oviposit. For
each valid trial, we recorded adult feeding damage, oviposition and larval development. In general, we tested
eight replicates per plant species in the tribe Cardueae
and four in the more distantly related taxa. We doubled
the number of replicates if there were any signs of larval development.
Lab choice
An ovipositing female was placed inside a wooden
sleeve box (73 43 43 cm; length, width and height)
containing cut leaves of four to five species of test
plants for 5 days. Each species was represented by a
cluster of two cut leaves held in a vial of water with a
foam stopper. Yellow starthistle leaves were included
as a positive control in each trial. Adult feeding damage
and oviposition were recorded. The number of valid
replicates ranged from four to 18 for each of six nontarget species tested.
Field choice
We conducted experiments during 3 years (2002
2004) at three sites in eastern Turkey (Askale, Horasan,
at) where C. basicorne was naturally abundant (Smith
et al., 2006). We tested two accessions of yellow
starthistle: US (seed collected in Davis, California)
and Turkey (seed collected at the three Turkish sites)
and two commercial safflower, C. tinctorius L., varie
ties CW1221 (linoleic, CalWest) and S317 (oleic, SeedTec). All test plants were transplanted into the field sites
as soon as C. basicorne feeding damage was observed
on wild yellow starthistle plants. Plants were harvested
as soon as C. basicorne pupae were observed in wild
yellow starthistle plants and were either dissected or
individually bagged to allow adults to emerge. Adult
insects were identified by either Dr. Enzo Colonnelli
(University of Rome La Sapienza, Italy) or Dr. Boris
Korotyaev (Russian Academy of Sciences, St. Petersburg). Larvae were preserved in 99% ethanol for DNA
extraction.
Results and discussion
No-choice oviposition
In no-choice oviposition tests, C. basicorne females
oviposited on 94% of plant species in the subtribe Centaureinae (Smith, 2007), including safflower and the
native species Centaurea americana Nutt. and Centaurea rothrockii Greenm. (results for Centaureinae are in
Fig. 2). There was no larval damage to plants outside
the tribe Cardueae. The highest occurrence of insect
larval development was observed on C. solstitialis
and C. cyanus, but there was significant development
on C. melitensis L. (tocalote), C. benedictus (blessed
thistle), safflower and Crupina vulgaris Cass. (common crupina). These results corroborate the previous
observation that C. basicorne can develop in safflower
(Clement et al., 1989). However, neither safflower nor
bachelors button appears to be normal hosts for C. basicorne because they do not form a rosette. Thus, when
young larvae tunnel down a leaf on either of these
plants, they cannot reach the root crown. Such larvae
develop in the woody outer portion of the stem, not in
the central pith. The relatively thin cortex provides a
limited space for the insect and as the plant continues
to grow, it sometimes crushes the pupa.
Lab choice
In the sleeve-box choice experiment, adult feeding and oviposition was significantly greater on yellow starthistle than on any of the six other non-target
323
XII International Symposium on Biological Control of Weeds
Figure 2.
Proportion of trials in which a female oviposited on test plants (eggs) and in which insects completed
development to at least pupal stage (pupae) in the no-choice host-specificity experiment. Individual females were held inside a plastic tube attached to the leaf of a non-target test plant for 5 days and on yellow
starthistle for 2 to 3 days.
species tested. About 72% of eggs were deposited on
yellow starthistle, 20% on bachelors button, 5% on C.
melitensis, 1% on C. americana and 1% on safflower
(Fig. 3). These results indicate that C. basicorne females are more attracted to yellow starthistle than to
bachelors button or any of these other non-target test
plants. However, bachelors button and, to a lesser de-
gree, safflower appear to be at risk of some attack, at
least under these confined laboratory conditions.
Field choice
The infestation of the yellow starthistle test plants
was between 48% and 92% (US and Turkish plants
324
Refining methods to improve pre-release risk assessment of prospective agents
Ca. tinctorius
Eggs (x10)
Total FH
Ce. rothrockii
Ce. americana
Ce. sulphurea
Ce. melitensis
Ce. cyanus
Ce. solstitialis
0
Figure 3.
Table 1.
10
20
40
60
Number per 5-day trial
Infestation of root crowns or lower stems of
test plants by apionid weevils (including larvae,
pupae and adults) during 3 years at three field
sites in eastern Turkey. Ceratapion basicorne
was reared from yellow starthistle but not from
safflower (Smith et al., 2006).
Test plant
Yellow
starthistle
Safflower
No.
safflower
plants
sampled
(US)
(Turkey)
Oleic
Linoleic
83 b
28 b
59 b
100 a
67 a
87 a
0c
0c
19 cb
0c
0c
16 cc
45
38
40
70
80
90
37 a
45 a
77 a
0b
8 bd
0b
57
39
98 a
100 a
0b
26 be
combined) at the three sites, indicating that there was
a substantial infestation rate to challenge the safflower
plants (Table 1). We reared 87, 145 and 297 individuals of Ceratapion from yellow starthistle at Horasan,
at and Askale, respectively, but many were immature and others were in poor condition for taxonomic
determination. All the adults that could be identified
were C. basicorne (29, 30 and 92 from Horasan, at
and Askale, respectively), except for two C. orientale
(Gerstaecker). No safflower plants were infested by internal feeding insects at either Horasan or at. Thirty
safflower plants were infested at Askale, but none of
the insects reared from these plants were C. basicorne.
We reared only C. scalptum (Mulsant and Rey), C. orientale and C. onopordi (Kirby) from these safflower
plants. Subsequent identification of preserved larvae
using molecular genetics has confirmed these results
(Antonini et al., 2008, this proceedings).
Conclusions
250
99
Values followed by the same letter in the same row are not significantly different (chi-square test, P < 0.01)
b
Adults identified: 4 C. scalptum, 1 C. orientale, 2 C. onopordi
c
Adults identified: 2 C. scalptum
d
Three unidentifiable adults
e
Adults identified: 8 C. scalptum, 2 C. orientale
a
50
Oviposition and adult feeding by Ceratapion basicorne during choice oviposition experiments in sleeve
boxes (one female for 5 days exposed to cut leaves of four to five plant species at a time, always including
yellow starthistle). Number of eggs was multiplied by ten for visibility on the same scale; FH Number of
adult feeding holes, each approximately 1 mm2; error bars = SE.
Proportion of plants
infested (%)a
Site
2002
Horasan
at
Askale
2003
at
Askale
2004
Horasan
Askale
30
Although C. basicorne can develop on safflower in
laboratory experiments, we never found evidence of
attack during 3 years of field studies. The results of
our no-choice, choice and field experiments concur
with current theory that no-choice experiments tend
to overestimate risk to non-target plants under field
conditions (physiological vs ecological host range).
To avoid rejecting prospective agents that might actually be suitably host-specific, more emphasis should be
placed on observations in the natural environment (i.e.
325
XII International Symposium on Biological Control of Weeds
Type of Test
Strong evidence of specificity?
No
Field screening test
Yes
Larvae mobile?
Yes
No
Larval feeding tests
Offspring selectively distributed? No
Yes
Adults in lab?
Yes
Adult oviposition test
No
Figure 4.
Revised decision tree for choosing the general type of host-specificity test. The emphasis is on what stages
are capable of selecting the host. When suitable adults cannot be obtained by rearing or field collection,
then field experiments should be used rather than larval transfer. Use of choice or no-choice experiments
could be appropriate for any of these three types of tests.
natural history). This includes both examining non-
target plants during foreign exploration and conducting
relatively simple field experiments. Preliminary studies
should focus on the most critical results for the particular agent. In this example, the critical issue was the
attack on C. americana, C. rothrockii and safflower because they are the closest relatives to yellow starthistle
deemed beneficial in North America. Preliminary experiments should focus on the developmental stage and
conditions under which the agent chooses the target; in
this case, females ovipositing on rosette leaves in early
spring. Sheppards (1999) flowchart provides a good
guide, but with the caveat that for species in which the
female chooses the host plant, adult oviposition should
be tested regardless of whether the agent is easy to rear
in the lab. The modified flowchart (Fig. 4) shifts the
emphasis away from whether or not the agent is easy
to rear in the laboratory towards which stage or stages
choose the host plant. Conducting larval no-choice experiments for a species whose larvae are not capable of
choosing their host plant is more likely to mistakenly
eliminate it than if the adults were tested.
References
Alonso-Zarazaga, M.A. (1990) Revision of the sub-genera
Ceratapion S. Str. and Echinostroma Nov. of the genus
Ceratapion Schilsky, 1901. Fragmenta Entomologica.
Roma 22, 399528.
Antonini, G., Audisio, P., De Biase, A., Mancini, E., Rector, B.G., Cristofaro, M., Biondi, M., Korotyaev, B.A.,
Bon, M.C. and Smith, L. (2008) The importance of molecular tools in classical biological control of weeds: two
case studies with yellow starthistle candidate biocontrol
agents. In: Julien, M.M., Sforza, R., Bon, M.C., Evans,
H.C., Hatcher, P.E., Hinz, H.L. and Rector, B.G. (eds)
Proceedings of the XII International Symposium on Biological Control of Weeds, Montpellier, France. CAB International, Wallingford, pp. 263269.
Briese, D.T. (2005) Translating host-specificity test results
into the real world: The need to harmonize the yin and
yang of current testing procedures. Biological Control 35,
208214.
Clement, S.L. and Cristofaro, M. (1995) Open-field tests in
host-specificity determination of insects for biological
control of weeds. Biocontrol Science and Technology 5,
395406.
Clement, S.L., Alonso-Zarazaga, M.A., Mimmocchi, T.
and Cristofaro, M. (1989) Life history and host range of
Ceratapion basicorne (Coleoptera: Apionidae) with notes
on other weevil associates (Apioninae) of yellow starthistle in Italy and Greece. Annals of the Entomological Society of America 82, 741747.
Garcia-Jacas, N., Susanna, A., Mozaffarian, V. and Ilarsian, R.
(2000) The natural delimitation of Centaurea (Asteraceae:
Cardueae): ITS sequence analysis of the Centaurea jacea
group. Plant Systematics and Evolution 223, 185199.
Hayat, R., Guclu, S., Ozbek, H. and Schon, K. (2002) Contribution to the knowledge of the families Apionidae and
Nanophyidae (Coleoptera: Curculionoidea) from Turkey,
with new records. Phytoporasitica 30, 2537.
Keil, D.J. and Turner, C.E. (1993) Centaurea. In: Hickman,
J.C. (ed.) The Jepson manual: higher plants of California.
University of California Press, Berkeley, California, pp.
222223, 227.
326
Refining methods to improve pre-release risk assessment of prospective agents
Maddox, D.M. (1981) Introduction, phenology, and density
of yellow starthistle in coastal, intercoastal, and central
valley situations in California. US Department of Agriculture, Agricultural Research Service, Agricultural
Research Results, ARR-W-20, July 1981, USDA-ARS,
Oakland, CA.
Marohasy, J. (1998) The design and interpretation of hostspecificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News & Information 19, 1220.
Piper, G.L. (2001) The biological control of yellow starthistle
in the western U.S.: four decades of progress. In: Smith,
L. (ed.) Proceedings of the First International Knapweed
Symposium of the Twenty-First Century, March 1516,
2001, Coeur dAlene, Idaho. USDA-ARS, Albany, CA,
USA, pp. 4855.
Pitcairn, M.J., Piper, G.L. and Coombs, E.M. (2004) Yellow
starthistle. In: Coombs, E.M., Clark, J.K., Piper, G.L. and
Cofrancesco, Jr., A.F. (eds) Biological Control of Invasive Plants in the United States. Oregon State University
Press, Corvallis, OR, pp. 421435.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, J.
(2006) Yellow starthistle continues its spread in California. California Agriculture 60, 8390.
Roch, C.T. and Roch, Jr., B.F. (2000) Identification of knapweeds and starthistles in the Pacific Northwest. Pacific
Northwest Extension Publication PNW432. Washington
State University, Pullman, University of Idaho, Moscow,
Oregon State University, Corvallis, Montana State University, Bozeman, MT, 22 pp.
Rosenthal, S.S., Davarci, T., Ercis, A., Platts, B. and Tait, S.
(1994) Turkish herbivores and pathogens associated with
some knapweeds (Asteraceae: Centaurea and Acroptilon)
that are weeds in the United States. Proceedings of the
Entomological Society of Washington 96, 162175.
Sheppard, A.W. (1999) Which test? A mini review of test usage in host specificity testing. In: Withers, T.M., Barton
Browne, L. and Stanley, J. (eds) Host Specificity Testing
in Australasia: Towards Improved Assays for Biological
Control. Papers from the Workshop on Introduction of
Exotic Biocontrol Agents Recommendations on Host
Specificity Testing Procedures in Australasia, Brisbane,
October 1998. Scientific Publishing, Indooroopilly, QLD,
Australia, pp. 6069.
Sheppard, A.W., van Klinken, R.D. and Heard, T.A. (2005)
Scientific advances in the analysis of direct risks of weed
biological control agents to nontarget plants. Biological
Control 35(3), 215226.
Smith, L. (2007) Physiological host range of Ceratapion basicorne, a prospective biological control agent of Centaurea
solstitialis (Asteraceae). Biological Control 41, 120133.
Smith, L. and Drew, A.E. (2006) Fecundity, development
and behavior of Ceratapion basicorne (Coleoptera: Apionidae), a prospective biological control agent of yellow starthistle. Environmental Entomology 35, 1366
1371.
Smith, L., Hayat, R., Cristofaro, M., Tronci, C., Tozlu, G. and
Lecce, F. (2006) Assessment of risk of attack to safflower
by Ceratapion basicorne (Coleoptera: Apionidae), a prospective biological control agent of Centaurea solstitialis
(Asteraceae). Biological Control 36, 337344.
Spafford Jacob, H. and Briese, D.T. (eds) (2003) Improving
the selection, testing and evaluation of weed biological
control agents. Proceedings of the CRC for Australian
Weed Management Biological Control of Weeds Symposium and Workshop, University of Western Australia,
Perth, Sept. 13, 2002. CRC for Australian Weed Management Technical Series . 7.
Turner, C.E., Johnson, J.B., and McCaffrey, J.P. (1995) Yellow starthistle. In: Nechols, J.R., Andres, L.A., Beardsley,
J.W., Goeden, R.D. and Jackson, C.G. (eds) Biological
Control in the Western United States: Accomplishments
and Benefits of Regional Research Project W-84, 1964
1989. University of California, Division of Agriculture
and Natural Resources, Oakland. Publ. 3361, pp. 270
275.
van Driesche, R., Heard, T., McClay, A. and Reardon, R.
(eds) (2000) Proceedings of Session: Host Specificity of
Exotic Arthropod Biological Control Agents: The Biological Basis for Improvement in Safety. Forest Service, Morgantown, West Virginia, USA. FHTET-99-1.
Wanat, M., (1994) Systematics and phylogeny of the tribe
Ceratapiini (Coleoptera, Curculionoidea, Apionidae).
Genus, International Journal of Invertebrate Taxonomy
(Suppl. 3), 1406.
Withers, T.M., Barton Browne, L. and Stanley, J. (eds) (1999)
Host specificity testing in Australasia: towards improved
assays for biological control. Papers from the Workshop
on Introduction of Exotic Biocontrol Agents Recommendations on Host Specificity Testing Procedures in Australasia, Brisbane, October 1998. Scientific Publishing,
Indooroopilly, QLD, Australia.
Zwlfer, H. (1965) Phytophagous insect species associated
with Centaurea solstitialis L. in south-western Europe.
Report on Investigations Carried Out in 1965. Commonwealth Institute of Biological Control, Ascot, UK.
327
Host-specificity testing on Leipothrix
dipsacivagus (Acari: Eriophyidae),
a candidate for biological control of
Dipsacus spp.
A. Stoeva,1 B.G. Rector2 and V. Harizanova1
Summary
Leipothrix dipsacivagus Petanovic & Rector is the first eriophyid mite recorded from hosts in the
genus Dipsacus L. and is considered a potential candidate for biological control of invasive teasels
(Dipsacaceae). Host-specificity testing on L. dipsacivagus (Acari: Eriophyidae) was carried out under
insectary conditions for 4 months, from 19 April until the end of August, 2006. The laboratory colony
of L. dipsacivagus was descended from mites collected on Dipsacus laciniatus L. in Klokotnitsa,
Bulgaria. They were tested in choice and no-choice tests on Dipsacus fullonum L., Knautia arvensis
L., Cephalaria sp. and Scabiosa sp. In the choice experiments, individual D. laciniatus plants infested with L. dipsacivagus were placed in cages with one to two plants of each test species. There
were four replicates. Data were recorded at 10, 20, 30, 40, 60 and 90 days after infestation of test
plants. After 10 days, only D. laciniatus was infested. After 20 days, mites were vagrant on all the
plants except Scabiosa, and colonies were established on all Dipsacus plants. Some Cephalaria and
Scabiosa plants had vagrant mites at days 20 and 30, respectively, but all of these were dead by days
30 and 40, respectively. By day 60, all Knautia plants were colonized, although these colonies later
died. The mite successfully colonized only D. laciniatus and D. fullonum and temporarily colonized
K. arvensis. In the no-choice tests, each of four plants of a given test species was infested with five
mites/plant in a single cage. There were three replications (cages) for each test-plant species. Mites on
plants other than Dipsacus spp. began to die after 10 days. Reproducing populations of L. dipsacivagus established only on D. laciniatus and D. fullonum.
Keywords: teasel, invasive plants, mites, Dipsacaceae.
Introduction
Leipothrix dipsacivagus Petanovic & Rector is the first
eriophyid mite recorded from hosts in the genus Dipsacus L. (Petanovic and Rector, 2007). This species was
first collected in Serbia in 1999 but was misidentified
(Petanovic, 2001). It was subsequently collected during
surveys conducted in Serbia, Bulgaria and France in
2005, described as a new species (Petanovic and Rector, 2007), and is now a candidate for biological control
of invasive teasels (Dipsacus spp., Dipsacaceae) in the
Agriculture University, Faculty of Plant Protection, Department of Entomology, 12 Mendeleev Street, 4000 Plovdiv, Bulgaria.
2
USDA-ARS, European Biological Control Laboratory, Campus International de Baillarguet, Montferrier-sur-Lez, France.
Corresponding author: A. Stoeva <[email protected], vili@au-
plovdiv.bg>.
CAB International 2008
USA. According to Petanovic and Rector (2007), the
mite occurs on both the upper and lower leaf surfaces
of Dipsacus spp. as a vagrant causing rust-like symptoms, wrinkles on the longitudinal folds of the leaves,
witches broom of the plant (i.e. reduced internode
length and deformed leaves), stunting, delayed flowering and malformation of the flower heads. Analyses
of symptomatic plant tissues for presence of microbial
plant pathogens were negative (Petanovic and Rector,
unpublished data). A study of the injuries caused by the
mite (Pecinar et al., 2007) showed that the injury to the
leaves is conspicuous at the morphological and physiological as well as the anatomical level.
Eriophyid mites are frequently considered to be
promising candidates for weed biological control due
to their high host specificity, rapid life cycle and severe damage to their host plants (Littlefield and Sobhian, 2000; Rancic and Petanovic, 2002; Sobhian et
al., 2004). In one study, more than 80% of eriophyid
328
Host-specificity testing on Leipothrix dipsacivagus
species occurring on weeds were monophagous (Boczek and Petanovic, 1996), which is a key criterion for
a successful weed biological control agent. A critical
part of any biological control program is to test the
host range of candidate agents to establish the level of
risk posed to non-target plants and ensure that agent
releases will not do more harm than good (Cullen and
Briese, 2001).
The purpose of the present paper is to report on initial results concerning host-specificity testing of the
new eriophyoid mite L. dipsacivagus, a potential candidate for biological control of Dipsacus spp., in the
USA.
Methods
Origin and maintenance of test population
L. dipsacivagus individuals were collected from
cutleaf teasel, Dipsacus laciniatus L., plants in a field
near Klokotnitsa, Bulgaria (4200.43N, 2527.41E)
and brought to the insectary of the Department of Entomology, Agricultural University in Plovdiv, Bulgaria.
Species identification was made by R. Petanovic, Department of Entomology, Faculty of Agriculture, University of Belgrade, Serbia. The original mite colony
was set up in August 2005 and maintained under insectary conditions on potted D. laciniatus plants.
laciniatus plant was placed together with either three
or seven pots of a test plant species, depending on the
size of the cage. Plants were arranged in a randomized
design within the test cages (four pots in the small and
eight pots in the large cages). Pots were arranged such
that the leaves of adjacent test plants within each cage
were touching to facilitate migration of the mites between plants. For each plant species, four replications
were made with each cage representing a replication.
The plants were checked regularly under a stereomicroscope to monitor the migration of the mites. At 10,
20, 30, 40, 60 and 90 days, the number of test plants
with mites present was recorded. The infestation was
recorded as the number of infested plants in each cage
and the percentage of infestation was calculated.
No-choice test Potted plants of D. fullonum, K.
arvensis, Scabiosa sp. and Cephalaria sp., were tested.
D. laciniatus plants were tested in separate cages as a
control. Five adult mites were transferred under a stereo microscope to the leaves of each test plant, using
a fine brush. For each replicate, four plants of one test
species were arranged in a small cage with three replications per species. The plants were checked under a
stereomicroscope regularly. At 10, 20, 30, 40, 60 and
90 days, the infestation of test plants was recorded.
In both, the choice and no-choice tests migration,
reproduction and feeding damage were recorded in addition to the presence of mites.
Test plants
Several closely related plant species from the family Dipsacaceae were chosen: Dipsacus fullonum L.,
Knautia arvensis L., Scabiosa sp. and Cephalaria sp.
Plants of D. laciniatus, the original host of the colony,
were used as a control. The test plants were grown in
plastic pots 8 cm in diameter from field-collected seed.
Plants were used in tests after they had formed their
first two foliar leaves. Test plants were inspected to ensure that they were in healthy condition at the time of
testing.
Host-specificity testing
The host-specificity tests were conducted under laboratory conditions from 19 April until the end of August, 2006. Two types of cages made from clear, plastic
panels with fine nylon mesh tops were used for the experiments: small (20 20 40 cm) and large (20 40
40 cm). During the tests, there was approximately 16
h of light per day in the insectary with temperatures of
approximately 22C during the day and 15 C at night.
Relative humidity was 50% to 60% in the insectary
and 70% to 80% within the plastic cages. Two different
tests were designed for studying the host-specificity:
choice tests and no-choice tests.
Choice test Free migration was allowed from
any infested D. laciniatus plant to other plants in the
same cage. In each cage, one pot with an infested D.
Results and discussion
Choice test
The initial results from the choice tests are presented
in Table 1 and Fig. 1. By the tenth day after the beginning of the experiment, migration was observed only
on D. laciniatus plants. After 20 days, migration was
observed to all test species, except for Scabiosa sp. At
that time, all plants in all control cages were 100% infested, while in the test cages, infestation varied from
0% on Scabiosa sp. to 50% on D. fullonum plants.
After 30 days, 100% of the D. fullonum plants were
infested, and after 60 days, 100% of the K. arvensis
plants were infested. Migration to the other two species, Cephalaria sp. and Scabiosa sp., was observed at
20 and 30 days, respectively.
Soon after migrating, most of the mites on plants other
than Dipsacus spp. died. By the 30th day on Cephalaria
sp. and by the 40th day on Scabiosa sp., all the mites
on these plants were dead. No further migration was
observed to any of the test plants of these two species.
Besides the differences in the duration of dispersal,
differences in the development of the mite depending
on the host plant species were also noted. On the control (D. laciniatus), the mite began reproducing after
10 days, while on D. fullonum and K. arvensis plants,
mites were reproducing after 20 days and 30 days, respectively. After 30 days, there were no differences in
329
XII International Symposium on Biological Control of Weeds
Table 1.
Results of host-specificity testing with Leipothrix dipsacivagus at lab conditions at the Agricultural
University-Plovdiv in 2006.
Test plant species
Test 1. Choice test
Dipsacaceae
D. laciniatus L. (control)
Dipsacus fullonum L.
K. arvensis L.
Scabiosa sp.
Cephalaria sp.
Test 2. No-choice test
Dipsacaceae
D. laciniatus L. (control)
Dipsacus fullonum L.
K. arvensis L.
Scabiosa sp.
Cephalaria sp.
No of plants
tested
No of infested plants (days after infestation)
10 days
20 days
30 days
40 days
60 days
90 days
Feeding
damage
20
20
20
20
20
17
0
0
0
0
20
10
6
0
7
20
20
12
7
13
20
20
17
13
a
20
20
20
a
a
20
20
20
a
a
Yes
Yes
Yes
No
No
12
12
12
12
12
12
12
9 (3)
8 (4)
12
12
12
6 (6)
a
a
12
12
a
a
a
12
12
a
a
a
12
12
a
a
a
12
12
a
a
a
Yes
Yes
Yes
No
Number in brackets shows the number of plants with dead mites.
a
All mites are dead.
the activity of the populations on these two test-plant
species and the control. The mite was reproducing and
increasing its population density on D. laciniatus, D.
fullonum and K. arvensis plants (with one exception)
until the end of the experiment.
Feeding damage was observed only on plant species on which the mite successfully reproduced (i.e., D.
laciniatus, D. fullonum and K. arvensis). For all three
of these plant species, the oldest leaves of the rosettes
wilted and became chlorotic and eventually withered.
This feeding damage to rosette leaves under laboratory conditions was quite different from the damage to
bolting and flowering Dipsacus spp. plants, observed
by the authors in the field in 20042006, near the villages of Klokotnitsa (4200.43N, 2527.41E), Lozen
(4237.57N, 2330.34E), Gorski Izvor (4200.93N,
2526.14E) and Dalbok Izvor, Bulgaria (4211.34N,
2504.81E; Stoeva et al., unpublished data) nor like that
reported by Petanovic and Rector (2007). This could be
due to physiological differences between rosettes and
bolting or flowering plants or due to fundamental differences between field and insectary conditions.
On Scabiosa sp. and Cephalaria sp., although mites
had migrated, they did not establish populations. Approximately 20 to 25 days after migration, the mobile
forms had died. No eggs or immature stages were found
on these plants. There was some evidence of feeding,
although the leaves did not become chlorotic or dessicated.
An additional experiment was conducted for the test
plant species on which development of mite populations had established. From the cages, housing choice
tests involving D. fullonum and K. arvensis plants, the
D. laciniatus plant, which had been the plant originally
infested with mites in the choice-test cages, was re-
moved. The cages were left with only the test plants
(D. fullonum or K. arvensis), onto which mites had migrated and established populations. On the D. fullonum
plants, mite populations continued to develop, whereas
on K. arvensis, all mites died 30 days after removal of
the D. laciniatus plants.
The development of the population on K. arvensis
plants in the choice tests (when the test plants were
touching the infested D. laciniatus plant) and the cessation of population development after removal of the D.
laciniatus plant could be explained as a result of induction of a state of central nervous excitation (Marohasy,
1998) by the presence in the two Dipsacus spp. (but not
in K. arvensis), of volatiles or other compounds that
stimulate mite feeding or that are necessary for mite
reproduction.
The results from the choice test showed that L. dipsacivagus migrates, feeds, reproduces and establishes
sustained populations on D. laciniatus and D. fullonum,
while it can feed temporarily on K. arvensis but cannot
sustain itself on this host in the absence of D. laciniatus. Neither Cephalaria nor Scabiosa proved suitable
as hosts to L. dipsacivagus in this experiement.
No-choice test
Ten days after the beginning of the no-choice experiments, in which all plants were directly infested
(five mites per plant), there were still live mites in all
replications on all the tested plants (Table 1 and Fig. 1).
Mites began to die after 10 days on some of the Scabiosa sp. and Cephalaria sp. plants and after 20 days
on K. arvensis plants. By the 30th day, all the mites on
all the plants of these species were dead. There was
330
Host-specificity testing on Leipothrix dipsacivagus
Figure 1.
Infestation of Leipothrix dipsacivagus onto test plants in choice and no-choice tests under laboratory
conditions.
no reproduction on these species, although there was
some evidence of feeding. In contrast, on D. laciniatus
and D. fullonum plants, there were eggs and immatures
present on the tenth day and the population density
was increasing. After 40 days, feeding damage was
observed on these two species as described from the
choice test.
The results from the no-choice test show that the
mite L. dipsacivagus reproduces and establishes populations on D. laciniatus and D. fullonum plants, both of
which are noxious weeds in the USA, and cannot colonize the confamilial species K. arvensis, Scabiosa sp.
or Cephalaria sp. These preliminary results are promising, in that the effective host range of this mite appears
to be restricted well within the family Dipsacaceae, if
not within the genus Dipsacus. This would bode well
for the prospects of this mite as a candidate for bio-
logical control of invasive teasels in the New World, as
there are no native or economically important members
of the family Dipsacaceae there.
Due to ambiguities in the results of choice tests of
L. dipsacivagus on K. arvensis, further no-choice testing using larger initial infestations are under way. In
addition, testing remains for the remainder of the hostspecificity test list for this biological control candidate,
which comprises approximately 40 species, mostly
outside the Dipsacaceae, including rare and/or threatened native American plants in closely related families
(e.g. Caprifoliaceae).
Acknowledgements
The authors would like to thank Prof. Dr. R. Petanovic of
the Department of Entomology, Faculty of Agriculture,
331
XII International Symposium on Biological Control of Weeds
University of Belgrade, Serbia for the species identification the mite and O. Todorov for technical assistance.
Thanks also to J. Kashefi, USDA-ARS-EBCL, Thessaloniki, Greece for providing the seeds of some of the
plant species.
References
Boczek, J. and Petanovic, R. (1996) Eriophyid mites as
agents for biological control of weeds. In: Moran, V.C.
and Hoffman, J.H. (eds) Proceedings of the 9th International Symposium on Biological Control of Weeds. University of Cape Town, pp. 127131.
Cullen, J.M. and Briese, D.T. (2001). Host plant susceptibility to eriophyid mites for weed biological control. In: Halliday, R.B., Walter, D.E., Proctor, H.C., Norton, R.A. and
Colloff, M.J. (eds) Acarology: Proceedings of the 10th
International Congress. CSIRO Publishing, Melbourne,
pp. 342348.
Littlefield, J.L. and Sobhian, R. (2000) The host specificity of
Phyllocoptes nevadensis Roivainen (Acari: Eriophyidae),
a candidate for biological control of leafy and cypress
spurges. In: Spencer, N.R. (ed.) Proceedings of the X International Symposium on Biological control of Weeds.
Montana State University, Bozeman, Montana, USA, pp.
621626.
Marohasy, J. (1998) The design and interpretation of hostspecificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News and
Information 19(1), 13N20N.
Pecinar, I., Stevanovic, B., Rector, B.G. and Petanovic, R.
(2007) Anatomical injuries caused by Leipothrix dipsacivagus Petanovic and Rector on cut-leaf teasel, Dipsacus laciniatus L. (Dipsacaceae). Archives of Biological
Sciences 59, 363367.
Petanovic, R., (2001) Three new species of eriophyid mites
(Acari: Eriophyoidea) from Serbia with notes on new taxa
for the fauna of Yugoslavia. Acta Entomologica Serbica
4, 127137.
Petanovic, R.U. and Rector, B.G. (2007) A new species of
Leipothrix (Acari: Prostigmata: Eriophyidae) on Dipsacus
spp. in Europe and reassignment of two Epitrimerus spp.
(Acari: Prostigmata: Eriophyidae) to the genus Leipothrix.
Annals of the Entomological Society of America 100,
157163.
Rancic, D. and Petanovic, R. (2002) Anatomical alterations
of Convolvulus arvensis L. leaves caused by eriophyoid
mite Aceria malherbae Nuzz. Acta entomologica serbica
7(1/2), 129136.
Sobhian, R., McClay, A., Hasan, S., Peterschmitt, M. and
Hughes, R. B. (2004) Safety assessment and potential
of Cecidophyes rouhollahi (Acari, Eriophyidae) for biological control of Galium spurium (Rubiaceae) in North
America. Journal of Applied Entomology 128, 258266.
332
Impact of larval and adult feeding of
Psylliodes chalcomera (Coleoptera:
Chrysomelidae) on Centaurea solstitialis
(yellow starthistle)
C. Tronci,1 A. Paolini,1 F. Lecce,1 F. Di Cristina,1 M. Cristofaro,2
S.Ya. Reznik3 and L. Smith4
Summary
The flea beetle, Psylliodes chalcomera (Illiger) (Coleoptera: Chrysomelidae), is a promising candidate biocontrol agent for yellow starthistle, Centaurea solstitialis L., a weed of primary importance in
the western USA. Two sets of trials were performed to evaluate the impact of insect feeding at larval
and adult stages. In the first experiment, larval impact was assessed by inoculating different numbers
of neonate larvae on bolting yellow starthistle plants. The impact was evaluated on fresh/dry biomass,
size of plant, numbers of seed heads, seed production and germinability. The second experiment was
focussed on the evaluation of adult feeding behaviour, including possible gregarious aspects, which
could result in a relative increase of damage on yellow starthistle. The larval transfer test showed
a significant impact of larval feeding on yellow starthistle, although limited to seed production. P.
chalcomera adult feeding was negatively influenced by the presence of other individuals on the same
substrate. On the other hand, the individual feeding rates, especially for females, suggest a potential
defoliation impact. Comparing the outcome of both experiments with the infestation rates observed
in the field, P. chalcomera shows potential to damage yellow starthistle, although more studies are
needed to assess the impact of such damage.
Keywords: impact assessment, flea beetle, feeding damage.
Introduction
In 2001, a population of the flea beetle Psylliodes
chalcomera (Illiger) (Coleoptera: Chrysomelidae) was
observed developing on yellow starthistle, Centaurea
solstitialis L. (Asteraceae: Cardueae), in the vicinity
of the village of Volna, Krasnodar territory, southern
Russia. This insect is known to attack Carduus species
(Dunn and Rizza, 1976; Dunn and Campobasso, 1993),
but it had not been previously observed feeding on spe-
Biotechnology and Biological Control Agency, Via del Bosco 10,
00060 Sacrofano, Rome, Italy.
2
ENEA C.R. Casaccia, s.p. 25, Via Anguillarese 301, 00060 S. Maria di
Galeria, Rome, Italy.
3
Russian Academy of Sciences, Zoological Institute, St. Petersburg,
Russia.
4
USDAARS, 800 Buchanan Street, Albany, CA 94710, USA.
Corresponding author: C. Tronci <[email protected]>.
CAB International 2008
cies in the genus Centaurea. Subsequent behavioral
and molecular genetic studies showed that the population associated with yellow starthistle was biologically
different from populations associated with Onopordum
sp. in Russia and Carduus sp. in Italy (De Biase et al.,
2003, 2004). Consequently, we are evaluating individuals from this population to determine if they would be
suitable candidates for the biological control of yellow
starthistle. An important part of this evaluation is to assess the potential impact that the insect can have on the
target plant (Balciunas and Coombs, 2004).
Yellow starthistle is an important rangeland weed
in the western USA, especially in regions with a mild
moist winter and dry summer (Mediterranean climate;
Maddox, 1981; Sheley et al., 1999). It is a winter annual forb native to southern Europe and southwestern
Asia. Seeds germinate in the late fall or early spring,
rosettes develop until late spring and then plants bolt
and flower until they senesce due to lack of moisture
or freezing. P. chalcomera has one generation per year
333
XII International Symposium on Biological Control of Weeds
(Dunn and Rizza, 1976; Cristofaro et al., 2004). Overwintering adults feed on yellow starthistle foliage in
the early spring and oviposit on or near the plants. Larvae tunnel in the leaf midribs and young stems and exit
the plant to pupate in the soil. Emerging adults feed
briefly, mate and then aestivate and hibernate until the
following spring.
Methods
Insects
Experiments on the impact of larval and adult feeding were carried out utilizing 65 adults of P. chalcomera, collected near Volna (450736N; 364135
E; altitude 16 m) on March 28 to 31, 2005. Based on
our previous experience, adults collected at this time of
year should be completing diapause and ready to begin
feeding and ovipositing.
In the laboratory, the insects were placed in a 3-l
glass beaker with crumpled paper and fresh yellow
starthistle leaves at room temperature (18C to 25C)
and with a 16:8 L/D cycle (natural and artificial light
combined) for a week to allow complete reactivation
after diapause. During this time mating was often observed.
Females and males were then separated. Males were
kept in the same beaker described above, while females
were individually placed in Petri dishes (25C, 16:8 L/
D cycle) with small bouquets of fresh yellow starthistle
leaves with the aim to select ovipositing individuals.
Laid eggs were collected daily, counted and placed in
sterile Petri dishes over wet filter paper to allow hatching.
After 1 week, ovipositing females were put back to
the common container with males, where the insects
were allowed to feed on fresh yellow starthistle leaf
bouquets replaced every other day.
Larval transfer
The larval transfer test was carried out on early
bolting, US biotype, greenhouse grown potted yellow
starthistle plants. Two treatments and one negative control were set up transferring 10, 20 and 0 larvae per
plant, respectively, with ten replicates each. Before the
start of the experiment, plant height, root-crown diameter and number of internodes were recorded for each
plant.
First-instar larvae that emerged from the eggs maintained on filter paper were transferred with a fine brush
onto leaf axils, where larvae are known to enter the
plant under natural conditions. To help the access of
the larvae into the stem tissue, a small opening through
the leaf tricomes was dug at each leaf axil with a sharp
instrument. When enough neonate larvae were available, the transfers were done simultaneously on each
replicate.
This first phase of the test was carried out in laboratory at 18C to 26C and 16:8 L/D cycle (natural and
artificial light combined). Two weeks after the transfer
of the first larva, each plant was enclosed in a 60- by
23-cm fine (1 mm) nylon mesh cage, supported by an
inner aluminum frame and fastened around the outside of the pot with a 3-cm wide elastic band. The pots
were then moved to a shade house outside the laboratory (6C to 34C min/max, 18.7C mean temperature,
14:10 to 15:9 L/D conditions; April to June, 2005).
Forty days after transfer of the last larva, the cages
were inspected to recover emerged adults. Such inspections were repeated every other day until the 60th day
after the transfers, when all successfully developed
adults were considered to have emerged. Cages were
then removed and pot soil inspected for dead adults or
pupae. At this time, corresponding to the full flowering stage, half the plants from each treatment (n = 5)
were harvested, carefully cleaned and weighed, and the
number and stage of the flower buds, root-crown diameter and plant height were also recorded. Harvested
plants were then dissected to estimate the mean number
of galleries per plant, their position and average length.
After dissection, the material from each plant was put
together in paper bags, dried at 65C for 72 h in a ventilated stove and weighed.
The remaining plants (n = 5 for each treatment)
were left undisturbed to allow the completion of the
life cycle and seed production. When all flower heads
were senescent, plants were harvested, measured, dissected and weighed following the above described
methods. In addition, the seeds from ten mature flower
buds sampled from each plant were collected, counted
and separated into mature and immature. A germination test in Petri dishes was carried on sub-samples of
mature and immature seeds from each plant.
Adult feeding
The aim of this experiment was to assess the feeding
impact of variable numbers of P. chalcomera adults on
fresh cut yellow starthistle leaves by measuring the leaf
area consumed per day. In addition, we wanted to investigate if the presence of one or more other individuals of the same or opposite sex on the same substrate
would determine any significant increase or reduction
of the amount of leaf tissue consumed in a specific time
by an adult. The following combinations of males (M)
and females (F) were tested: 1F, 2F, 5F, 1M, 2M, 5M,
1F + 1M, 2F + 2M and 1F + 1M. Each treatment was
replicated ten times, except 5M (five replicates) and 5F
(three replicates).
The trials were carried out in sterile 10-cm diameter
Petri dishes in a climatic cabinet (25C, 50% to 60%
RH, 16:8 L/D cycle). The insects were allowed to feed
on one single freshly cut yellow starthistle leaf, laid on
a moist disk of filter paper for 24 h. The Petri dishes
334
Impact of larval and adult feeding of Psylliodes chalcomera on Centaurea solstitialis
also contained a moist sterile cotton plug (approximately 1 cm3), a substrate to encourage oviposition.
Before each trial, leaves were carefully washed,
rinsed and scanned into a TIF format digital image using a desktop flatbed scanner. At the end of each 24
h testing session, the leaves were again scanned. After each test, insects were placed back in a common
container for at least one full day before insects were
randomly selected for the next feeding trial. Eggs were
collected daily, counted and placed in sterile Petri
dishes over wet filter paper to allow hatching.
Image analysis
Before performing image analysis, each image was
edited with Photoshop LE v.5.0 (Adobe Systems, San
Figure 1.
Jose, CA) to colour the background and the eaten areas with two uniform tints (respectively, black and
red). Such edits were achieved selecting background
and eaten areas of the image with the magic wand tool,
adjusting the tolerance level to fit their full width and
finally coloring them with the paint bucket tool.
Next, the amount of leaf area eaten by adults of
P. chalcomera was evaluated utilizing public domain
software (Image v.4.0.3.2 beta for Windows, National
Institute of Health, Bethesda, MD). Scanned leaf image dimensions were converted from pixels into square
centimetres using the calibration function of the software and using a strip of scaled paper that was included
in all scans as reference (ONeal et al., 2002).
Once calibrated, the red channel of each image was
first inverted to negative and then converted to black
The length of galleries and the number of tunnels made by larvae of Psylliodes chalcomera
when 0, 10 or 20 first-instar larvae were transferred onto yellow starthistle (mean 95% CI).
335
XII International Symposium on Biological Control of Weeds
and white with the threshold function of the program.
The threshold level was manually adjusted to fit the actual eaten area that was finally measured with the measure function of the software.
Field infestation rates
P. chalcomera field infestation rates were assessed
in May 2001 near Volna and Primorskii (451508 N;
365217E; elevation 5 m) on yellow starthistle, Italian thistle, Carduus pycnocephalus L., and nodding
or musk thistle, Carduus thoermerii Weinm. In May
and June 2003, two samples were repeated on the same
population of yellow starthistle at Volna location only.
At each sampling, 20 to 65 whole plants from dense
patches were harvested, dissected and the numbers of
larvae were recorded. In 2003, the crown diameter and
height of the plants were also measured, and their fresh
weight was evaluated.
Statistical analysis
Data were analyzed using analysis of variance for
classified effects or linear regression for quantitative
effects. The relationship between total gallery length
and number of seeds per capitulum was modeled by using least squares nonlinear regression with the Weibull
equation using the quasi-Newton estimation method in
the computer program Statistica (release 5.1, StatSoft
Inc., Tulsa, OK).
Figure 2.
Results and discussion
Larval transfer experiment
Establishment of larvae, as measured by the number
of tunnels, was the same whether ten or 20 larvae were
transferred to yellow starthistle plants; however, the total length of tunnelling increased with the number of
larvae transferred (Fig. 1). Relatively few larvae were
able to complete their development up to the adult
stage: A total of five and seven adults were recovered,
respectively, from 20- and 10-larvae treatments, which
was only 2.5% and 7% survivorship, respectively. Although some larval mortality was expected, due to the
experimental conditions that required overcrowding
larvae on the plants to obtain the maximum impact, an
additional cause of mortality can probably be ascribed
to the dispersal behaviour of newly transferred larvae.
We observed some larvae moving away from the intended insertion point on the plant and they fell from
the plant and possibly died.
Larval damage (total tunnel length) did not affect
wet weight, dry weight, number of buds, number of mature flowers, number of senescent flowers or total number of capitula. This is probably because (1) these plant
characters are influenced by plant growth that occurred
before the insect damage occurred, (2) the plants were
able to compensate for the damage or (3) the damage
was too small to affect the plant. However, feeding by
P. chalcomera larvae significantly reduced the number
of seeds produced per capitulum (whether or not dry
The relationship between total gallery length, made by Psylliodes chalcomera tunneling in yellow starthistle, which increased with greater numbers of larvae, and the number of seeds per mature capitulum.
336
Impact of larval and adult feeding of Psylliodes chalcomera on Centaurea solstitialis
weight was included as a covariate). The data were fit
by the Weibull equation (Y = c exp(-((X/a)b)), where
c = 31.786 3.548 (SE), a = 3.839 2.465, b = 1.188
1.175, R = 0.830). The equation predicts that 8 cm of
larval tunnelling is able to reduce seed production by
90% and 14 cm by 99% (Fig. 2). This suggests that the
larval damage reduced the plants ability to provide nutrients to developing seeds. The number of capitula per
dry weight of plant decreased slightly with increasing
length of larval tunnelling (slope = -0.098 0.035 SE;
F(1, 27) = 8.0; P < 0.009), suggesting that larval damage
decreased resources available to develop capitula.
Adult feeding experiment
Females feed more than males [area eaten (1F vs
1M): females = 0.416 0.058 (SE), males = 0.027 +
0.007 cm2, F(1, 18) = 43.7, P < 0.0001; and feeding scars
per individual: females = 11.700 + 1.627 (SE), males
= 2.300 + 0.517, F(1, 18) = 30.33, P < 0.0001] (Fig. 3).
Feeding per insect was not affected by crowding at the
observed levels for either males or females, when each
Figure 3.
sex was analyzed alone (slope of linear regression for
area eaten or number of feeding scars per individual,
P > 0.05). However, there was a significant interaction
between sex and number of insects for both response
variables that was related to a tendency of feeding by
females to decrease with crowding while feeding by
males remained constant.
Field infestation rates
The infestation rate of yellow starthistle plants observed in southern Russia ranged from 36% to 90%
of yellow starthistle plants sampled in May or June
in 2001 and 2003 (Table 1). The number of larvae per
infested plant ranged from 2.3 to 6.9. No larvae were
observed in nearby plants of C. pycnocephalus at the
Volna site, but 20% of C. thoermeri were infested at
the Primorskyi site (Fig. 4). Although C. thoermeri
plants were larger than yellow starthistle, the number
of larvae per infested plant was lower. These results
suggest that either C. thoermeri is a more suitable host
than C. pycnocephalus or that the insect populations on
The effect of crowding different numbers of adult males and females on mean feeding rates per individual
of Psylliodes chalcomera on leaves of yellow starthistle during 24 h (1M one male, 1F one female, 2F2M
two females with two males etc.)
337
XII International Symposium on Biological Control of Weeds
Table 1.
The infestation rate and the number of larvae of P. chalcomera per infested plant species
observed at two sites in southern Russia (SE).
Location
Date
Host plant
Volna
Volna
Primorski
Primorski
Volna
Volna
10 May 01
10 May 01
11 May 01
11 May 01
19 May 03
16 June 03
C. solstitialis
Carduus pycnocephalus
C. solstitialis
Carduus thoermeri
C. solstitialis
C. solstitialis
Figure 4.
Figure 5.
Infestation
rate
38%
0%
36%
20%
63%
90%
No. larvae/
infested plant
2.33 0.44
0%
2.71 0.48
1.20 0.20
4.34 0.54
6.91 0.63
No. plants
sampled
40
20
50
50
65
48
The frequency distribution of number of larvae of Psylliodes chalcomera attacking some Cardueae plants in the field at sites in southern
Russia.
The relationship between number of larvae of Psylliodes chalcomera per plant of yellow
starthistle and plant height in the field at sites in southern Russia (sampled in May and June
2003).
338
Impact of larval and adult feeding of Psylliodes chalcomera on Centaurea solstitialis
these two plants may be genetically different. The latter
theory is supported by observations in laboratory host
plant specificity experiments and analysis of molecular
genetics, which indicate at least three host-associated
genetically isolated populations of this morphological
species (Cristofaro et al., 2004; De Biase et al., 2003,
2004).
The number of larvae per plant increased with plant
size, as represented by either plant height or root diameter (Fig. 5). Although plants were taller on the June
sample than in May [64.2 2.0 (SE) vs 34.1 1.4 cm],
sample date did not significantly affect the relationship
between the number of larvae and plant height. The
best-fitting regression equation was Y = -1.01 (0.90
SE) + 0.111(0.018) X, where X = plant height (cm)
(F1,111 = 39.00, P = 0.0001), but the regression only explained about a quarter of the variation. Because larvae
were probably too small to significantly affect plant
height at the time of sampling, the results suggest that
larger plants either attract higher rates of infestation or
that they are able to support more larvae (which may
be cannibalistic). If this tendency of finding more larvae on larger plants holds true, it is an attractive property for a prospective biological control agent because
higher numbers are probably necessary to impact larger
plants.
Conclusions
Feeding damage caused by larvae at the levels that we
observed in the laboratory, where they produced one
to five galleries and zero to two adults per plant, was
not sufficient to reduce plant size of yellow starthistle
plants. However, larval damage greatly reduced the
number of seeds per capitulum and the number of capitula per plant biomass, which is very interesting damage
to inflict on an annual plant. Transfer of about 12 larvae
corresponded to 8 cm of tunnelling, which cause 90%
reduction of seed production. In Russia, the number of
larvae observed in the field was much lower (generally
less than four larvae); however, it is not known whether
the studied population is limited by natural enemies. If
that was the case, then impact of this insect could be
substantial when established in a region lacking these
natural enemies (Torchin et al., 2003). Further studies
are ongoing to improve the survivorship of transferred
larvae in laboratory to evaluate the larval feeding impact with infestation rates similar as the ones recorded
in natural conditions.
Although adult females fed more (11.0 scars, 0.41
cm2 per day) than males (2.3 scars, 0.03 cm2) and, in
the field, feeding is likely to occur during 2 to 3 months
in the spring, unless they have gregarious behaviour or
preferentially attack young plants, this level of damage is not likely to significantly affect plant growth.
Our crowding experiment showed that the adults feed
less when crowded, and they generally have not been
observed to aggregate on plants in the field. However,
preference for or impact on young rosettes has not been
studied.
References
Balciunas, J.K. and E.M. Coombs. (2004) International code
of best practices for classical biological control of weeds.
In: Coombs, E.M., Clark, J.K., Piper, G.L. and Cofrancesco, A.F., Jr. (eds) Biological Control of Invasive Plants
in the United States. Oregon State University Press, Corvallis, OR, pp. 130136.
Cristofaro, M., Dolgovskaya, M. Yu., Konstantinov, A., Lecce,
F., Reznik, S. Ya., Smith, L.,Tronci, C. and Volkovitsh,
M.G. (2004) Psylliodes chalcomerus Illiger (Coleoptera:
Chrysomelidae: Alticinae), a flea beetle candidate for
biological control of yellow starthistle Centaurea solstitialis. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings
of the XI International Symposium on Biological Control
of Weeds. CSIRO Entomology, Canberra, Australia, pp.
7580.
De Biase, A., Antonini, G. and Audisio, P. (2003) Genetic
analyses on taxonomic status and mtDNA variation in
natural populations of Psylliodes spp. cfr. chalcomera
(Coleoptera, Chrysomelidae, Alticinae). Technical Report, University of Rome La Sapienza, Italy, 7 pp.
De Biase, A., Mancini, E. and Audisio, P. (2004) Genetic
analyses on taxonomic status and mtDNA variation in
natural populations of Psylliodes spp. cfr. chalcomerus
(Coleoptera, Chrysomelidae, Alticinae). Technical Report, University of Rome La Sapienza, Italy, 10 pp.
Dunn, P.H. and Rizza, A. (1976) Bionomics of Psylliodes
chalcomera, a candidate for biological control of musk
thistle [Carduus nutans]. Annals of the Entomological Society of America 69, 395398.
Dunn, P.H. and Campobasso, G. (1993) Field test of the weevil Hadroplonthus trimaculatus and the fleabeetle Psylliodes chalcomera against musk thistle (Carduus nutans).
Weed Science 41, 656663.
Maddox, D.M. (1981) Introduction, phenology, and density
of yellow starthistle in coastal, intercoastal, and central
valley situations in California. US Department of Agriculture, Agricultural Research Service, Agricultural
Research Results. ARR-W-20, July 1981, USDAARS,
Oakland, CA.
ONeal, M., Landis, D. and Isaacs, R. (2002) An inexpensive,
accurate method for measuring leaf area and defoliation
through digital image analysis. Journal of Economic Entomology 95, 11901194.
Sheley, R.L., Larson, L.L. and Jacobs, J.J. (1999) Yellow
starthistle. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State Univ. Press., Corvallis, OR, pp. 408416.
Torchin, M. E., Lafferty, K.D., McKenzie, V. J. and Kuris, A.
M. (2003). Introduced species and their missing parasites.
Nature 421, 628630.
339
Syphraea uberabensis (Coleoptera:
Chrysomelidae) potential agent for
biological control of Tibouchina herbacea
(Melastomataceae) in the archipelago
of Hawaii, USA
C. Wikler1 and P.G. Souza2
Summary
Biological invasions are one of the major threats to Hawaiis biodiversity. Herbacious glory tree,
Tibouchina herbacea Cogn. (Melastomataceae), native to America, is regarded as one of the harmful
plant species due to its fast growth, small wind dispersed seed and the absence of natural enemies.
Since 1998, potential agents have been studied in Brazil for a classical biological control. This paper
presents results from host range and impact tests conducted under field and laboratory conditions
for Syphraea uberabensis Bechyn (Coleoptera: Chrysomelidae), which is indicated as having great
potential as a control agent for T. herbacea. A detailed description of the adults of this insect has not
been published, and we describe it in this paper.
From a group of 20 species of plants in ten families investigated, S. uberabensis fed only on the
two species of Tibouchina demonstrating that Tibouchina supplies the physiological and biological
needs of this insect. S. uberabensis completes its life cycle on the leaves of T. herbacea and does not
attack other plant parts. During summer months, the life cycle is completed in approximately 35 days,
lengthening to 80 days in the cooler months. The main impact to the plant was caused by the thirdinstar larvae and adults, and the damage can kill plants in less than 2 weeks. In laboratory conditions,
25% of leaf damage caused leaf death and leaf drop. A batch of S. uberabensis was sent to the Quarantine Service of USDA in Hawaii in 2005 where further host-specificity tests are being conducted.
Keywords: host range, impact tests, Tibouchina, Hawaii.
Introduction
Herbaceous glory tree, Tibouchina herbacea Cogn.,
native to southeast Brazil, Uruguay, Paraguay and Argentina, has become a particularly troublesome species
in the Hawaiian Archipelago (Almasi, 2000). Its vigorous spread by tiny seeds and sprouts is beyond conventional control techniques, and it has been the target of
extensive field research in Brazil since early exploratory
work was conducted in 1994 by Burkhart (1994). Since
1998, a biological control research program targeting
UNICENTRO Forest Protection Laboratory, Rua Theresa P. Moura, 70,
Pilarzinho, 82100-440, Curitiba, Paran, Brazil.
2
UNICENTRO Forest Protection Laboratory, BR 153, km 7, Bairro
Riozinho 84500-000, Irati, Paran, Brazil.
Corresponding author: C. Wikler <[email protected]>.
CAB International 2008
this aggressive and tenacious weed has been underway
in southern Brazil, and potential biological agents are
being evaluated. Among them, the flea beetle Syphraea
uberabensis Bechyn (1955) is the highest ranked potential candidate according to the field and laboratory
studies conducted in Irati, Brazil (Mueller and Wikler,
2001). Though some species of Syphraea are known to
feed on the roots of the target weed, S. uberabensis has
only been observed feeding on leaves.
The genus Syphraea is described by Baly (1876) as
oval, compact, small black or blue-black flea beetles. S.
uberabensis are 34 mm in length and have a dark blue
color. The antennas have robust articles from the base
to the apex compared with the anterior tibia; the elytra have simple and very fine punctuations (Bechyn,
1955). This paper provides further description of S.
uberabensis, its biology and its impact on T. herbacea
and results of host specificity tests.
340
Syphraea uberabensis potential agent for biological control of Tibouchina herbacea
Methods and materials
Seedlings of T. herbacea from different provenances
were reared for the experiments in greenhouses at the
Irati campus. Cultures of S. uberabensis were obtained
from T. herbacea and Tibouchina cerastifolia Cogn.
plants that occurred naturally at the Irati Campus and
Mananciais da Serra, Paran State and from the locality
of Nova Petrpolis, Rio Grande do Sul State.
Insect rearing, biological observations and experiments were conducted under laboratory conditions in a
controlled temperature room ranging from 29 2C day
to 22 2C night, under artificial light banks on a 12-h
photo phase. Morphological descriptions were made of
insects that were reared under these conditions.
Biological observations and host range experiments
were conducted in Petri dish (10-cm diameter) and in
plastic bottles of 200 ml, with wet filter paper at the
bottom. Leaves of Tibouchina species were used to
feed individual larva and adult Syphraea. The larvae
and adults were examined every second day, and host
plant leaves were substituted daily to reduce the incidence of diseases and to provide fresh food.
In the plastic bottles, seedlings of both species of Tibouchina plants (20 to 30 cm tall) were placed in about
10 cm of sterilized soil collected from the same sites
as the plants to provide the whole plants to the insects.
Frass was not removed, and no additional plant material was provided.
Based on the phylogenetic system of Cronquist
(1981), 20 plant species that were closely related to
the T. herbacea and occurred in similar habitats were
selected for host-specificity tests (Hight et al., 2003).
Plants from the families Alzateaceae, Crypteroniaceae,
Oliniaceae, Penaeaceae, Punicaceae, Rhynchocalycaceae, Sonneratiaceae and Trapaceae were not included
in the tests due their absence in the study region.
For the no-choice tests, neonates larvae reared at
the laboratory and adults collected in the Irati Campus
field were placed in Petri dish containing moist filter
paper and leaves from the test plant. Four replicates of
each test plant were used. After all larvae and adults had
died in each dish, the leaves were observed for feeding
indicated by scraping of leaf epidermal cells and for
larval frass. Leaf area and consumed areas were measured by tracing the leaf outlines and the eaten areas
on millimetre paper and counting the areas damaged
or missing. Counts of the eggs and initial instars were
conducted using a Wild stereoscopic microscope at 10
and 40. Data were analyzed by conventional statistics
using Microsoft Excel.
Results
Description of adults of S. uberabensis
Body: Elongated, slightly broader posteriorly; robust
legs; thorax, abdomen, legs and antennae relatively
covered with fine short hairs; coloration deep metallic
blue, 2.8 0.10 mm long and 1.5 0.03 mm wide.
Head: Top of the head with strong depressed area
before frontal calli; punctuation behind frontal calli,
longitudinal carina narrow, elevated, eyes small and
oblong, entire; antennae more or less thickened.
Thorax: Pronotal punctures small, larger antebasal sulcus but not deeply impressed; lateral margins of pronotum
narrow; prosternal narrow with dense and long setation,
extending a little beyond posterior margin of procoxae;
metasternum densely covered with fine soft hairs.
Elytra: Large punctures densely distributed and deeply
impressed toward the base getting a little apex; elytra not
smooth; few short setae on posterior margin of elytra.
Biology of S. uberabensis
Most information presented in this paper is from the
laboratory experiments. When possible, we have provided ranges for summer and winter field conditions.
Mating occurred predominantly on the under surface
of the leaf toward the apex of the leaf, although, in
both laboratory and field situations, we occasionally
observed mating on the upper surface. Mating occurred
during the night, early hours of daylight and evening,
but on rainy days and periods of high humidity, it occurred throughout the day. Mating was effected by the
female partially opening their elytra not only to facilitate the appropriate juxtaposition of the male but also
allowing for a quick disassociation in case of predation.
We observed that mating frequency was reduced when
food was scarce, especially in autumn and winter when
the plants were in decline or reduced to their over-
wintering, short shoots.
The eggs were white and elliptical, measuring 0.58
0.01mm by 0.25 0.01mm. Oviposition was observed
only in the laboratory because it occurred principally
during the cooler hours of darkness and to a lesser extent the early morning or late afternoon. The female
pushed the leaf hairs apart with the hind legs and deposited eggs among the hairs on both surfaces of the
leaf and on the stem. On T. herbacea, the eggs were
interspersed between the abundant leaf hairs, but on T.
cerastifolia, the eggs were cemented on the leaf cuticle.
Eggs were normally laid singly, although rarely two or
three eggs were found together. It was not known if
these eggs were laid at the same time or that females
later visited the same site (Wikler and Souza, 2005).
Egg laying began 8 3 days after copulation in the lab
and approximately 12 days after copulation in the field
in autumn. The maximum number of eggs oviposited
during a single session of 3 h in the lab was 42. The
average number of eggs laid in the Petri dishes was 44
0.5 per week, although in the first eight weeks, it is
higher than 65 eggs per couple and during this period
most individuals died (n = 57 mating pairs; Fig. 1).
Eggs took between 12 and 21 days to hatch, and the
emerging larvae were active and mobile.
341
XII International Symposium on Biological Control of Weeds
120
100
EGGS
80
60
40
20
0
1 2
3 4
5 6
7 8
9 10 11 12 13 14 15 16 17 18 19 20 21
W EEKS
Figure 1.
Weekly average number of eggs oviposited by mated females of Syphraea uberabensis (Coleopera: Chrysomelidae) under laboratory conditions in a controlled temperature room.
First-instar larvae were 1.95 0.02 mm long, whitish and becoming yellowish on the second day. They
have three to four segments, and the head is not well
defined. This stage was brief lasting an average of 1.87
days. Feeding damage by these young larvae was almost imperceptible. The second-instar larva was light
yellowish color, with five to seven segments and a well
defined head, although it was often difficult to distinguish and was best observed from the underside. There
was a small posterior protuberance. Initially, the larvae
were 3 mm long, growing to 5 mm before ecdysis. This
stage lasted from 6 to 12 days. Feeding damage was
insignificant and did not exceed 1 cm2 per week. Thirdinstar larvae were dark yellow, with seven segments, a
well-defined head and a large posterior protuberance. It
was 6 mm long. Feeding damage was more than 1.27
cm2 per week, producing the most significant damage
to the plant, greater even than the adult. This stage
lasted 12 to 25 days, the first half feeding and the second half entering the soil in prepare for pupation. The
pupae were initially the same color as the third-instar
larva, darkening only a few days before eclosion.
Adults emerged as soon as 10 days after pupation,
although in the field in winter it took at least 30 days.
In the laboratory, of 48 insects, only 50% survived 7
weeks, 25% for 11 weeks and all had died by the end
of the 21st week
Mortality during development of this insect was
high; 65.7 17.2% of the eggs survived to first instar,
55.8% 21.4 % to the second instar, 51.2% 20.0% to
third instar and 7.6 7.7% of eggs survived to adulthood. In only one collect in Rio Grande do Sul, two
generalist Hemipterans (not yet identified) were found
attacking the adult insects in the field. Fungi also attacked the larvae and pupae during periods of high humidity in the laboratory.
Feeding occurred without preference on both
younger and older leaves. Visual observations of leaf
consumption by adults and especially third-instar larvae was extensive and resulted in reduced plant viability, lack of flower maturity and reduced seed production
(C. Wikler and P. Souza, unpublished data).
No-choice tests
In no-choice tests, S. uberabensis laid eggs on and
larvae and adults fed on plants in the genus Tibouchina
but not on any other test plant (Table 1). Additional
feeding trials conducted in the laboratory showed that
larvae completely defoliated T. herbacea in no-choice
tests and were able to complete development to adults
on this species. In choice tests, larval preferences for T.
herbacea and T. cerastifolia were equal.
Effects on T. herbacea and T. cerastifolia
In the experiments of biological impact, adults and
larvae of S. uberabensis demonstrated great potential
342
Syphraea uberabensis potential agent for biological control of Tibouchina herbacea
Table 1.
The list of test plants used in the host specificity tests for Syphraea S. uberabensis
and the results, where X indicates positive results.
Family
Anacardiaceae
Euphorbiaceae
Lythraceae
Melastomataceae
Monimiaceae
Myrtaceae
Onagraceae
Poaceae
Rosaceae
Thymelaeaceae
Species
Rhus sandwichensis Gray
Schinus terebinthifolius Raddi
Lithraea brasiliensis Marchand
Manihot esculenta Crantz
Phyllanthus tenellus Roxb.
Lafoensia pacari St. Hil.
T. herbacea ( DC. ) Cogn.
T. cerastifolia Cogn.
Peumus boldus Molina
Psidium cattleianum Sabine (red form)
Psidium cattleianum Sabine (yellow form)
Psidium guava L.
Campomanesia xanthocarpa O. Berg.
Eugenia uniflora L.
Eucalyptus grandis W. Hill ex. Maiden
Ludwigia sp
Bambusa vulgarisSchrad
Pyrus malus L.
Pyrus communis L.
Daphnopsis racemosa Griseb
to be used in the control of T. herbacea due the extensive damage caused to the plant. The leaves were skeletonized removing completely the plant matter, leaving
only the stem and vein structure. As a consequence,
plant growth was reduced, and flowering and consequently seed production were prevented.
The consumed leaf area was higher by the larvae
(third instar) that in the adult stage. The consumption
difference was not significant because, on average, the
larvae consumed less than half of a square centimetre
more than the adults. The weekly consumption area
of the adults was on average 1.15 cm2, approximately
12.8% of the total leaf area, and the consumption of
the larvae was on average of 1.28 cm2, approximately
14.2% of the total leaf area.
In the plastic bottles experiment, it was observed
that consumption of about 25% of the leaf was enough
to dry it out and cause leaf drop. The plant showed high
vulnerability to the S. uberabensis attack, which caused
the defoliation of all of the plants and they all died, on
average, after 4 weeks.
In the field, as in the laboratory, the leaves of both Tibouchina species demonstrated low or no regenerating
capacity after the attack of the S. uberabensis, drying
soon after a period of 2 weeks of the insect damage.
Discussion
Laboratory and field investigations confirmed that S.
uberabensis is strictly specific to the genus Tibouchina
and therefore a safe biological control agent for T.
herbacea in Hawaii. According to Harris (1971), the
loss of mature leaves is normally most damaging to
the plant, as these leaves represent the direct photo-
Adults
Larvae
Eggs
X
X
X
X
X
X
synthetic capacity of the plant. The attack by S. uberabensis is therefore meaningful, as no preferences based
on the age of the leaves were found. As result of these
studies and the potential of this insect as a biological
control agent, 2000 insects were sent in 2005 to the
Quarantine Service of USDA in Hawaii where further
host-specificity tests are being conducted.
Acknowledgements
For the kind assistance in field collection and laboratory experiments, we are very thankful to Alexryus
Augusto Altran, Jean Marcos Lubczyk, Ronan Felipe
de Souza and Mateus Marochi. We are very grateful
to Dr. Clifford W. Smith and the funding from US National Park Service via its Cooperative Studies Unit at
the University of Hawaii and US Geological Service,
Pacific Islands Research Center, through the Research
Corporation, University of Hawaii, as well as that from
CNPq. We also thank FUPEF and FAU for their administrative support.
References
Almasi, K.N. (2000) A non-native perennial invades a native
forest. Biological Invasions. Kluwer, The Netherlands.
Baly, J.S. (1876) Description of new genera and species of
Halticinae. Transactions of the Entomological Society of
London 3, 433449.
Bechyn, J. (1955) Quatrime Note Sur Les Chrysomeloidea
Notropicaux Des Collections de LInstitut Royal Des Sciences Naturelles de Belgique. Bulletin De LInstitut Royal
Des Sciences Naturelles de Belgique 31 (74), 112.
Burkhart, R. (1994) Natural Enemies of Tibouchina herbacea
Collections made in South America between December
343
XII International Symposium on Biological Control of Weeds
1993 and April 1994. Unpublished report. Hawaii State
Department of Agriculture.
Cronquist, A. 1981. An Integrated System of Classification of
Flowering Plants. Columbia University Press, New York,
NY.
Harris, P. (1971) Weed vulnerability to damage by biological
control agents. In: Dunn, P.H (ed) Proceedings of the 2nd
International Symposium of Biological Control of Weeds,
pp. 2939. Commonwealth Agricultural Bureaux, Farnham Royal, England.
Hight, S.D., Horiuchi, I., Vitorino, M.D., Wikler, C. and
Pedrosa Macedo, J.H. (2003) Biology, host specificity tests, and risk assessment of the sawfly Heteroper-
reyia hubrichi, a potential biological control agent of
Schinus terebinthifolius in Hawaii. BioControl 48, 461
476.
Mueller, Jr.V. and Wikler, C. (2001) Testes para utilizao de
Syphrea uberabensis Bechyn, 1955 (Coleoptera: Chrysomelidae) no controle biolgico de Tibouchina herbacea. In: UNICENTRO (ed.) Anais do XIII Seminrio de
Pesquisas, VIII Semana de Iniciao Cientfica. Guarapuava, PR. V.1 N.1 P. 334.
Wikler, C. and Souza, P.G. (2005) Estudos bioecolgicos
de Syphraea uberabensis (Coleoptera: Chrysomelidae)
Bechyn 1955. AMBIENCIA. Editora da UNICENTRO.
Guarapuava, PR V.1 N. 1, pp. 103112.
344
Host-specificity testing of Prospodium
transformans (Uredinales: Uropyxidaceae),
a biological control agent for use against
Tecoma stans var. stans (Bignoniaceae)
A.R. Wood1
Summary
Yellow bells, Tecoma stans (L.) Juss. ex Humb., Bonpl. & Kunth (Bignoniaceae), originally from
Meso- and South America, is an emerging weed in the warm, moist regions of South Africa. The microcyclic, gall-inducing, rust fungus Prospodium transformans (Ellis & Ever.) Cummins (Uredinales:
Uropyxidaceae) is a Neotropical pathogen of T. stans. It has been observed to be damaging to its host
under natural conditions and is being considered as a biological control agent for use against this
weed in South Africa. Two isolates were collected, one from Guatemala and the other from southern
Mexico, and established and maintained in quarantine in South Africa. Host-specificity testing was
conducted using both isolates against 14 species of Bignoniaceae (eight indigenous to southern Africa) and a further eight indigenous species in closely related families. No symptoms were produced
on any of these species, except for small chlorotic spots on Fernandoa magnifica Seem. Sporulating
galls were produced on all control T. stans plant leaves, petioles and stems. Leaves of the Bignoniaceae plants tested were examined microscopically at 7 days after inoculation. Fungal colonies of
approximately 300 m in diameter developed in control T. stans leaves, but the rust did not colonize
any other species. Therefore, P. transformans is considered safe for introduction into South Africa,
and permission for its release will be sought.
Keywords: South Africa, emerging environmental weed, yellow bells.
Introduction
Tecoma Juss. (Lamiales, Bignoniaceae) is a genus of
14 species, mainly occurring in the Neotropics but with
two species in Africa (Gentry, 1992). The genus can
be divided into two groups, one with narrow tubular
bird-pollinated flowers and the other with capanulate
bee-pollinated flowers. The two African species are included in the former group, which is the more diverse.
The latter group includes Tecoma stans (L.) Juss. ex
Humb., Bonpl. & Kunth and three other more narrowly
distributed segregate species (Gentry, 1992).
T. stans var. stans is a small tree with a widespread
natural distribution, occurring throughout Mesoamerica
and the Caribbean, as well as much of South America
(Gentry, 1992). Within this range, it is morphologically
ARC-Plant Protection Research Institute, P. Bag X5017, Stellenbosch,
7599, South Africa <[email protected]>.
CAB International 2008
variable, intergrading in places with the other two described varieties (var. velutina DC. and var. angustata
Rehder; Gentry, 1992).
This plant is naturalized in South Africa, where it
invades natural and disturbed vegetation and is therefore a declared weed (Henderson, 2001). Although not
yet regarded as a major weed, it is considered to have
the potential to invade a large proportion of the country
(Nel et. al., 2004). It is currently increasing in abundance and has been chosen as a target of a biological
control programme aimed at preventing it from emerging as a weed of national importance (Olckers, 2004).
Prospodium Arth. (Uredinales: Uropyxidaceae) is a
Neotropical genus of about 50 species predominantly
parasitizing members of the Bignoniaceae, with the
rest on the Verbenaceae (Cummins and Hiratsuka,
2003). One species, Prospodium tuberculatum (Speg.)
Arth., has been introduced into Australia for the biological control of Lantana camara L. (Tomley and Riding, 2002), and another, P. tumefaciens Lind., has been
proposed as a potential agent for use against Aloysia
345
XII International Symposium on Biological Control of Weeds
gratissima (Gill. et Hook.) Troncose in the USA (Cordo
and DeLoach, 1995).
Three Prospodium spp. have been recorded as occurring on T. stans, namely the macrocyclic P. appendiculatum (Wint.) Arth. and the microcyclic Prospodium
transformans (Ellis & Ever.) Cummins and P. elegans
(Schroet.) Cummins. The latter two are presumed to
have been derived by a contraction of the life cycle of
P. appendiculatum (Cummins, 1940). All three may be
considered as potential biological control agents for
use against T. stans in South Africa. One of these, P. appendiculatum, is adventitious in Brazil and is currently
being assessed for its effectiveness as a biological control agent against T. stans in that country (Vitorino et.
al., 2004).
In its native range (Caribbean basin, Guatemala,
Mexico), P. transformans has only been recorded from
T. stans var. stans and var. velutina (as T. mollis Humb.,
Bonpl. & Kunth) (Cummins, 1940). This rust fungus
causes galls up to 3 cm in diameter on petioles, stems
and seed pods, on which initially pycnia then telia develop. These are the only two stages of this species life
cycle. The teliospores may germinate as soon as they
develop (Shuttleworth, 1953). Because of its known,
limited host range and the damage that it causes to its
host plant in its native range, studies were undertaken
to assess the suitability of introducing P. transformans
in South Africa for the biological control of T. stans
var. stans. Results reported in this paper deal with hostspecificity testing of P. transformans before seeking
approval for release of this species in South Africa.
Lamiales (Table 1) were inoculated in the same manner
as above. Six plants of each species in the Bignoniaceae
were inoculated, three using the southern Mexican and
three using the Guatemalan isolates. Three plants of
each species in the other families were inoculated with
only the isolate from southern Mexico. The plants were
then observed for gall development and sporulation for
1 month after inoculation. For every batch of plants
inoculated, a plant of local T. stans was inoculated in
the same manner, at the same time. Only if sporulating
galls developed on these control plants were the results
recorded, otherwise the plants were re-inoculated. Every plant was inoculated twice, the second time on new
growth not previously inoculated.
Microscopic examination
Two plants of each of the tested Bignoniaceae species were inoculated as above. Seven days after inoculation, two leaves from each plant were harvested and
prepared for microscopic examination using the wholeleaf clearing and staining technique of Bruzzese and
Hassan (1983). The stained leaves were examined at
400 magnification for penetration and development of
mycelium by P. transformans. A control plant of local
T. stans was included for each inoculation.
Table 1.
Plant species
Bignoniaceae
Fernandoa magnifica Seem.
Fernandoa sp.
Jacaranda mimosifolia D. Don.
Kigelia africana (Lam.) Benth
Macfadyena unguis-cati (L.) A.H. Gentry
Markhamia obtusifolia (Baker) Sprague
M. zanzibarica (Bojer ex DC.) K. Schum.
Podranea ricasoliana (Tanfani) Sprague
Pyrostegia venusta (Ker Gawl.) Miers
Rhigozum obovatum Burch.
Spathodea campanulata P. Beauv.
Tecoma capensis (Thunb.) Lindl.
T. stans (L.) Juss. ex Humb., Bonpl. &
Kunth var. stans
Acanthaceae
Duvernoia adhatodoides E. mey. ex Nees
Mackaya bella Harv.
Oleaceae
Jasminum multipartitum Hochst.
Schrophulariaceae
Freylinia tropica S. Moore
Halleria lucida L.
Verbenaceae
Lantana rugosa Thunb.
Lippia rehmania H. Pearson
Lippia scaberima Sond.
Methods and materials
Source and maintenance of cultures
Two isolates of P. transformans, originally collected
from southern Mexico and Guatemala, were maintained in the quarantine glasshouses at ARC-Plant Protection Research Institute, Stellenbosch, South Africa,
by repeated inoculation of potted T. stans plants. The
plants were grown from seed collected in South Africa.
These two isolates were kept in separate glasshouses
to prevent cross-contamination. Plants were inoculated
by dusting dry teliospores on the petioles and adaxial
surfaces of immature leaflets (pinnae) using a small
paintbrush, spraying the leaflets with water using an air
brush or atomizer until very small droplets were visible to the naked eye, and then sealing the plants within
a plastic bag. The inoculated plants were placed in an
incubator at 18C (Shuttleworth, 1953) for 48 h and
then transferred to the quarantine glasshouses with a
day/night temperature cycle of 25/19C.
Host-specificity testing
Plants of indigenous and locally cultivated species
in the Bignoniaceae and other selected families of the
List of plant species included in host-specificity
testing of Prospodium transformans.
Origin
Af
Af
e
SA
e
Af
SA
SA
e
SA
Af
SA
e
SA
SA
SA
SA
SA
SA
SA
SA
SA Indigenous to South Africa, Af indigenous elsewhere in Africa, e
exotic to Africa.
346
Host-specificity testing of Prospodium transformans
Results
Host-specificity testing
On the control T. stans plants, chlorotic flecks were
visible on the leaf blades approximately 5 days after
inoculation. Galls began to develop on these flecks
soon after; small galls on the leaf blades (reaching approximately 5 mm in diameter after a month) but larger
galls on the petioles or stems (up to 30 mm long). Pycnia developed on the galls beginning approximately 12
days after inoculation, and then telia appeared approximately 18 days after inoculation. No symptoms developed on any of the plant species tested, except for small
chlorotic spots on Fernandoa magnifica Seem.
Microscopic examination
Mycelium of P. transformans colonized an area of
approximately 300 m diameter for each infection in
leaves of T. stans 7 days after inoculation, associated
with the chlorotic flecks visible to the naked eye. No
such mycelium was observed on any of the other Bignoniaceae species examined.
Small groups of dead epidermal cells were observed
on leaflets of one F. magnifica leaf, and empty basidiospores were still attached to many of these. Small crystals were concentrated in the surrounding epidermal
cells, and the underlying parenchyma cells appeared
more densely distributed compared to surrounding areas. On other leaves of F. magnifica, no dead cells were
observed, but areas of dense parenchyma cells and
large crystals occurred. No mycelium was observed in
or around these areas. The differences between these
leaf reactions were probably due to differences in leaf
age.
Single dead cells at the point of penetration were
observed in Tecoma capensis (Thunb.) Lindl., indicating a hypersensitive reaction. Neither mycelium
nor any plant reaction was visible for Fernandoa sp.,
Jacaranda mimosifolia D. Don., Macfadyena unguiscatii (L.) A.H. Genrty, Markhamia obtusifolia (Baker)
Sprague, Markhamia zanzibarica (Bojer ex DC.) K.
Schum., Podranea ricasoliana (Tanfani) Spraque,
Pyrostegia venusta (Ker Gawl.) Miers, Rhigozum obovatum Burch. or Spathodea campanulata P. Beauv.
Discussion
The Bignoniaceae is a small family in southern Africa,
having only 16 species in nine genera in this region
(Diniz, 1988; Smithies, 2003). Representatives of six
of these genera were tested and found not to become
infected with P. transformans, including T. capensis,
a congener of the target weed. It was observed microscopically that a hypersensitive reaction occurred in
T. capensis. Additionally tested members of the Bignoniaceae native to South America (cultivated and/or
naturalized in South Africa) and indigenous representatives of other families in the Lamiales all showed
no symptoms of infection. The only plant tested that
showed any symptoms (chlorotic spots) was F. magnifica. Microscopic examination revealed no fungal
mycelium associated with these chlorotic spots; rather,
dense parenchyma and accumulated crystals occurred.
These probably indicate that the chlorotic flecks were
due to a plant defence reaction.
Because of the high level of host specificity demonstrated, it is considered that this rust fungus is safe for
introduction into South Africa for the biological control of T. stans var. stans. This conclusion is supported
by the narrow host range recorded in its native distribution (Cummins, 1940). Permission for its release will
be sought from the relevant authorities.
Acknowledgements
The Working-for-Water Programme of the Department
of Water Affairs and Forestry funded this project, and
are gratefully acknowledged. Drs S. Neser and H.G.
Zimmerman collected the two rust isolates used.
References
Bruzzese, E. and Hasan, S. (1983) A whole leaf clearing and
staining technique for host specificity studies of rust fungi. Plant Pathology 32, 335338.
Cordo, H.A. and DeLoach, C.J. (1995) Natural enemies of
the rangeland weed whitebrush (Aloysia gratissima: Verbenaceae) in South America: potential for biological control in the United States. Biological Control 5, 218230.
Cummins, G.B. (1940) The genus Prospodium (Uredinales).
Lloydia 3, 178
Cummins, G.B. and Hiratsaka, Y. (2003) Illustrated genera
of rust fungi. APS Press, St Paul, pp. 113114.
Diniz, M.A. (1988) Bignoniaceae. Flora Zambezica 8,
6185.
Gentry, A.H. (1992) Bignoniaceae Part II (Tribe Tecomeae).
Flora Neotropica Monograph, vol. 25(II). New York
Botanic Garden, New York, pp. 273293.
Henderson, L. (2001) Alien weeds and invasive plants, a
complete guide to declared weeds and invaders in South
Africa. Plant Protection Research Institute Monograph
no. 12. Plant Protection Research Institute, Pretoria, South
Africa (300 p).
Nel, J.L., Richardson, D.M., Rouget, M., Mgidi, T.N.,
Mdzeke, N., Le maitre, D.C., van Wilgen, B.W.,
Schonegevel, L., Henderson, L. and Neser, S. (2004)
A proposed classification of invasive alien plant species
in South Africa: towards prioritizing species and areas
for management action. South African Journal of Science
100(January/February), 5364.
Olckers, T. (2004) Targeting emerging weeds for biological control in South Africa: the benefits of halting the
spread of alien plants at an early stage of their invasion.
South African Journal of Science 100 (January/February),
6468.
347
XII International Symposium on Biological Control of Weeds
Smithies, S.J. (2003) Bignoniaceae. In Plants of southern Africa: an annotated checklist. (eds Germishuizen, G. and
Meyer, N.L.). Strelitzia 14, 312313.
Shuttleworth, F.S. (1953) Studies on sub-tropical rusts. I.
Prospodium transformans. Mycologia 45, 437449.
Tomley, A.J. and Riding, N. (2002) Prospodium tuberculatum, lantana rust, a new agent released for the biocontrol of the woody shrub Lantana camara. In: Spafford
Jacob, H., Dodd J. and Moore. J. (eds) Proceedings of
the 13th Australian Weeds Conference. Plant Protec-
tion Society of Western Australia, Perth, Australia, pp.
389390.
Vitorino, M.D., Pedrosa-Macedo, J.H., Menezes, A.O. Jr., Andreazza, C.J. Bredow, E.A. and Simes, H.C. (2004) Survey of potential biological agents to control yellow bells,
Tecoma stans (L.) Kunth. (Bignoniaceae), in southern Brazil. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale,
W.M., Morin, L. and Scott, J.K.. (eds) Proceedings of the XI
International Symposium on Biological control of Weeds.
CSIRO Entomology, Canberra, Australia, pp. 186187.
348
Study on the herbicidal activity of vulculic
acid from Nimbya alternantherae
M.M. Xiang, L.L. Fan, Y.S. Zeng and Y.P. Zhou1
Summary
The detached leaves of 28 species of plants and the plants of alligator weed were tested for toxic
activity of vulculic acid isolated from Nimbya alternantherae. Results indicated that the toxin was
non-host-specific and could cause leaf blight and withering of alligator weed. The effect of the toxin
on the ultrastructure of leaf tissue of alligator weed was studied by treating the mature leaves with
the toxin. It was shown that the damage on leaf tissue included plasmolysis and vacuolation in cells,
lamellae disorder and vacuolation in chloroplasts as well as disappearance of ridges and vacuolation
in mitochondria after treatment with vulculic acid at a concentration of 50 mg/ml for 12 h. These results suggest that the target sites for vulculic acid action may be the plasma membranes, the lamellae
of chloroplast and the ridges of mitochondria of the alligator weed leaf.
Keywords: mycotoxin, alligator weed, ultrastructure, pathogenic mechanism.
Introduction
Alligator weed, Alternanthera philoxeroides (Martius)
Grisebach, is widely known as a serious exotic weed.
Some progress has been achieved in the biological
control of this weed with plant pathogenic fungi and
their metabolites. Pathogenic fungi reported on alligator weed include Rhizoctonia solani Khn (Singh and
Devi, 1991), Colletotrichum sp. (Tan and Gu, 1992),
Cercospora alternantherae Ellis & Langlois (Barreto
and Torres, 1999) and Nimbya alternantherae (Holcomb & Antonopoulus) Simmons & Alcom (Xiang
et al., 1998; Barreto and Torres, 1999). The latter two
species are considered to have a potential as biological
control agents for alligator weed (Barreto and Torres,
1999; Barreto et al., 2000; Xiang et al., 2002a). Moreover, Wan et al. (2001) found that the crude preparation
of the toxin from Alternaria alternata (Fr.) Keissler was
strongly toxic to the leaves of the alligator weed. The
metabolic product of Fusarium sp. could also induce leaf
lesions and withering of the weed (Tan et al., 2002).
N. alternantherae can cause purplish leaf spots and
defoliation (Barreto and Torres, 1999; Xiang et al.,
2002a) and is a highly host-specific pathogenic fungus
(Xiang et al., 2002a). Xiang et al. (2002b) found that
Zhongkai University of Agriculture and Technology, School of Agriculture and Landscape Architecture, Dongsha Street, Fangzhi Road,
Guangzhou, 510225, China
Corresponding author: M.M. Xiang <[email protected]>.
CAB International 2008
the filtrate of this fungus had herbicidal activity and
could inhibit radicle growth in sorghum, and Zhou et
al. (2006) isolated vulculic acid from the filtrate. To
provide a foundation for developing a biochemical herbicide, the herbicidal activity of this toxin was studied
and is reported in this paper.
Materials and methods
Materials
Toxin material: Vulculic acid was isolated from the
filtrate of N. alternantherae using the method of Zhou
et al. (2006), which was slightly modified. N. alternantherae was cultured in modified Fries (Xiang, 2005)
at 28C, 200 rpm for 7 days, and the culture liquid was
filtered. Methanol was added into the filtrate in the volume ratio of 1:3, stirred, then filtered and evaporated at
60C to get a mush. The crude product was extracted
with ethyl acetate from the mush, and dissolved with
benzene and acetone (in volume ratio of 1:1) at 60C
and filtered. After crystallizing and re-crystallizing in
an ice-bath two to three times with benzene and acetone in 1:1 ratio, crystals with a few impurities were
mixed with silica gel in a mass ratio of 1:1 and put into
a column of 300 15mm, eluting with benzene and acetone (1:1). The eluting liquid was evaporated at 60C
until 30 to 40 ml remained and then re-crystallized in
an ice-bath. The toxin obtained was dried in a vacuum
desiccator, and its purification was evaluated by highperformance liquid chromatography and infrared spectra (IR).
349
XII International Symposium on Biological Control of Weeds
Plant material: Alligator weed was grown in nutritional liquid that is commonly used for leafy vegetable
production. The other 27 species of plants tested for
toxic activity were all seedlings grown in a greenhouse
or taken from the field (see Table 1).
Table 1.
Sensitivity of the plants tested to vulculic acid.
Plant species
Red pepper (Capsicum
annuum L.)
Tomato (Lycopersicon
esculentum Mill)
Eggplant (Solanum
melongena L.)
Spinach (Spinacia oleracea L.)
Water cress (Nasturtium
officinale R.Br.)
Radish (Raphanus sativus L.)
Gynura (Gynura bicolor L.)
Yerbadetajo (Eclipta
prostrate L.)
Red tasselfpower [Emilia
sonchifolia (L.) DC.]
Crowndaisy oxeyedaisy
(Chrysanthemum
coronarium L.)
Lettuce (Lactuca sativa var.
crispa L.)
China crabdaisy [Wedelia
trilobata (Osb.) Merr.]
Japanese youngia [Youngia
japonica (L.) DC.]
Celery (Apium graveolens L.)
Coriander (Coriandrum
sativum L.)
Carrot (Daucus carota L. var.
sativus DC.)
Pea (Pisum sativum L.)
Sour dallisgrass (Paspalum
conjugatum Berg.)
Corn (Zea mays L.)
Sweet potato (Ipomoea
batatas Lam.)
Dashen [Colocasia esculenta
(L.) Schott]
Cucumber (Cucumis sativus L.)
Nutgrass cypressgrass (Cyperus
rutundus L.)
Red woodsorrel (Oxalis
corymbosa DC.)
Asia plantain (Plantago
asiatica L.)
Alligator weed [A. philoxeroides
(Martius) Grisebach]
Thorny amaranth (Amaranthus
spinosus L.)
Concentration of the toxin
(mg/ml)
Methods
Effective range of the toxin: The purified toxin was
diluted to give concentrations of 50, 30, 10 and 5 mg/ml
with double-distilled water. The healthy leaves of the
plants were washed with tap water and then three times
with double-distilled water, dried on sterile filter paper, cut into 0.5 0.5 cm pieces and then were placed
into test tubes with 2 ml of the toxin solution. Ten plant
pieces were added to each tube, and then was a duplicate for each concentration. The tubes with solution
and pieces were put into an illuminated incubator at
25C for 24 h, and pathological changes were recorded.
Effect of the toxin on alligator weed plant: The purified toxin was diluted to concentrations of 300, 200,
100 and 50 mg/ml with distilled water. Alligator weed
stems that had just begun to develop roots were dug up
in the field, washed in tap water and grown in nutrient
solution for 3 to 4 days. Then, the plants were washed
three times with distilled water, dried on sterile filter
paper and transplanted into tubes containing 10 ml of
the different toxin dilutions. Three plants were included
in each tube, and tubes were duplicated for each concentration. Distilled water was used in control tubes.
The treated plants were grown at room temperature for
24 h, and then pathological changes were recorded.
Effect of the toxin on the ultrastructure of alligator
weed leaf: The purified toxin was diluted to a concentration of 50 mg/ml with double-distilled water. Healthy
mature leaves of alligator weed were washed with tap
water and then three times with double-distilled water, dried on sterile filter paper and cut into pieces of 2
to 3 mm transversely. The pieces were placed into test
tubes with 2 ml of the toxin solution and decompressed
for 20 min. Then, the tubes with solution and pieces
were put into an illuminated incubator at 25C for 12 h.
Double-distilled water was used in the control tube.
The samples were fixed with 2.5% glutaraldehyde,
then with 1% osmic acid. After being dehydrated using a standard method, the samples were embedded
in Epon 812. Microtome sections were dyed with uranium acetate, then lead citrate and observed through
FEI-Tecnai 12 transmission electron microscope.
50
30
10
CK
+
-
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
Effect of the toxin on alligator weed plant
The toxin inhibited root growth and induced wilt of
alligator weed plants after treatment for 24 h under all
of the concentrations tested (Fig. 1A, B).
Note: + means sensitive, - means insensitive
Results
Effective range of the toxin
Vulculic acid was a non-host-specific toxin. It had
toxic activity towards many plants from different families or genera, including alligator weed, after treatment
for 24 h under almost all of the concentrations tested
(Table 1). It caused brown blight on the leaf pieces.
350
Study on the herbicidal activity of vulculic acid from Nimbya alternantherae
Figure 1.
A Roots of alligator weed treated with the toxin for 24 h. B Plants of alligator weed treated with the toxin for 24 h.
Figure 2.
A Normal ultrastructure in the alligator weed control leaf. B Plasmolysis and vacuolated cells in the treated leaf with the toxin. C Normal lamellae of chloroplast in
the control leaf. D Normal structures of mitochondria in the control leaf. E Disordered lamellae and vacuolated chloroplasts in the treated leaf with the toxin.
F Vacuolated mitochondria and disordered lamellae of chloroplast in the treated
leaf with the toxin. CH Chloroplast, M mitochondria, N nuclei, CW cell wall.
351
XII International Symposium on Biological Control of Weeds
Effect of the toxin on ultrastructure of
alligator weed leaf
In tissue samples after treatment with vulculic acid,
at a concentration of 50mg/ml for 12h, the plasmalemma of leaf cells was invaginated and detached from
the cell wall in almost all places, and the cells became
vacuolated. However, the plasmalemma did not appear ruptured (see Fig. 2B). Chloroplasts were severely
damaged with disorder lamellae and a lot of vacuoles
in the treated tissue with the toxin (Fig. 2E). Mitochondria from leaf cells treated with the toxin showed significant disorganization, such as the decrease of ridges
and the vacuolation of mitochondria (Fig. 2F).
Discussion
The toxin, vulculic acid, isolated from N. alternantherae, was reported to inhibit the pollen germination
of black pine, Pinus thunbergii Parl., by up to 85.3% at
a concentration of 10 mg/l (Kimura et al., 1991). Before our study, its toxicity to other plants had not been
reported. The preliminary screening results showed
that vulculic acid is a non-host-specific toxin and could
inhibit root growth and induce wilt of alligator weed.
Thus, the advantage of vulculic acid as an herbicide
compared to N. alternantherae lies in its wider host
range and better prospect for product development.
In this study, vulculic acid was toxic to the plasmalemma, the lamellae of chloroplast and the ridges of
mitochondria of alligator weed leaf cells after treatment
at a concentration of 50 mg/ml for 12 h. These results
suggest that the target sites for the toxin action may be
on the plasma membranes, the lamellae of chloroplast
and the ridges of mitochondria of alligator weed leaf.
However, this is the first and preliminary study on the
ultrastructural effect of vulculic acid on plant tissues.
Further studies are required to determine which one of
the three target sites is damaged first and the minimum
concentration of toxin needed to cause the damage.
Acknowledgements
This study was supported by the National Natural Science Foundation of China, the National Key Technolo-
gies R&D Programme, the Natural Science Foundation
of Guangdong Province and the Key Technologies
R&D Programme of Guangdong.
References
Barreto, R.W. and Torres A.N.L. (1999) Nimbya alternantherae and Cercospora alternantherae: two new records
of fungal pathogens on Alternanthera philoxeroides (alligatorweed) in Brazil. Australasian Plant Pathology 28,
103107.
Barreto, R., Charudattan A., Pomella A. and Hanada R.
(2000) Biological control of neotropical aquatic weeds
with fungi. Crop Protection 19, 697703.
Kimura, Y., Nishibe M., Nakajima H. and Hamasaki, T.
(1991) Vulculic acid, a pollen germination inhibitor produced by the fungus, Penicillium sp. Agricultural Biological Chemistry 55, 11371138.
Singh, N.I. and Devi, R.K.T. (1991) New host records of fungi from India. Indian Phytopathology 43, 594595.
Tan, W.Z. and Gu, C.Y. (1992) Studies on the biological characteristic and the increase and decline of Colletotrichum
sp. from alligatorweed. Journal of Yunnan Agricultural
University 7, 249-251.
Tan, W.Z., Li, Q.J. and Qing, L. (2002) Biological control
of alligatorweed (Alternanthera philoxeroides) with a Fusarium sp. BioControl 47, 463479.
Wan, Z.X., Qiang, S., Xu, S.B., et al. (2001) Culture conditions for production of phytotoxin by Alternaria alternata
and plant range of toxicity. Chinese Journal of Biological
Control 17, 1015.
Xiang, M.M. (2005) Study on the herbicidal activity of Nimbya alternantherae and its toxin. Dissertation. South China Agricultural University, Guangzhou, China.
Xiang, M.M., Liu, R. and Zeng, Y.S. (1998) Nimbya alternantherae a new record of the genus Nimbya from China. Mycosystema 17, 283, 288.
Xiang, M.M., Liu, R. and Zeng, Y.S. (2002a) Herbicidal activity of metabolite produced by Nimbya alternantherae, a leaf
spot pathogen of Alternanthera philoxeroide. Chinese
Journal of Biological Control 18, 8789.
Xiang, M.M., Zeng, Y.S. and Liu, R. (2002b) Host range, condition for conidium-producing and efficacy in alligatorweed
control of Nimbya alternantherae. Acta Phytopathologica
Sinica 32, 285287.
Zhou, Y.P., Xiang, M.M., Jiang, Z.D., Li, H.P., Sun, W., Lin,
H.L. and Fan, H.Z. (2006) Separation, purification and
structural identification of the phytotoxin from Nimbya
alternantherae. Chemical Journal of Chinese Universities 27, 14851487.
352
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Optimization of water activity and placement of
Pesta-Pseudomonas fluorescens BRG100
biocontrol of green foxtail
S.M. Boyetchko, R.K. Hynes, K. Sawchyn, D. Hupka and J. Geissler
Agriculture & Agri-Food Canada, 107 Science Place, Saskatoon, SK, Canada S7N 0X2
Pseudomonas fluorescens BRG100 was selected from earlier screening studies for pre-emergent bioherbicidal activity to green foxtail and wild oat. A granular formulation, Pesta, has been developed to
deliver P. fluorescens BRG100. Delivery and placement of sufficient numbers of BRG100 to inhibit or
suppress germination of the weed is one of the key challenges in bioherbicide product development.
However, optimization of BRG100 survival, placement and dispersion from the Pesta granule in the
target zone has not been fully established. Increased shelf-life of BRG100 in Pesta may be acquired
by increasing BRG100 cell membrane integrity, optimizing the water activity of the granules (aw), a
useful measure of the free (unbound) water that is available for use by microorganisms. Addition of
maltose, 3% w/w, reduced survival of BRG100 in peat culture and in Pesta granules prepared from peat
powder cultures as compared to peat powder culture and resulting Pesta without maltose. Survival of
BRG 100 in Pesta was greatest with the water activity (aw) adjusted to 0.2 as compared to 0.5 and 0.8
aw. Placement of Pesta in-row and side-banded with green foxtail was examined in a greenhouse study.
Evidence of phytotoxin damage to green foxtail by Pseudophomins A and B was observed.
Impact of natural enemies on the potential damage of
Hydrellia sp. (Diptera: Ephydridae) on Egeria densa
G. Cabrera Walsh,1 F. Mattioli1 and L.W.J. Anderson2
USDAARSSouth American Biological Control Laboratory, Bolivar 1559, B1686EFA,
Hurlingham, Buenos Aires, Argentina
2
USDAARSExotic and Invasive Weeds Research Unit, Department of Plant Sciences, UC Davis,
Mail Stop 4, One Shields Avenue, Davis, CA 95616-0000 USA
1
Egeria densa Planchon (Brazilian Elodea or Brazilian waterweed) is a South American submerged
perennial in the Hydrocharitaceae that has become a weed in North America, Australia, New Zealand,
South Africa and parts of Asia and Europe. It crowds out other plant species by forming dense stands,
negatively affecting the native biota, as well as water sports, fishing, navigation, delivery of irrigation
water and hydropower production. The larva of Hydrellia sp. (Diptera: Ephydridae) from Argentina
feeds on the mesophyl, producing chlorosis (bleaching) of two to three whorls per larva, and mining
the stem in between them. Under laboratory conditions, a single gravid female can cause the defoliation of whole stems. In the field, this insect has several natural enemies that attack the larvae and the
pupae. We discuss its potential impact on the weed under an enemy release situation, considering
Hydrellia has both specific and generalist natural enemies.
CAB International 2008
353
XII International Symposium on Biological Control of Weeds
Towards to study of the sunflower broomrape fungi
disease in Georgia
C. Chkhubianishvili, I. Malania, E. Tabatadze and L. Tsivilashvili
Kanchaveli L. Institute of Plant Protection, 82, Chavchavadze Avenue, 0162 Tbilisi, Georgia
The parasite plant, sunflower broomrape, Orobance cumana, is a major pest and is widely spread in
sunflower production regions of East Georgia. Experiments were conducted on influence of the introduced fungus, Fusarium oxysporum f. sp. orthoceras (FOO) on O. cumana and its host sunflower
strain Donskoy-60. These investigations indicate that FOO is a potent biological agent to control
against O. cumana in conditions experienced in Georgia, and it is an important component in integrated
pest management for sunflower management.
Biological control of Imperata cylindrica in
West Africa using fungal pathogens
A. Den Breeyen,1 R. Charudattan,1 F. Beed,2 G.E. MacDonald,3
J.A. Rollins1 and F. Altpeter3
University of Florida, Plant Pathology Department, Gainesville, FL, USA
2
International Institute of Tropical Agriculture (IITA), Cotonou, Benin
3
University of Florida, Agronomy Department, Gainesville, FL, USA
Imperata cylindrica (cogon grass), a noxious, rhizomatous grass with a pan-tropical distribution represents one of the most serious constraints to crop production and poverty alleviation in West Africa. The
fungi, Bipolaris sacchari and Drechslera gigantea, have shown potential as bioherbicides to control
cogon grass (var. major) in the southeastern USA. Biological control may however prove to be ineffective if the West African cogon grass (var. africana) is genetically heterogeneous from the southeastern
USA cogon grass. The objectives of this study are to assess the genetic diversity between the West
African and southeastern USA cogon grass populations and to determine the virulence of the southeastern USA and West African isolates of B. sacchari and D. gigantea on the West African cogon grass
population. A further objective is to determine the potential of three biotrophs: two rust fungi, Puccinia
imperatae and Puccinia fragosoana, and a head smut, Sporisorium schweinfurthiana, associated with
cogon grass in South Africa, where cogon grass is not a weed, for control of cogon grass. Interim results indicate that there are no differences between the USA and West African fungal isolates in terms
of their virulence on the var. africana. The genetic variation results and the implications for fungal
biocontrol on cogon grass in West Africa will be discussed.
354
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Impact of Ischnodemus variegatus (Hemiptera: Blissidae)
on the invasive grass Hymenachne amplexicaulis in Florida
R. Diaz,1 W.A. Overholt,1 J.P. Cuda2 and P.D. Pratt3
Biological Control Research and Containment Laboratory, Fort Pierce, FL 34945, USA
University of Florida, Department of Entomology and Nematology, Gainesville, FL 32611, USA
3
USDA, Invasive Plant Research Laboratory, Fort Lauderdale, FL 32608, USA
1
Invasions of exotic grasses constitute a major threat to aquatic ecosystems. West Indian Marsh Grass,
Hymenachne amplexicaulis (Rudge) Nees, which is native to South America, is considered a major
environmental weed in southeastern USA and Australia. In Florida, an adventive insect was recently
found causing severe damage to H. amplexicaulis. This insect was identified as Ischnodemus variegatus (Hemiptera: Blissidae) and is considered native to South America. The host range of this herbivore
and its potential to control H. amplexicaulis were evaluated under laboratory, greenhouse and field conditions. We tested 60 plants under no-choice conditions for development and five plants for oviposition
of the insect. I. variegatus had higher survival from nymph to adult on H. amplexicaulis than on other
tested plants. Development to the adult stage also occurred on Panicum hemitomon, Panicum anceps,
Paspalum urvellei (all Poaceae) and Thalia geniculata (Marantaceae). Oviposition choice tests demonstrated that I. variegatus females will lay eggs on several non-target grasses. Greenhouse experiments
demonstrated that feeding damage of I. variegatus reduces the growth rate, chlorophyll levels and
biomass of H. amplexicaulis seedlings. Field sampling of naturally occurring populations in Florida
indicated that I. variegatus density, under favourable climatic conditions, increase during the summer
and can experience outbreaks that severely reduce H. amplexicaulis survival and reproduction.
Ecological basis for biological control of
Arundo donax in California
T.L. Dudley,1 A. Lambert1 and A. Kirk2
1
University of California, Marine Science Institute, Santa Barbara, CA, USA
2
USDAARS European Biological Control Lab, Montpellier, France
Arundo donax invades California riparian areas and is a target for biological control. Candidate agents
have been identified, but their eventual release will depend upon evidence that damage is substantial
and novel. As part of a program comparing Arundo growth traits and damage in California and the
Mediterranean region (its presumed origin), we documented the presence in southern California of
Tetramesa romana (Walker) (Hymenoptera: Eurytomidae), the same stem-boring sawfly being tested
in quarantine laboratories for future introduction. Primary or secondary shoots <10 cm diameter are
occupied, with densities up to 34 larvae per 100 cm of culm, and mortality of secondary shoots is common. The wasp has been shown to infect new hosts under experimental field conditions, so we have
an opportunity to test its efficacy and host range without the artefacts that plague standard quarantine
testing. Field studies continue on both continents to document life cycles and impacts and determine
whether this insect can utilize alternative hosts such as Phragmites australis and other native or economic grasses. If safety can be shown, this wasp may be amenable to re-distribution to other infested
ecosystems in the western USA.
355
XII International Symposium on Biological Control of Weeds
Biology and host specificity of Puccinia arechavaletae,
a potential agent for the biocontrol of
Cardiospermum grandiflorum
A. Fourie and A.R. Wood
ARC-PPRI, Weeds Research Division, Private Bag X5017, Stellenbosch 7599, South Africa
Cardiospermum grandiflorum (balloon-vine), a subtropical climber from South America, is an emerging weed in some regions in Southern Africa. It is a highly invasive, vigorous climber that invades
mostly forest margins and watercourses in sub-tropical areas such as KwaZulu-Natal and the Kruger
National Park in Mpumalanga, South Africa. Robust stems with tendrils enable C. grandiflorum to
climb and form a dense canopy, which completely smothers the underlying indigenous vegetation.
The fruit capsules are carried by wind and float freely on water, dispersing the plant along waterways,
and the hard seeds result in extended seed longevity in the soil. It is classified as a category 1 weed in
South Africa, meaning that it is prohibited and must be controlled or eradicated where possible, and it
was consequently targeted for biological control. An isolate of the rust fungus, Puccinia arechavaletae,
was collected in South America, from a biotype of C. grandiflorum that matches the morphology of the
South African weed. This isolate was successfully established on C. grandiflorum in quarantine laboratories at the ARC-PPRI in South Africa. Host-range testing showed that this isolate of P. arechavaletae
is specific to C. grandiflorum, making it a promising agent for the biocontrol of this environmental
weed.
Potential for host-specific biological control agents at
population/subspecies level?
P. Hfliger1 and B. Blossey2
1
CABI Switzerland Centre, Rue des Grillons 1, 2800 Delmont, Switzerland
Cornell University, Ecology and Management of Invasive Plants Program, Department of Natural Resources,
Fernow Hall, Ithaca, NY 14853, USA
Host-specificity testing and post-release evaluations of biological control agents show that closely
related plant species (same genus) are most likely to be at risk of non-target attack. Most biocontrol
programs aim for species-specific control agents, yet in the program targeting invasive Phragmites
australis in North America, the existence of a native subspecies (Phragmites australis americanus)
requires specificity at the subspecies level. Is it realistic to expect to find herbivores that distinguish
between native and introduced P. australis? Field surveys in North America found native specialist
herbivores attacking only native P. australis, suggesting distinct differences between native and introduced genotypes that are recognizable by herbivores. Preliminary multiple-choice oviposition tests and
no-choice larval development tests with several herbivore species considered as potential biological
control agents in Europe showed the ability of these herbivores to complete development on native
and introduced P. australis genotypes. However, we also found a preference of these herbivores for
introduced P. australis. These preferences together with differences in plant morphology may allow a
biocontrol program to proceed even if no subspecies-specific control agents are available.
356
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Combined effects of herbicides and rust fungi on
Rumex obtusifolius
P.E. Hatcher and F.J. Palomares-Rius
The University of Reading, School of Biological Sciences, Whiteknights, Reading, RG6 6AS, UK
The rust fungus, Uromyces rumicis, has been investigated as a potential biocontrol agent for the perennial grassland weed Rumex obtusifolius in Europe. Although the rust infection reduces plant growth, it
does not kill the plant, cannot infect the young leaves and causes only a moderate reduction in seed production. Thus, it is unlikely to be a successful biocontrol agent on its own. In this study, we investigated
the possibility that low doses of herbicides might act as synergists to the rust. Out of several herbicides
tested, asulam, thifensulfuron-methyl and dicamba increased Uromyces spore germination when applied at 1/300th recommended concentration, and the first two also had a positive effect at 1/150th
concentration. When applied on R. obtusifolius at these concentrations, both the rust and herbicides had
significant effects alone. Together they often had an additive, but not a synergistic effect.
Host-specificity and potential of Kokujewia ectrapela
Konow for the control of Rumex spp.
Y. Karimpour
Urmia University, Faculty of Agriculture, Department of Plant Protection, Urmia, Iran
The host specificity and food consumption of Kokujewia ectrapela Konow (Hym., Argidae) were studied to evaluate the potential of this sawfly as a non-endogenous biological control agent of Rumex
spp. in Australia. This oligophagous, multivoltine sawfly is an indigenous species on Rumex spp. (Polygonaceae) in Russia, Transcaucasia and Iran. Results of no-choice feeding tests with second instars
on 27 plant species belonging to 13 families showed that K. ectrapela completed its life cycle mainly
on plants of Rumex and occasionally fed on Polygonum persicaria L. Under laboratory conditions,
weights of consumed food by larvae were measured at 25C. Feeding activity of three larval instars of
K. ectrapela on Rumex obtusifolius L. were investigated. Weight of consumed leaves differed between
instars. During the 3 days of first instar development, one larva consumed 0.041 0.001 g of R. obtusifolius leaves. The next two instars, with durations of 4 and 5 days, consumed significantly larger
amount of leaves, viz., 1.227 0.006 and 3.058 0.014 g, respectively. Total weight of consumed
leaves by all three instars of a single larva, during 12 days of the developmental time, amounted to
4.310 0.01 g.
357
XII International Symposium on Biological Control of Weeds
Growth and phenology of three Lythraceae species in
relation to feeding by the leaf beetles, Galerucella spp.
E.J.S. Katovich,1 R.L. Becker,1 L.C. Skinner2 and D.W. Ragsdale3
University of Minnesota, Department of Agronomy and Plant Genetics, St. Paul, MN, USA
2
Minnesota Department of Natural Resources, St. Paul, MN, USA
3
Department of Entomology, University of Minnesota, St. Paul, MN, USA
Previous studies have characterized the development of Galerucella calmariensis and Galerucella
pusilla on Lythrum salicaria and on the non-target Lythraceae species, Lythrum alatum and Decodon
verticillatus. The impact of Galerucella on these species, when grown in outdoor mesocosms that more
closely mimics ecological host range, has not been reported. The first objective of this study was to
compare the growth and seed capsule production of L. salicaria, L. alatum and D. verticillatus, with
and without Galerucella. With L. salicaria, larval feeding on apical and lateral shoot buds resulted in
fewer seed capsules compared to control plants. No measured plant growth or reproductive parameters
were reduced as a result of Galerucella feeding on D. verticillatus. Presence of Galerucella on L.
alatum resulted in a reduction of seed capsules in 1 year of a 2-year study. A second objective of our
study was to compare the phenology of the three Lythraceae species in relation to that of Galerucella.
In the northern USA, flowering and seed development in D. verticillatus occurred a month later than
in L. salicaria or L. alatum. The delayed flowering of D. verticillatus resulted in avoidance of shoot
meristem feeding damage caused by the first generation of Galerucella.
Corynespora cassiicola f. sp. benghalensis, a new natural
enemy of Commelina benghalensis: infection parameters
D.C. Lustosa and R.W. Barreto
Universidade Federal de Viosa, Departamento de Fitopatologia, CEP 36571-000, Viosa, MG, Brazil
Commelina benghalensis (wandering Jew) is a herbaceous plant from Asia, which became one of
the worst crop invaders after its recent arrival in Brazil, particularly in soybean and coffee. A survey
aimed at discovering fungal pathogens that might already be established in Brazil revealed a limited
mycobiota. Among the fungi that were collected was the eye-spot fungus Corynespora cassiicola.
This is usually regarded as a polyphagous pathogen, but a host-range evaluation has shown that the
strain that attacks C. benghalensis is a highly specific forma specialis and is sufficiently damaging to
allow further considerations on its use as a mycoherbicide. Growth, sporulation and spore germination were evaluated under different conditions. The fungus was insensitive to light regimes. Optimum
sporulation was between 2530C, whereas optimal conidial germination was between 2025C. The
effect on disease severity of a combination of different dew periods and temperatures was tested, and
the level of damage was proportional to the duration of the dew period, independent of temperature.
Delays to the onset of dew period after inoculum application were also tested, and levels of disease
severity were still high with delays of up to 24 h.
358
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Potential use of Trichilogaster acaciaelongifoliae as a
biocontrol agent of Acacia longifolia in Portugal
H. Marchante,1 H. Freitas2 and J. Hoffmann3
Escola Superior Agrria de Coimbra, 3040-216, Coimbra, Portugal
University of Coimbra, Department of Botany, Calada Martim de Freitas, 3001-455 Coimbra, Portugal
3
University of Cape Town, Department of Zoology, Rondebosch 7700, South Africa
1
Acacia longifolia (long-leafed wattle) was introduced into Portugal more than 150 years ago to bind
coastal dunes. Its distribution has increased greatly after fire events, and it is now one of the worst invasive plant species along the Portuguese coast. Trichilogaster acaciaelongifoliae (an Australian gall
wasp) is being tested as a potential biological control agent of A. longifolia. If released, this would be
one of the first planned introductions of a classical biological control agent against an environmental
weed species in Europe. T. acaciaelongifoliae is a monospecific bud-galling pteromalid wasp that
prevents its host plant from flowering normally and deforms vegetative growth. Seed production by A.
longifolia has been reduced by more than 90% since the introduction of the wasp into South Africa in
1987. In specificity tests, females are confined on a set of test-plant species. Flower and vegetative buds
are then dissected to detect eggs. Species on which eggs are found will then be the subject to larval
development tests. The results so far have been promising. At the same time, climate studies are being
undertaken to determine whether any regions of Portugal are unsuitable for T. acaciaelongifoliae and
how best to move the wasps from the southern to the northern hemisphere.
Diclidophlebia smithi (Hemiptera, Psylloidea):
a potential biocontrol agent for Miconia calvescens
E.G.F. Morais,1 M.C. Picano,1 R.W. Barreto,2 G. Silva,1 M.R. Campos1 and R.B. Queiroz1
Universidade Federal de Viosa, Departamento de Biologia Animal, Viosa, MG 36570-000, Brazil
2
Universidade Federal de Viosa, Departamento de Fitopatologia, Viosa, MG 36570-000, Brazil
Miconia calvescens (Melastomataceae) is a shrub or small tree native from the Neotropics that became
an aggressive invader of natural ecosystems of the Pacific Islands after its introduction as an ornamental. Intensive searches for insects and pathogens to be used as biocontrol agents were conducted in its
native range. Among the insects collected in Brazil, the newly described species was Diclidophlebia
smithi (Hemiptera: Psyllidae),, which is often found attacking terminal buds, inflorescences and fruits
of M. calvescens. It appeared to be causing significant impact on its host in the field. Population density
of D. smithi was recorded for 2 years at three different localities in the state of Minas Gerais (Brazil). It
occurs throughout the year, but its population is more abundant between April and October (the dry and
cool autumnwinter period). No parastioid was found attacking D. smithi, and the sole significant natural enemy of this psyllid was a predatory Syrphidae larva. Diclidophlebia sp. n. has five nymph instars
during its life cycle, which takes between 40 and 67 days. Preliminary specificity tests have indicated
that D. smithi is highly host-specific, attacking only M. calvescens. Impact studies have shown that D.
smithi affects significantly shoot development and flower and fruit bearing in M. calvescens.
359
XII International Symposium on Biological Control of Weeds
Supplementary host-specificity testing of
Puccinia melampodii, a biocontrol agent of
Parthenium hysterophorus
K. Ntushelo and A.R. Wood
ARC-PPRI, Weeds Division, Private Bag X5017, Stellenbosch 7599, South Africa
Parthenium hysterophorus, native to South and Central America, is an invasive weed in KwaZuluNatal, Mpumalanga and the northwest provinces of South Africa. The micro-cyclic rust fungus, Puccinia melampodii, has been successfully used in Australia and is being considered for release in South
Africa. As this rust was subjected to comprehensive host-specificity testing in Australia, only supplemental testing was necessary. Testing to five selected South African sunflower cultivars and three out
of eight indigenous Heliantheae (Asteraceae) species was undertaken. Plants were inoculated with basidiospores of P. melampodii and incubated at 25C for 24 h. For each sunflower cultivar, three plants
were tested in four replications, two replications for the indigenous plants, and three P. hysterophorus
plants were included as control in every replication. Rust symptom development was monitored, and
no symptoms developed on the sunflower plants and the indigenous Heliantheae, except for a few
pustules that developed on two Spilanthes mauritiana plants in one replication. All Parthenium control
plants were heavily infected. The conclusion was therefore made that the tested plant species and the
sunflower cultivars are unlikely to be infected by P. melampodii in the event that this agent is released
to control Parthenium in South Africa.
Is Prosopis meeting its match in Baringo?
W.O. Ogutu,1,2 H. Mueller-Schaerer,1 U. Schaffner,3 P.J. Edwards4 and R. Day2
Universit de Fribourg/Prolles, Dpartement de Biologie/Ecologie & Evolution, Chemin du Muse 10,
1700 Fribourg, Switzerland
2
CABI Africa, ICRAF Complex, United Nations Avenue, Gigiri, P. O. Box 633-00621, Nairobi, Kenya
3
CABI EuropeSwitzerland, Rue des Grillons 1, 2800 Delmont, Switzerland
4
Geobotanical Institute, Swiss Federal Institute of Technology Zuerichbergstrasse 38 8044 Zurich, Switzerland
1
The genus Prosopis is native to arid and semi-arid regions of Asia, Africa and America. Neotropical
species, such as Prosopis juliflora and Prosopis pallida, have been introduced worldwide for multipurpose use and their ability to survive poor conditions. Prosopis introductions into Kenya occurred
mainly in 1980s, and it has since spread to neighbouring areas threatening the livelihoods of humans
and ecosystems. In response, Food and Agriculture Organization supports a project to manage Prosopis. One objective is to introduce, test and release the Prosopis seed-feeding bruchid Algarobius
prosopis from South Africa. The bruchid is undergoing specificity tests in quarantine. This beetle was
imported on the assumption that Prosopis is spreading because it is outside its natural range and lacks
natural enemies to regulate its population. In an effort to understand the ecology of Prosopis, we assessed the biodiversity (arthropods and microsorganisms) associated with Prosopis at Baringo, Kenya.
There an indigenous insect fauna is associated with Prosopis. Some of these insects cause significant
damage to the trees reducing reproductive potential and timber value. Can these insects be incorporated
in a management strategy for controlling Prosopis? Studies to determine the relationship between the
insects, Prosopis and the indigenous flora and to clarify the status and genetic variation of the Prosopis
species and their assumed hybrids are underway.
360
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
A lace bug as biological control agent of
yellow starthistle, Centaurea solstitialis L. (Asteraceae):
an unusual choice
A. Paolini,1 C. Tronci,1 F. Lecce,1 R. Hayat,2 F. Di Cristina,1 M. Cristofaro3 and L. Smith4
Biotechnology and Biological Control Agency, Via del Bosco 10, 00060 Sacrofano, Rome, Italy
Atatrk University, Faculty of Agriculture, Plant Protection Department, 25240 TR Erzurum, Turkey
3
ENEA C.R. Casaccia, s.p. 25, Via Anguillarese 301, 00060 S. Maria di Galeria, Rome, Italy
4
USDAARS, 800 Buchanan Street, Albany, CA 94710, USA
1
Tingis grisea is a univoltine sap-feeding lace bug distributed throughout Central and Southern Europe
and the Middle East reportedly associated with the genus Centaurea. During 2002, high T. grisea infestation levels were recorded on one yellow starthistle population in Eastern Turkey. Field observations
showed that significant damage was caused to the host plant especially when individuals were feeding
on the same plant in large numbers. Life cycle and biology observations allowed assessing the duration
of the five nymphal stages of T. grisea in laboratory conditions as well as female fecundity and longevity. Starvation and oviposition no-choice tests were carried out to determine the host-specificity level
of the insect. Results showed a clear oligophagous behaviour closely restricted to the genus Centaurea.
In addition, among the three species on which full larval development was ascertained (Centaurea solstitialis, Centaurea sulphurea and Centaurea cyanus), yellow starthistle was clearly preferred in terms
of number of eggs laid and number of adults obtained.
Potential biological control of Lantana camara in the
Galapagos using the rust Puccinia lantanae
J.L. Rentera and C. Ellison
Charles Darwin Research Station, Introduced Species Program, Puerto Ayora, Isla Santa Cruz,
Galapagos, Ecuador
2
Invasive Species Management, Introduced Plants Program, CABI Bioscience, Silwood Park, Ascot,
Berkshire SL5 7TA, UK
Laboratory experiments were carried out in England to test the specificity and environmental requirements of a Peruvian isolate of the fungus, Puccinia lantanae Farlow, known to attack the invasive plant
Lantana camara L., a serious problem in Galapagos. Eight species of plants representing five families
were inoculated with the fungus and kept in a dew chamber for 48 h. Lantana peduncularis Andersson and L. camara were sourced from Galapagos; other species related to Lantana were sourced from
other places. Dew periods of 5, 8, 11, 14 and 20 h were tested to determine the period necessary for
basidiospore formation and host infection. Only L. camara from Galapagos and Peru developed visible
symptoms 6 days after inoculation, and after 15 days, sori were fully developed. No non-target species
developed macroscopic symptoms. Most importantly, the rust did not attack the closest host relative
from Galapagos, the endemic L. peduncularis. Eight hours in the dew chamber was enough to induce
basidiospore formation and host infection, but times up to 20 h induced progressively more sori. Although we have not completed yet the experiments to determine the host-range specificity, P. lantanae
shows a promise as a biocontrol agent for L. camara in Galapagos.
361
XII International Symposium on Biological Control of Weeds
Biology and host specificity of Puccinia conoclinii
for biocontrol of Campuloclinium macrocephalum
in South Africa
E. Retief and A.R. Wood
ARC-PPRI, Weeds Research Division, Private Bag X5017, Stellenbosch 7599, South Africa
Campuloclinium macrocephalum (pompom weed) is a perennial herb, which was presumed to be
brought into South Africa for ornamental purposes. Initially, it could only be found in disturbed sites
such as roadsides, but eventually, it spread to wetlands, open savanna and grasslands and is very prominent in the Gauteng Highveld. During summer, numerous annual shoots are produced. During winter, it
survives as a rootstock in a dormant state underground. Due to prolific seed production and its survival
abilities, this plant is multiplying and spreading rapidly. The ideal control method would be to use a
biocontrol agent to damage the aerial parts of the plant. This will weaken the plant, reduce seed production and will deplete the nutrients stored in the roots. A rust fungus, which was tentatively identified as
Puccinia conoclinii (only urediniospores observed), was collected from C. macrocephalum in Northern Argentina and introduced into the quarantine laboratories at the ARC-PPRI, Stellenbosch. Hostspecificity testing revealed C. macrocephalum to be the only host of this rust isolate, and these results
were confirmed by microscopic examination on the host and other closely related plant species.
Status of tree of heaven, Ailanthus altissima, in Virginia,
USA and quarantine evaluation of Eucryptorrhynchus
brandti (Harold) (Coleoptera: Curculionidae),
a potential biological control agent
S.M. Salom, L.T. Kok, S. Yan, N. Herrick and T.J. McAvoy
Virginia Tech, Department of Entomology, Blacksburg, VA 24061-0319, USA
Tree of heaven (TOH), Ailanthus altissima (Mill.) Swingle, is an imported invasive weed tree from
China that has become established throughout much of continental USA. It colonizes disturbed forest
sites and often out-competes native vegetation. Short-term cultural and chemical controls of this weed
are expensive and have limited efficacy. Two curculionid species, Eucryptorrhynchus brandti (Harold)
and E. chinensis (Olivier), are primary mortality agents of A. altissima in China and have no other
known hosts. The objectives of our project are (1) to assess the pest status of A. altissima in Virginia
and (2) to evaluate E. brandti, the more numerous of the two species, for its potential as a biological
control agent. A statewide survey showed significant presence of TOH but no native herbivores with
potential of controlling it. Economic analysis of mechanical and chemical control indicates biological
control to be an attractive alternative. E. brandti requires live trees for development. Therefore, quarantine studies have focussed on developing a rearing technique and testing host specificity on native
plants approved by the Technical Advisory Group for Biological Control Agents of Weeds. Results
indicate that E. brandti feeds only on TOH, with greatly reduced feeding observed on corkwood, Leitneria floridana Chapman, and paradise tree, Simarouba glauca DC.
362
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Host use by the biological control agent
Longitarsus jacobaeae among closely related plant species?
U. Schaffner,1 P. Pelser2 and K. Vrieling3
CABI Switzerland Centre, Rue des Grillons 1, 2800 Delmont, Switzerland
2
Miami University, Department of Botany, Oxford, OH 45056, USA
3
Leiden University, Institute of Biology, Plant Ecology Section, 2300 RA Leiden, The Netherlands
1
The selection of representative test plant species for host-specificity testing is an important first step in
pre-release studies of classical biological control projects. This may be a challenging task, particularly
in projects where the target weed belongs to a species-rich genus. We present results from the hostrange testing of a cold-adapted Swiss biotype of the flea beetle, Longitarsus jacobaeae L. (Chrysomelidae), a candidate for the biological control of tansy ragwort in Canada. Until recently, L. jacobaeae was
considered to be monophagous under field conditions. We carried out adult feeding and oviposition as
well as larval development bioassays with a large number of European and North American Senecio
species to assess the fundamental host range of the Swiss biotype and to compare the results from the
bioassays with the phylogeny, the plant secondary metabolite profiles and the physiology of the test
plant species. The results will be discussed in the context of herbivoreplant interaction theories and of
current recommendations for setting up test plant lists in biological control projects.
Towards predicting establishment of Longitarsus bethae,
root-feeding flea beetle introduced into South Africa for
potential release against Lantana camara
D.O. Simelane
ARC-Plant Protection Research Institute, P/Bag X134, Queenswood 0121, South Africa
In an attempt to improve the selection of effective weed biocontrol agents, a study was conducted
to demonstrate that the prospective agent, a root feeding flea beetle Longitarsus bethae (Coleoptera:
Chrysomelidae), would establish by virtue of having the ability to cope with important environmental
and ecological conditions in the release areas. Although L. bethae had been proven to be adequately
host-specific to L. camara (lantana), releasing it into the environment would be a risk not worth taking unless there are reasonable grounds that it would establish and become prolific in its new range. A
multi-factorial experiment was arranged to determine how four environmental factors (soil moisture,
soil texture, the presence of T. scrupulosa and lantana variety) acting in combination might influence
the survival of immature stages of this beetle. Soil moisture and clay content had the most substantial
effect on survival of L. bethae, while the presence of T. scrupulosa and the type of lantana variety serving as a host had minimal influence on the beetles. Based on this investigation, a survival-prediction
equation was derived. This was used to identify three geographic regions that are likely to be suitable,
marginally suitable and unsuitable for L. bethae in South Africa. Based on host-specificity tests conducted previously and the results of the current study, it was strongly justified to release L. bethae in
South Africa.
363
XII International Symposium on Biological Control of Weeds
Host-specificity testing the French broom psyllid
Arytinnis hakani (Loginova)
T. Thomann1 and A.W. Sheppard2
CSIRO European Laboratory, Campus de Baillarguet, 34980 Montferrier-sur-Lez, France
2
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
The broom psyllids are known to have strong co-evolutionary relationships with their related host
plants. For this reason, the French broom psyllid, Arytinnis hakani (Loginova) (Homoptera: Psyllidae),
was selected as potential biological control agent against its host plant, Genista monspessulana (L.)
Johnson, a Mediterranean leguminous shrub invasive in Australia and California. Between 2002 and
2006, two types of host-specificity test were conducted on potted plants: (a) choice-without-target
tests, which evaluated the capacity of the insect to lay eggs on test plant species in the absence of the
natural host and (b) no-choice starvation tests, where the first-instar nymphs are forced to develop on
test plant species other than the natural host. Over 92 species were tested in 47 genera covering ten
plant families. The tests revealed that A. hakani can potentially develop on plant species from four
genera within the Genisteae tribe (including the target), with nymphal development on species from
two genera within the Thermopsidae tribe. The high number of species with nymph development in the
genus Lupinus (16 of 25 tested) may lead us to reconsider A. hakani as a potential biological control
agent against G. monspessulana in the USA. Further work on imported exotic lupines of economic
importance to Australia is required to assess potential for release there.
Prospects for the biocontrol of Banana Passionfruit in
New Zealand with a Septoria leaf pathogen
N.W. Waipara,1 A.H. Gourlay,2 A.F. Gianotti,1 J. Barton,2 L.S. Nagasawa3
and E.M. Killgore3
Landcare Research, Private Bag 92170, Auckland, New Zealand
2
Landcare Research, PO Box 40, Lincoln, New Zealand
3
Hawaii Department of Agriculture, 1428 South King Street, Honolulu, HI 96814, USA
1
Seven closely related vine species of Passiflora, all with the common name banana passion fruit and of
South American origin, have naturalized and become serious environmental weeds in various regions
throughout New Zealand. Banana passion fruit is capable of smothering trees, particularly those at
forest margins and in forest gaps. It often prevents regeneration of native plants and has therefore been
classified as a priority weed for biocontrol by invasive plant biosecurity managers in New Zealand.
It is also a significant environmental threat in Hawaii where it is known as banana poka. A successful
classical biological weed control programme was undertaken with the release in 1996 of a virulent leaf
pathogen, Septoria passiflorae. A similar biological control programme was initiated in New Zealand
to explore the efficacy and safety of S. passiflorae for its potential introduction against this rapidly
expanding and hybridizing weedy complex. Pathogenicity testing showed the fungus to be a virulent
pathogen against the banana passion fruit weed complex, with promising biocontrol prospects. However, its release in New Zealand may be prevented due to its potential damage to the closely related
commercially cultivated species Passiflora edulis.
364
Abstracts: Theme 4 Pre-release Specificity and Efficacy Testing
Novel preliminary host-specificity testing of
Endophyllum osteospermi (Uredinales)
A.R. Wood
ARC-Plant Protection Research Institute, P. Bag X5017, Stellenbosch 7599, South Africa
Chrysanthemoides monilifera ssp. monilifera (boneseed), indigenous to South Africa, is a serious invader of native vegetation in southeastern Australia. The rust fungus, Endophyllum osteospermi, causes
witches brooms on bone seed in South Africa but has a long latent period, typically between 6 and 24
months between infection and the initiation of the witches brooms. This long latent period makes the
logistics of doing traditional host-specificity testing, in which all test plant species are inoculated and
observed for symptom development, unfeasible for this rust fungus. Germination of aecidioid teliospores and penetration by basidiospores were observed on the surface of excised leaves of 36 test plant
species at 4 days after inoculation and were compared to that on bone seed. Germinating aecidioid
teliospores aborted on 14 test plant species, while no penetration was attempted on further 14 test plant
species. Penetration only occurred, or was attempted, on eight of the 36 test plant species in addition
to boneseed. Inoculating whole plants of nine selected test plant species confirmed the above results.
Therefore, only the test plant species in which penetration occurred, or at least was attempted, need to
undergo comprehensive host-specificity testing.
Potential of Ustilago sporoboli-indici for biological control
of five invasive Sporobolus grasses in Australia
K.S. Yobo1, M.D. Laing1, W.A. Palmer2,4 and R.G. Shivas3
University of KwaZulu-Natal, Department of Plant Pathology, Private Bag X01, Scottsville, Pietermaritzburg
3209, South Africa
2
Queensland Department of Natural Resources and Water, Alan Fletcher Research Station, PO Box 36,
Sherwood, QLD 4075, Australia
3
Indooroopilly Research Station, Department of Primary Industries and Fisheries, 80 Meiers Road,
Indooroopilly, QLD 4068, Australia
4
CRC for Australian Weed Management, Glen Osmond, SA, Australia
Sporobolus pyramidalis, Sporobolus africanus, Sporobolus natalensis, Sporobolus fertilis and Sporobolus jacquemontii, known collectively as the weedy sporobolus grasses, are exotic weeds causing
serious economic loss in grazing areas along Australias entire eastern coast. In one of the first attempts
to provide biological control for a grass, a smut fungus, Ustilago sporoboli-indici, has been found to
attack the leaves and flowering parts of S. pyramidalis, S. africanus and S. natalensis in South Africa.
The potential of this pathogen as a classical biological control agent for all five weedy Sporobolus spp.
found in Australia was evaluated in the glasshouse. The smut attacked S. pyramidalis, S. africanus,
S. natalensis and S. fertilis but not the New World S. jacquemontii. Host range trials with ten native
Australian Sporobolus spp. were also conducted. The extent of damage caused by the smut fungus to
two weedy Sporobolus spp. (S. fertilis and S. natalensis) and two native Australian Sporobolus spp.
(S. creber and S. elongatus) under glasshouse conditions was determined by measuring biomass and
effects on flower and seed formation. The prospects for the smut as a biocontrol agent are assessed.
365
This page intentionally left blank
Theme 5:
Regulations and Public Awareness
Session Chair: Dick Shaw
367
This page intentionally left blank
Keynote Presenter
Regulation of biological weed control agents
in Europe: results of the EU Policy Support
Action REBECA
R.-U. Ehlers1
Summary
Biological weed control uses invertebrate biological control agents (IBCAs) as well as plant pathogens. Unlike North America, Australia and New Zealand, Europe has no central agency responsible
for an environmental risk assessment (ERA) prior to the introduction of exotic IBCAs, and no classical biological control agents for weed control have been released in Europe. However, many exotic
and native insect and mite species are currently used in European horticulture. Since the establishment in Europe of the exotic coccinelid Harmonia axyridis (Pallas) and concerns about potential displacement of indigenous coccinelids, proposals and guidelines for the regulation of IBCAs have been
prepared. The European Union (EU) Policy Support Action Regulation for Biological Control Agents
(REBECA) has contributed to the development and harmonization of guidelines and implementation
of regulation procedures. However, regulation is organized on a national level. For progress in biological weed control, a well-recognized and knowledgeable European-wide organization dealing with
the risk assessment and authorization of IBCA is urgently needed. Microbial weed control agents and
products of natural origin have to be registered following the rules of the EU directive 91/414, which
treats such biological control agents (BCAs) almost like synthetic chemical plant-protection products.
REBECA has reviewed current legislation and guidance documents and made proposals for alternative, less bureaucratic and more efficient regulation procedures maintaining the same level of safety
for human health and the environment but accelerating market access, increasing the availability of
BCAs and lowering registration costs. Information on the progress of the REBECA Action is available on-line (www.rebeca-net.de). In Europe, the current expertise on risk assessment of biological
weed control agents is limited to a few experts. A lack of knowledge at the level of regulation authorities will result in exaggerating risks and implementation of unnecessary regulation. It is therefore
recommended to start a dialogue between weed control scientists and regulation experts immediately
in order to prepare the ground for the use of biologicals in weed control in Europe.
Keywords: regulation, registration, data requirements, REBECA.
Why regulation of biological
control agents?
Regulation is implemented by governments when human activities cause or threaten to cause damage to the
society. In order to avoid, prevent or minimize impacts,
Department for Biotechnology and Biological Control, Institute for
Phytopathology, ChristianAlbrechtsUniversity, HermannRodewald
Str.9, 24118 Kiel/Germany <[email protected]>.
CAB International 2008
1
regulation is necessary. It should concentrate on safety
aspects and try to minimize impacts on trade, economy
and the environment.
Biological control agents (BCAs) for weeds, with
very few exceptions, can be regarded as environmentally safe. In biological weed control, invertebrate
agents including nematodes (IBCAs) and microbial
pathogens have been released. These BCAs have been
the most successful, cost effective, safe and environmentally friendly method of weed management. Nonetheless, our societies demand information on the risks
and safety of BCAs.
369
XII International Symposium on Biological Control of Weeds
Management of risks in
modern societies
When regulating the use of BCAs, risks and benefits
have to be carefully considered. Any kind of exaggeration of risks is causing tradeoff effects (Graham and
Wiener, 1995). Regulation might keep older, riskier
technology (e.g. synthetic pesticides) in use. While
policy and society demands a reduction of chemical
control, overregulation of biological control can result in a more widespread use of chemicals or increase
damage caused by invasive weeds. In order to minimize
tradeoff effects, some fundamental rules of regulation
should be followed. Prior to the development of regulation, a costbenefit analysis should assess the magnitude of any problem and try to estimate the potential
environmental damage. The result of the costbenefit
analysis will answer the question: Do benefits of regulation justify costs of regulation? A risktrade-offanalysis should follow. Once trade-offs are identified in a
quantitative and qualitative way, target risks and countervailing risks must be weighed and affected population (e.g. farmers vs endangered species) be estimated
(Graham and Wiener, 1995). The last step is to develop
effective and inexpensive tools. The search for cheaper
and more effective tools to achieve the basic goal is of
major importance and might produce creative solutions
for risk assessment. These three principles are simple
but also quite powerful. If they were taken seriously
and implemented in the right way, they can improve
risk regulation and potentially save money and damage
to the environment. The analysis ensures that policy is
driven by full appreciation of relevant risks and not by
hysteria and alarm (Sunstein, 2002). Unfortunately, the
implementation of regulation of BCAs is not always
driven by this analytical approach but by the power
play between interest groups and tradition. Rules,
which have been developed in the past to protect the
environment from synthetic chemical plant protection
products, are now implemented on innovative, biological agriculture tools.
The REBECA action
As set out in the European Union (EU) Community Agriculture Policy (CAP) package 2003, developing agricultural techniques, which are both ecologically sound
and economically viable, will require new and powerful tools and assessment methods for the management
of weeds, pests and diseases in European agriculture,
horticulture and forestry. Biological control agents are
part of these tools. However, despite considerable research efforts on biological and natural control agents
(beneficial insect, mites and nematodes, microbial plant
protection products, plant derived substances and semiochemicals), the number of such products on the market
or projects in biological weed control in Europe is cur-
rently extremely low compared to other Organisation
for Economic Co-operation and Development (OECD)
countries. Therefore, the EU Commission published a
call for proposals to organize a policy support action.
As a result, the REBECA Action Regulation for Biological Control Agents was started in January 2006.
It reviews current legislation, guidelines and guidance
documents at Member State and EU level and compares these with similar legislation in other countries,
where the introduction of new BCAs has proven to be
more successful than in the EU. The studies are available on the REBECA webpage. Based on specific input by researchers, regulators and product developers,
proposals for appropriate and balanced risk assessment
demands for BCAs have been developed. It is expected
that no compromise to the level of safety will be made.
In fact, more adapted risk assessment strategies might
produce greater safety than occurs in the existing
system. Proposals for a balanced regulatory environment will lead to better access to BCAs for growers
and farmers and, therefore, to further reductions in the
use of chemical pesticides. The results will serve as a
scientific basis for reviewing current legislation and
guidance for BCAs. The REBECA Action will last until December 2007 and will present and has discussed
the outcome at a final conference in September 2021,
2007 in Brussels. Results are also available on the REBECA webpage.
Risks related with the use of IBCAs
No human activity is without potential risks. As a first
step, REBECA analysed the real and potential risks related with the use of biological control agents. Risks
related with IBCAs have been summarized by the
REBECA Action (see results and presentations of REBECA Wageningen Workshop at www.rebeca-net.de):
(1) Human and animal health: The probability of risks
to humans is considered to be remote and limited to allergic reactions and bites and stings. (2) Plant and crop
damage and development to nuisance: There are few
reports of crop damage, e.g. Macrolophus caliginosus
(Wagner) on tomato, or related problems such as the
contamination of crop products, e.g. Harmonia axy
ridis (Pallas) in grapes in the USA. The latter can also
be a nuisance when entering into houses. Plant damage
is of less importance in the use of IBCAs to control
plant pests; however, it is of major concern in biological weed control. (3) Environment: It was recognized
that most important risks of IBCAs, real or perceived,
are to the environment. These are: establishment of the
exotic species in a new country, parasitism or predation of non-target species, competition or displacement
of native species, perturbation of ecosystem functions,
e.g. pollination, introduction of contaminating agents
(pathogens, hyperparasites), and inter-breeding with
native species.
370
Regulation of biological weed control agents in Europe: results of the EU Policy Support Action REBECA
Data requirements for risk
assessment of IBCAs
The REBECA Action considered all aspects of biological control applications, classical and augmentative
(inundative) biological control as well as control of
plant pests and weed control. Data needed to perform
an environmental risk assessment (ERA) for IBCAs are
limited to exotic species. For indigenous species, only
data on identification should be required, and specimen
should be deposited in recognized collections. For the
ERA of exotic species, a hierarchical approach is recommended (Fig. 1).Within the scientific and regulatory
expert groups in the REBECA Action, risks posed by
exotic IBCAs were mainly related to host specificity,
whereas impacts to health were considered negligible.
For nonnative species, additional safety data should
inform about establishment, host range and dispersal.
The hierarchical system proposed by van Lenteren et al.
(2003), and updated in van Lenteren et al. (2006a, b)
should be adopted (Fig. 1). Besides the significant reduction of weed populations, another major measure
Figure 1.
for success of a weed control release is the dispersal
and establishment of a species. Consequently, dispersal
and establishment are a pre-requisite for successful biological weed control, and ERA can therefore concentrate on the host-range testing. For host-range testing,
REBECA experts recommended adoption of the testing
scheme proposed by van Lenteren et al. (2006b) and
to select nontarget species for host-specificity testing
as recommended by Kuhlmann et al. (2006), who proposed selection from three categories: (1) phylogenetically related; (2) occurs in the same ecological niche;
(3) unrelated safeguard species. Testing of the host
range is first done in laboratory trials using one nontarget species. If acceptance of nontarget hosts is observed in nochoice tests, further tests need to include
direct comparison of the acceptance and development
on nontarget species when the target species is simultaneously available. It is recommended that three treatments are compared with appropriate replication: (1)
target species alone (control); (2) nontarget alone (nochoice); (3) target and nontarget together (choice). Often, these laboratory experiments provide false positive
The hierarchical system proposed by van Lenteren et al. (2003), and updated in van Lenteren et al. (2006a, b) for risk assessment of invertebrate
biocontrol agents. Exit from the system with the answer yes will require a
permit for release, and with no the permit will be refused. Refusing might
be particularly conflicting as the past use of exotic IBCAs for pest control
has caused no or remote damage in Europe.
371
XII International Symposium on Biological Control of Weeds
results on the host range. If the non-targets are accepted
in these tests, experiments should be carried out under more natural conditions in contained environments
such as large cages in the country of origin. Field observations on the host range in the ecosystem of origin
might also be considered for a decision on the risks.
Direct and indirect effects are a summary of information gained from the available literature. When such
information is not readily available, these effects may
be estimated by expert knowledge or generated from
the data on establishment, host range and dispersal in
the ERA. Examples of direct effects would include effects on nontarget species and on other trophic levels
(such as intra-guild predation and omnivory), hybridization and enrichment and vectoring. Indirect effects
are those that occur when there is no direct interaction between the control agent and nontarget species,
such as competition and competitive displacement
(see van Lenteren et al., 2003; Bigler et al., 2006).
Indirect effects are difficult to quantify but are likely
to be related to the scale of the direct effects. Risks
related with the use of nematode are considered as exceptionally low as the dispersal is limited, and effects
on non-target hosts are transient (Ehlers, 2003). However, this does not include weed control nematodes
as they have not been considered yet. The REBECA
Action has checked the applicability of its recommendations for both weed and pest control agents and recommends the use of both approaches. Methods on the
procedures to assess environmental risks have been
developed and widely discussed among experts (e.g.
Bigler et al., 2006).
How to regulate IBCAs in Europe?
What is not yet decided is how regulation of IBCAs
should be implemented in Europe. Benefits by far
outweigh the risks posed by IBCAs. Therefore, these
lowrisk products should not be overregulated in order
to enable the market excess also for small- and mediumsized enterprises or make possible cost-effective biological weed control programmes. The past has taught
us that one of the major reasons why IBCAs have
gained a market share of >150 million euros per year
is the reduced level of regulation for these organisms.
Of 90 nematode, mite and insect species in the list of
biological control agents widely used commercially in
Europe and neighbouring states, 40% originate from
outside of Europe (EPPO, 2003). The European Plant
Protection Agency (EPPO) had a Panel on Safe Use of
Biological Control, which recognized that these species had been widely used in several EPPO countries
and concluded that other EPPO countries may therefore presume with some confidence that these agents
can be introduced and used safely. However, EPPO
is not an organization which gives authorization for
use of exotic IBCAs. Authorization in Europe is currently organized on a national level by different plant
protection organizations, quarantine offices or agencies
responsible for the natural environment. No Europeanwide organization exists which could authorize the use
of IBCAs also for weed control. To minimize bureaucratic efforts and costs, the regulation of IBCAs should
not be in the hands of organizations currently regulating agrochemicals and should follow rules specifically
developed for IBCAs. REBECA has provided information on data requirements and forms for applications.
What has not been solved is the problem of national
regulation. As IBCAs cannot be restricted to national
borders, a Europe-wide organization of the authorization is necessary. Committees including all stakeholders and member states should decide on the risks and
produce consensus reports in order to agree on positive lists of those IBCAs which pose no major risks
to the environment. Similar approaches have been successful in the past (EPPO list). Such panels might be a
solution for future assessment of risks related with the
use of IBCAs in Europe and can also be implemented
to assess the risk of IBCAs for biological weed control.
However, such an organization must be recognized
by all member states. Therefore, agreement between
member states should be sought to identify one European organization which will be responsible for the risk
assessment and authorization of IBCAs.
Risks related with the use of
microbial agents
The risks related with the use of microbial BCAs are
pathogenicity to humans, mammals and other non-target organisms, effects of microbial secondary metabolites on non-targets, sensitization and allergic effects,
genotoxicity, development of resistance to antibiotics,
persistence in the environment, entering into or accumulation in the food chain, genetic instability, microbial contamination and other adverse effects on the
environment. Most of the risks are reflected in data requirements for registration according to EU Directive
91/414 (Table 1).
Data requirements and
registration of microbial BCAs
The directive 91/414 does not make any difference
between agents used to control plant pests or diseases
or weed control micro-organisms. However, the registration process is expected to handle information requirements with some flexibility on a case-to-case
basis and not ask for data when no potential risk can be
identified. Considering authorities follow these recommendations, many of the data requirements might not
be relevant or applicable for weed control agents, and
regulators might be able to waive data requirements,
which would significantly ease the registration process
for weed control micro-organisms.
372
Regulation of biological weed control agents in Europe: results of the EU Policy Support Action REBECA
Table 1.
Data requirements for risk assessment according to Annex II B of the European Union Directive 91/414.
Data requirements
Identity
Registration is on strain level, every new strain has to go through the process of risk assessment
Methods to identify and determine strain must be provided
Methods to identify impurities (contaminating microorganisms, relevant metabolites) must be provided
Biological properties
Origin and natural occurrence
Target
Mode of action
Specificity
Dispersal and colonization ability
Genetic stability
Production of metabolites
Human health TIER I
Medical data
Sensitization (inhalation and skin)
Acute toxicity, pathogenecity and infectivity (oral, inhalation, intraperitoneal/subcutaneous)
Genotoxicity (metabolites of particular concern)
Short-term toxicity and pathogenecity (repeated exposure)
First aid measures
Human health TIER II
In vivo studies in somatic cells
In vivo studies in germ cells
Residues
Persistence and multiplication
Non viable residues
Viable residues
Fate and behaviour in the environment
Persistence and multiplication (soil, water, air)
Mobility
Effect on nontarget organisms
Birds
Aquatic organisms (fish, algae)
Bees
Other arthropods
Earthworms
Nontarget soil microorganisms
The procedure for EU registration of plant protection products containing micro-organisms is well defined. An applicant produces a dossier on the active
ingredient (ai) including all studies and data as summarized in Table 1. Pre-submission meetings of applicant and regulation authorities are recommended by
the REBECA Action to discuss possible waivers and
to prevent expensive investigations on non-relevant
risks. The dossier is submitted to one national regulation authority of choice. This agency, the rapporteur
member state, checks the completeness of the data set
(completeness check) and then prepares a Draft Assessment Report. This report is sent to the EU SANCO
(EU Directorate General for Health and Consumer Affairs) office responsible for the registration at the EU
level, to the European Food Safety Agency (EFSA) and
all member states. European authorities can consult experts on the risk of the ai. SANCO discusses the risk
assessment with national regulation authorities. If the
authorities consider the risks can be managed appropriately without major damage to users, consumers and
the environment, then the ai is listed on the Annex 1 of
Dir. 91/414. The final decision on the inclusion of the
ai on the Annex 1 is taken by the Standing Committee
on the Food Chain and Animal Health (SCFCAH) of
EFSA together with SANCO (European Commission).
If included, the ai cannot yet be used. National registration of the commercial product has to follow.
Data requirements of the formulated product involve
trade name and composition (formulation), physical,
chemical, technical properties (storage stability, particle size distribution etc.), data on application (intended
use, mode of action, application method and rates,
number and timing of applications, proposed instructions for use including safety indications), information
on reentry periods, cleaning of application equipment,
373
XII International Symposium on Biological Control of Weeds
measures in case of an accident, efficacy data. Data on
human health, exposure under the proposed conditions
of use, residues, fate and behaviour in the environment
and effect on non-targets are similarly requested as for
inclusion on Annex 1. The procedure is laid down in
the Dir. 91/414 which was amended by Dir. 2001/36/
EC (Data Requirements MicroOrganisms). The European Parliament and Council are currently discussing
a proposal for a regulation concerning the placing of
plant protection products on the market, which will replace 91/414. All relevant information, guidelines, directives, position papers of stakeholders and results of
the REBECA Action in relation to this discussion are
available on the webpage.
Are costs too high to allow use of
microbial agents in weed control?
The situation with microbial BCAs is different. Due to
the existing regulation, products do not make it to the
market or are withdrawn immediately when regulation
is required. Some member state governments try to
find ways around 91/414. In Italy, until 2006, microbial
products which only had the Latin scientific name as a
product label did not need any registration. Switzerland
had experts reviewing reduced data files. As a result,
Switzerland has the largest number of BCA products
available on the market. Now, the country wants to
adapt to European rules. Some countries implemented
a listing for low-risk products, such as plant growth
promoters, and only demanded a minimum of safety
data. Weed control agents might be possible candidates for such lists of low-risk agents. Within the EU,
reduced data requirements for low-risk products are
also discussed. However, it is not yet well defined what
low-risk or basic products are.
Negative examples are also available for weed biological control. A European project considered fungal
pathogens to control the Giant Hogweed but dropped
the idea when participants learned that the agent had to
go through registration according to 91/414 (M. Cook,
2007, personal communication). Whereas classical control programmes are area-wide approaches supported
by public funds, the commercial application is using
BCAs to substitute chemical measures on a limited
area. In the past, the commercial use of weed control
agents in Europe was restricted to The Netherlands,
Germany and Belgium. Koppert (The Netherlands)
sold the product Biochon containing the fungus Chon
drostereum purpureum ([Pers.: ex Fr.] Pouzar) to control the introduced black prunus, Prunus serotina Ehrh.
First introduced to diversify the forest, it later turned
out to out-compete native shrubs and trees, and forestry
wanted to remove the plant. The commercial success
was limited. The product had been sold as a wood-rot
promoter. When criticism arose and registration ac-
cording to the directive 91/414 EEC was demanded,
the company withdrew the product from the market.
Costs related with the production of safety data would
have outranged the possible commercial turnover by
far (W. Ravensberg, 2007, personal communication).
Fortunately, regulation authorities now ask for less
data and are prepared to give waivers for certain data
requirements for microbial agents. Thus, costs for registration are lower than for chemical agents. However,
one must calculate with approximately 0.43.0 million euros. In area-wide weed control, the costs related
with the damage caused by invasive weeds might,
however, justify costs for registration. Economic estimates in Germany concluded that the control of Giant
Hogweed, Heracleum mantegazzianum Sommier and
Levier, reaches a yearly amount of 12.3 million euros
and of Japanese Knotweed, Fallopia japonica Houtt.,
32.3 million euros. The costs to treat patients suffering
from allergic asthma or rhinitis caused by Ambrosia
artemisiifolia L. was estimated at 32.1 million euros
(Reinhardt et al., 2003). However, these costs do not
give indication of the market size for a commercial
biological control product, particularly as for weed
control cheaper chemical products are available and
might be more effective when extinction during early
establishment is required. But for reduction of problems with, e.g. the Japanese Knotweed, the introduction of pathogens could be a solution. The question
is, of course, how to deal with registration. Following
the European rules, pathogens would at least need to
go through the registration procedure for inclusion in
Annex 1.
Going through the list of requirements (Table 1),
many questions could be answered immediately about
risks of microbial weed control agents, and the conclusion would be no risk. However, SANCO rules follow the precaution principle, and authorities might possibly request experimental data for minor risks as well,
which would make registration of a weed control agent
an expensive exercise. Should, however, authorities
give waivers for data on minor risks, then the registration might be less expensive than what has been spent
in the past for the registration of BCAs to control insects and diseases. Scientific information and consideration of expert opinions can produce more confidence
on the safety of weed control agents among authorities
and justify provision of waivers.
Major concern with microbial agents is infectivity
and pathogenicity to humans and possible effects of microbial metabolites. For weed control agents, these concerns might be of much less importance than for other
BCAs. Medical doctors are well informed about every
possible human pathogen or toxin producer causing food
poisoning. EU Dir. 2000/54 gives precise information
about which microbial organisms might cause damage
to workers. Thus, for some organism, a risk assessment
has already been done. Can we then waive the human
374
Regulation of biological weed control agents in Europe: results of the EU Policy Support Action REBECA
risk investigations if our microbial agent is not on the
list? Can we argue that humans and other non-targets
have co-evolved with these organisms and therefore
authorities can waive many of the data requirements on
the eco-toxicology? For instance, tests on earthworm
toxicology are required, but a literature search has not
resulted in any report of microbial pathogens found in
these animals. Is it necessary to answer questions on
residues possibly entering the food chain when the EU
project Risk Assessment of Fungal Biological Control
Agents (www.rafbca.com) has provided evidence that
fungal BCAs do not pose significant risks? Several requirements carried over from risk assessment of chemical products can certainly be waived. Should proposals
of the REBECA Action be acceptable to policy makers
and regulators, we might see a brighter future also in
the use of microbial weed control agents.
To exploit the use of biological weed control agents
in Europe, research projects should also consider the
use of micro-organisms. Weed control scientists should
contact experts experienced in the application for authorization of microbial control agents and together
discuss relevant risks and data requirements and estimate costs related with the registration process. Public support might be available as no registration for a
microbial biological weed control agent has ever been
applied for before. Such an experience can afterwards
serve as an example to define waivers and produce
guidance documents to facilitate future registration
procedures of microbial weed control agents. The scientific community dealing with biological weed control will not be able to progress without recognizing
the information requirements by public authorities.
Their request for data on the safety goes beyond the
assessment of non-target effects on plants and the environment, which is identified as the major risk by weed
control science. Contact between regulators and potential applicants should be made at an early stage of weed
control projects to exchange knowledge and produce
an environment for a scientifically based risk assessment, which considers concerns of all stakeholders. If
weed control science and regulation will not enter into
this dialogue, biological weed control in Europe will
not have a future.
References
Bigler, F., Babendreier, D. and Kuhlmann, U. (2006) Environmental Impact of Invertebrates for Biological Control
of Arthropods. Methods and Risk Assessment, CABI Publishing, Wallingford, UK. 299 p.
Ehlers, R.-U. (2003) Biocontrol Nematodes. In: Hokkanen,
H.M.T. and Hajek, A. (eds) Environmental impacts of
microbial insecticides. Kluwer Scientific Publishers, Dortrecht, NL, pp. 177220.
EPPO (2003) EPPO Standards on Phytosanitary Measures
Safe use of biological control. List of biological control agents widely used in the EPPO region [PM 6/3(2)].
Available at: http://archives.eppo.org/EPPOStandards/
biocontrol_web/bio_list.htm
Graham, J.D. and Wiener, J.B. (1995). Risk Versus Risk
Tradeoffs in Protecting Health and the Environment.
Harvard University Press, Cambridge, UK. 271p.
Kuhlmann, U. Schaffner, U. and Mason, P.G. (2006) Selection of nontarget species for host specificity testing. In:
Bigler, F., Babendreier, D. and Kuhlmann, U. (eds) Envi
ronmental Impact of Invertebrates for Biological Control
of Arthropods. Methods and Risk Assessment. CABI Publishing, Wallingford, UK, pp. 1537.
Reinhardt, F., Herle, M., Bastiansen, F. and Streit, B. (2003)
konomische Folgen der Ausbreitung von Neobiota. Forschungsbericht 201 86 211, Umweltbundesamt Berlin,
Germany, 248p.
Sunstein, C.R. (2002) Risk and Reason - Safety, Law and the
Environment. Cambridge University Press, Cambridge,
UK, 342p.
van Lenteren J.C., Babendreier, D., Bigler, F., Burgio, G.,
Hokkanen, H.M.T., Kuske, S., Loomans, A.J.M., Menzler
Hokkanen, I., van Rijn, P.C.J, Thomas, M.B., Tommasini,
M.G. and Zeng, Q.Q. (2003) Environmental risk assessment of exotic natural enemies used in inundative biological control. BioControl 48, 338.
van Lenteren, J.C., Bale, J.S,. Bigler, F., Hokkanen, H.M.T.
and Loomans, A.J.M. (2006a) Assessing risks of releasing
exotic natural enemies of arthropod pests. Annual Review
of Entomology 51, 609634.
van Lenteren, J.C., Cock, M.J.W., Hoffmeister, T.S. and
Sands, D.P.A. (2006b) Host specificity in arthropod biological control, methods for testing and interpretation of
the data. In: Bigler, F., Babendreier, D. and Kuhlmann, U.
(eds) Environmental Impact of Invertebrates for Biologi
cal Control of Arthropods. Methods and Risk Assessment.
CABI Publishing, Wallingford, UK, pp. 3863.
375
Avoiding tears before bedtime:
how biological control researchers
could undertake better dialogue with
their communities
L.M. Hayes, C. Horn and P.O.B. Lyver
Summary
While many people support biological control of weeds, there are many others who doubt its safety
and effectiveness, and universal agreement about what constitutes a weed is also rare. Scientists who
attempt to develop biological control programmes without ensuring effective two-way communication with their communities may, at best, experience poor uptake of their research and, at worst,
serious opposition that jeopardizes progress or even makes biological control programmes untenable. We describe a technique that allows researchers to engage in meaningful dialogue with their
communities. People are brought together in an environment where it is safe to express opinions and
concerns, where they are encouraged to listen and deepen their understanding, where they can discover common ground and build trust, and collectively find acceptable, or even novel, ways to move
forward. This process could be adapted for other topics and social settings as long as key elements
remain. We also describe some key messages about biological control of weeds that arose from this
dialogue and suggestions for how things could be done differently in order to gain greater acceptance
from the general public.
Keywords: communication, public engagement, decision making.
Introduction
While many people are supportive of research to develop more natural and sustainable methods of weed
control, others doubt the safety and effectiveness of
biological control. Even other scientists are sometimes
uncomfortable with or opposed to the practice (e.g.
see Louda et al., 2003), and universal agreement about
what constitutes a weed is rare (Stanley and Fowler,
2004). Currently, community perceptions are often
more oriented to the threat biological control agents
pose than the benefits of controlling weeds, and governments are becoming increasingly risk averse. These
factors are causing extensive delays and fewer agent
approvals in some countries while environmental damage caused by weeds continues unchecked - biological
control as a discipline could be heading towards a crisis
(Sheppard et al., 2003; McFadyen, 2004).
Landcare Research, PO Box 40, Lincoln, New Zealand.
Corresponding author: L.M. Hayes <[email protected]>.
CAB International 2008
These challenges are not just restricted to biological control of weeds. Around the world, the presumption that science knows what is best for society is under
challenge (Cribb, 2003), and scientists ignore public attitudes and values at their peril (House of Lords, 2000).
Hipkins et al. (2002) found wide agreement (69%) to
a statement that scientists should have to explain and
justify their research to the public.
Lobbying for changes to biological control regulatory procedures is likely to only be a small part of the
solution. Organizations undertaking biological control
also need to ensure that effective two-way communication takes place between practitioners, regulators, critics and stakeholders to enhance understanding (Sheppard et al., 2003) and also to identify and resolve any
potential areas of conflict or concerns or misconceptions at an early stage.
In New Zealand, and it would seem worldwide
(Sheppard et al., 2003), biological control priorities are
currently being decided and agents imported with little
or no input from most stakeholder groups or the general public. Consultation is often undertaken only when
376
Avoiding tears before bedtime
it is a regulatory requirement and is not used to enhance
the quality of decision making or improve public engagement. When human beings are suddenly presented
with a well-developed plan, they will often oppose it
because they feel backed into a corner.
What is dialogue?
We use the term dialogue to describe two-way communication where each party listens to and respects the
other point of view and seeks to find ways of moving
forward that are of benefit to both parties. Increasingly,
a number of dialogue processes are used worldwide for
dealing with social issues, e.g. focus groups, citizens
juries, consensus conferences, deliberative polling etc.
Traditionally, the response to concerns about science and technology has been to try to better educate
the public based on the presumption that if people
understand the science better they will be more supportive. This assumes that science is universally comprehensible (Wilsdon and Willis, 2004), when the level
of scientific literacy in the wider community is often
not particularly high. In addition, this communication
model misses the fact that science is not the only thing
that people want taken into account during decision
making. People often want to know about the need for
the technology, the motivation behind developing it,
the trustworthiness of the organizations undertaking
it (Kass, 2001) and any potential unintended consequences in the long term. Even when people understand
science better, they may not support projects if these
conflict with their belief and value systems.
People assess risk in different ways. What the public
may find acceptable may be at odds with objective risks
as understood by science (House of Lords, 2000). Decisions about biological control agents have to be made
on the basis of available information. Decision making
must cope with the fact that the world is imperfectly
understood, complex, dynamic and open-ended. Therefore, there can be major benefits in including people
with a wide range of different expertises and perspectives in discussions and decision making processes to
provide alternative perspectives that help to deal with
inherent uncertainty (Kass, 2001). By ensuring that
decision-makers listen to public values and concerns
and give the public some assurance that their views are
taken into account, the likelihood of decisions finding
acceptance is increased.
A novel approach to dialogue
This paper describes a novel dialogue approach that
initially linked ideas taken from traditional Mori protocols and The Seven Habits of Highly Effective People
(Covey, 1990) and was later pared down to key elements of both.
Mori are the indigenous people of New Zealand
and have traditionally used marae (communal meeting
area) to discuss or debate issues relevant to the community. Traditional practices and customs are used to
welcome visitors to meetings and onto the grounds of
the marae. During meetings, people commonly sit in a
circular fashion with their backs to the inside wall of
the wharenui (meeting house) often on mattresses on
the floor. It is also common for the local people and
visitors to sleep overnight in the meeting house if the
meeting lasts for more than a day. Observing traditional
Mori customs and values helps to foster a spirit of reverence amongst people and provides an alternate view
of the world.
The Seven Habits offers ideas for living more effectively; some that we incorporated into our process
included building trust, encouraging winwin thinking, achieving greater understanding through improved
listening skills, valuing differences and creating third
alternatives (not your way or my way but a better way
than any of us have thought of yet).
Initially, our process was used to undertake dialogue
on the use of biological control for weeds and 1080 (sodium monofluoroacetate) to control mammalian pests
(see Lyver et al., 2004; Hayes et al., 2004). Later, we
tested a pared-down process on the development of a
pest control strategy for the Hawkes Bay Region of
New Zealand (see Lyver et al., 2006). In this paper, we
focus mainly on the learning we gained from using our
dialogue process.
Methods
We tested our initial process four times between September 2003 and April 2004 at four different marae.
We invited a wide range of stakeholders to participate
including Mori, government departments, pest control
agencies, the Environmental Risk Management Authority, industry groups, scientists, students and lobby
groups. We aimed (with mixed success) to have a balanced gender and age mix. The number attending each
event varied from 16 to 31. Where attendees were not
local, we assisted with travel expenses. We also covered the cost of food and accommodation for all. Participants generally sat on mattresses on the floor in the
meeting house, and most slept over on them too.
At the meetings held on marae, Mori facilitators
helped participants to observe appropriate protocols. At
all the meetings, an independent facilitator managed the
group work. She established expectations, outlined the
Seven Habits, ensured ground rules were observed,
managed any conflict, taught listening skills, assisted
in the development of third alternatives and reflected
on the process. Main points of discussions were written
down on large flipcharts, and a summary of the entire
proceedings was produced.
On the first day, once the formalities and the expectations session were complete, our facilitator asked
participants to present their points of view and encouraged everyone to listen carefully since they would have
377
XII International Symposium on Biological Control of Weeds
to present someone elses viewpoint the following day
(and they did not know whose this would be). She
asked them to suspend judgement and assume they did
not understand. She allowed questions of clarification
only and stopped participants who challenged others
beliefs. At the end of the day, a social function allowed
people to chat in a relaxed and informal setting.
On the second day, facilitators gave a demonstration
of poor listening and good listening and then put participants into pairs (where possible with differing views).
Each person had 30 min to explain their viewpoint
again, and their partner was asked to reflect this information back to ensure they had clearly understood. At
the end of the hour, the group came back together, and
people took turns at presenting their partners perspective. Afterwards, the group identified common ground
and broke into smaller groups to think of third alternatives before a final summing up and farewells. Participants completed an evaluation survey before leaving.
The pared-down process was used once in May
2005 and was designed to see if the process would
work when real management outcomes were at stake
and when the meeting was held at a venue other than a
marae, in this case a winery. Chairs were arranged in a
semi-circle, and there were some novel objects to create
a sense of the unexpected, difference and novelty (e.g.
deck chairs with cartoon characters on them, cush balls
and an inflatable dog). People went home at the end
of the day. Time that previously had been devoted to
observing Mori protocols was spent working on outcomes. However, some key elements of Mori customs
were still used but adapted such as making people feel
welcome, introducing yourself fully and taking turns
at speaking. Instead of reporting back other peoples
perspectives to the wider group, small groups revisited
the task in hand based on their new understanding of
each others perspectives.
Results
Organizers perspectives and observations
Meeting preparation: Setting up the dialogue meetings took much more time than expected (weeks rather
than days). Many preliminary meetings, phone calls,
emails and letters were required to identify potential
participants and secure their participation. It was generally difficult to recruit stakeholders who were not already familiar with or involved with the issue. It was
easier to get men to attend and present information
than women. People from large organizations found it
easier to attend than the self-employed. People were
not always able to attend for the entire two days and,
since the dialogue process builds upon itself, this was
a disadvantage. Not all stakeholder groups were able
to attend, and some had reservations about attending a
meeting on a marae. There were no objections raised to
attending a meeting at a winery.
The actual events: Overall, we felt that the meetings
all worked well. The majority of participants showed
enthusiasm for the idea of dialogue, and even those
who came with some degree of doubt or suspicion participated fully without conflict under the guidance of
our facilitators.
Sharing meals and, on the marae, the experience of
communal living allowed people to get to know each
other quickly. Social events at the end of the first day
were an important and enjoyable opportunity for people to mingle and converse in a relaxed way. It was
not necessary for people to stay together overnight for
bonding to occur, but it did help. As people got to know
each other and discovered common ground, it became
more difficult to stop them from talking and come back
to group work. Uncovering common ground was a
vital part of the success of our dialogue process. For
example, realizing they had a shared concern for the
environment allowed participants to discuss contentious issues more constructively. We were pleasantly
surprised at the amount of humour and camaraderie
and the lack of aggression, particularly as some people
attending (especially the 1080 meetings) had a history
of prior conflict.
During the paired listening exercise, it was difficult for some participants to listen and reflect back,
as it is not something they commonly have to do, and
they needed help from the facilitators to stay on track.
Some people expressed surprise at discovering something new about their partners position. Often at the
beginning of the exercise, people were surprised that
a whole hour was set aside but at the end said they
could have used longer. Overall, the presenting back
of someone elses perspective to the wider group was
done well and often involved humour. Some found the
presenting-back useful because it allowed them to hear
what people were saying a second time. Others realized
how poorly we usually listen and how important good
listening is.
Participants evaluations of the
marae-based meetings
Overall satisfaction ratings and written comments:
The majority of stakeholders rated their overall experience as great or good (Figure 1). All but one of
the 81 participants said they would be prepared to participate in a similar dialogue process again. The most
common reasons given were that it was an excellent
non-confrontational forum to hear alternative views,
share information, network with people and improve
understanding and awareness. Some expressed surprise
and delight that they had been able to participate meaningfully and talk to scientists. They were surprised to
find that scientists were people like them with many of
the same values. The scientists at these meetings were
also enthusiastic about the interactions and to hear how
others perceive their work and appreciated the sugges-
378
Avoiding tears before bedtime
Figure 1.
Stakeholder ratings of the overall dialogue meeting experience (biocontrol 1: n = 19, biocontrol
2: n = 12, 1080 1: n = 21, and 1080 2 n = 29).
tions for better ways of doing things that were generated (see Appendix 1).
Long-term impressions: We talked with a few participants some time after the event. Overall, people were
positive about the effect that the meeting had had on
them and on their networks. They had often maintained
some contact with others they met at the meeting, even
those who had previously viewed each other with some
suspicion.
Participants evaluations of the
pared-down process
Overall satisfaction ratings and written comments:
Most of the 16 participants who filled in surveys rated
the meeting as great or good, and no one felt that
it was not great or bad. Reasons why the meeting
was worthwhile were similar to the marae-based meetings. Since many of this group had met before, it was
Figure 2.
noticed that some who are normally quiet spoke up
and contributed in ways that they had not done before.
Participants also commented that the process was fun
and challenging, the venue was good, the meeting was
well structured with different activities to keep people
interested and focused, it was a safe place in which to
exchange ideas, and useful outcomes were produced.
The scientists involved were heartened to see that their
work was valued and useful to others.
When asked about how they rated this style of meeting compared to other pest-management strategy meetings they had attended, there was a mixed response. Half
said that the meeting was better than other meetings
they had attended. A third did not know, and reasons for
this were not explored, but some had not been involved
in pest-management meetings before (Figure 2).
Post-meeting follow-up: We discussed how the meeting had gone with Hawkes Bay Regional Council staff
the week after the event and again 5 months later. In
Participants rating of dialogue meeting compared with other strategy development meetings
(n = 16).
379
XII International Symposium on Biological Control of Weeds
both sessions, they noted that the process had been
enjoyable, that people attending had participated well
and that it had been useful to expose people (including themselves) to a wider range of perspectives than
usual. They also noted that, compared with other processes they had used, the dialogue process had worked
well, as it got a lot more information and perspectives
on the table before any debate was generated. This
kept options open for much longer and provided a better overview of what was needed, wanted or expected.
More was achieved in a shorter time. The process also
created more ownership, buy-in, goodwill and enthusiasm. Involving scientists was considered to be beneficial as it allowed the group to learn more about science
being undertaken on pest management and the ways in
which science could strengthen the strategy. Overall,
the dialogue meeting was deemed memorable as it was
totally different from what people were used to, even
though some had initially expressed some discomfort about coming to something without a traditional
agenda.
Discussion
with and those with an established interest in the topic,
to take part. People need to feel safe about participating, especially in situations where there has been a history of conflict. Some groups will find it more difficult
to participate for reasons such as childcare or work
commitments.
Time required for dialogue
Adequate time is also needed to undertake a dialogue event. Participants ideally need to be present for
the entire time because the process builds upon itself.
Two days is a long time commitment, yet at the end
many felt they would have liked more time. It may be
possible to reduce the time needed by, for example,
sending out information for people to read beforehand,
but the same time pressures are likely to apply to any
preparation. It may also have been easier to convince
people to come to the meetings if they had been shorter
in duration, and there will always be trade-offs between
getting people to attend and what can be covered at the
event.
Facilitation
Dialogue may seem to be a lengthy and costly process,
but for potentially contentious issues, it could prove
quicker and cheaper in the long run. There is some
evidence that decision making involving appropriate
public engagement is quicker and less controversial because more effort is put into framing problems, debating options and agreeing solutions than pushing relentlessly forward with unpopular decisions (Kass, 2001).
Winstanley et al. (2005) suggest that researchers
benefit from discussions about social, cultural, ethical
and spiritual issues associated with their research. Certainly all the scientists who participated in our dialogue
events were surprised by just how useful they were in
terms of clarifying understanding of issues, developing ideas for better ways of doing things and forming
new relationships that would be helpful to them in the
future.
There are many different ways to undertake dialogue, and inevitably, some topics and groups will be
more challenging (Roper et al., 2004). Our process
worked well regardless of the topic, so we believe it
has wide applicability and could be adapted and simplified even further for use in different social settings
as long as key elements are not lost. So what are these
key elements?
Initial relationship building
The dialogue event itself is very much the tip of
the iceberg. The work required to develop relationships with stakeholders takes much more time than the
event itself. Others trialling dialogue processes have
also found this (Winstanley et al., 2005). It is easier
to convince groups you have already had prior contact
We found, like Cronin and Jackson (2004), that
an important underlying principle is respect for others and their point of view. Therefore, it is important
to have a skilled facilitator to run a dialogue event.
Everyone taking part needs to be treated with equal
respect, and ground rules (e.g. everyone gets a chance
to speak without fear of interruption or being challenged) must be established early on. As some of our
participants noted, the simple act of turn-taking and
preventing interjections can have a very big effect on
what gets put on the table for later discussion. This,
in turn, can have a major effect on the outcomes of that
discussion.
Some stakeholders will be keen to get some sense
of resolution from dialogue, which may or may not be
the reason for undertaking it. It is important, therefore,
to outline expectations at the start and formally close
the dialogue at the end. It also needs to be clear upfront
how any dialogue will contribute to decision making.
It is important not to create any sideshows that distract
from the dialogue. We found that the Seven Habits became a distraction as people were interested in learning
more about them, and it was better to just teach underlying principles.
Active listening
Although we highlighted the importance of listening
at an early stage, we found that most people had difficulty listening without interruption to points of view
that challenged their assumptions or points of view.
Other groups involved in dialogue have found the same
(Winstanley et al., 2005). However, giving people the
opportunity to walk in someone elses shoes for a
380
Avoiding tears before bedtime
short time can be a profoundly useful way of improving understanding and goodwill.
people who hosted, helped with, or attended one of our
dialogue events.
Enabling new behaviour
To be able to undertake good dialogue, people
need to be able to step out of established or traditional
expectations, roles and ways of doing things (Winstanley et al., 2005). The venue can help or hinder in
this respect. A neutral venue with appropriate seating
(Roper et al., 2004) where people can preferably sit
face to face is best. A venue, such as a marae, that
helps to break down normal patterns of behaviour and
engenders respect and tolerance for difference, may
be better than the usual meeting or conference rooms,
which may have history or expectations attached
to them. However, any room can be made better by
paying attention to seating arrangements and having
items that create a sense of novelty, difference and
curiosity.
Building relationships
The sharing of food is well known to be a good way
of getting people to relax and engage with each other.
Communal living can also help to break down barriers quickly but is not essential to good dialogue. Once
the ice is broken and people start talking to each other,
common ground is often found very quickly, which
helps to build trust and allow people to work together
constructively. It can be useful to explicitly identify
this common ground because points of difference may
be much smaller or even different from what people
previously thought.
Conclusions
Our process was a successful way of engaging stakeholders in dialogue about pest control. It was effective
because it allowed each participant to weigh up the
pros and cons for themselves in an environment where
those in authority were not able to move into persuasion mode and where scientific knowledge was only
one part of the equation. This process helped people
discover common ground to learn new things about the
standpoints of others and to build trust and understanding. It allowed constructive suggestions as to possibly
more effective new ways of doing things and areas
needing further thought or research. We believe that
this process or some variation of it could be used to
improve future decision making for all kinds of issues.
Dialogue is not something to be entered into lightly, but
the rewards can be enormous.
Acknowledgements
We thank the New Zealand Ministry for Research, Science and Technology for funding this work and all the
References
Covey, S.R. (1990) The Seven Habits of Highly Effective
People. Simon & Schuster, New York, NY, USA. 340 p.
Cribb, J. (2003) Water - The Australian Dilemma. Australia
Academy of Technological Sciences and Engineering,
Parkville, Victoria. Available at: www.atse.org.au/index.
php?sectionid=124 (accessed April 2007).
Cronin, K. and Jackson, L. (2004) Hands across the water.
Developing dialogue between stakeholders in the New
Zealand biotechnology debate. Ministry of Research, Science and Technology Report, Wellington, New Zealand.
Available as pdf at: www.morst.govt.nz/uploadedfiles/
Documents/ (accessed January 2006).
Hayes L., Horn C. and Lyver P. (2004) Taking the com
munity with you: a process for developing acceptable pest
control strategies. New Zealand Science Review 61,
6668.
Hipkins, R., Stockwell, W., Bolstad, R. and Baker, R. (2002)
Commonsense, trust and science: How patterns of beliefs
and attitudes to science pose challenges for effective communication. Ministry of Research, Science and Technology
Report, Wellington, New Zealand. Available as pdf at:
www.morst.govt.nz/uploadedfiles/Documents/ (accessed
January 2006).
House of Lords (2000) Science and Society. Science and Technology Select Committee, House of Lords, The United
Kingdom Parliament, UK. Available at: www.parliament.
the-stationery-office.co.uk/ (accessed January 2006).
Kass, G. (2001) Open channels: public dialogue in science
and technology. Report No. 153, Parliamentary Office of
Science and Technology, London. 41 p.
Louda, S.M., Pemberton, R.W., Johnson, M.T. and Follett,
P.A. (2003) Nontarget effects - the Achilles heel of biological control? Retrospective analyses to reduce risk associated with biocontrol introductions. Annual Review of
Entomology 48, 365396.
Lyver, P., Hayes, L.M. and Horn, C. (2004) A process for enhancing dialogue on biosecurity issues. Report to the Ministry of Research, Science and Technology, Wellington,
New Zealand. Available at: www.landcareresearch.co.nz/
research/social/documents/ (accessed January 2006).
Lyver, P., Hayes, L.M. and Horn, C. (2006) Using dialogue to
develop a more robust regional pest management strategy:
final report 2005/06. Report to the Ministry of Research,
Science and Technology, Wellington, New Zealand. Available as pdf at: www.landcareresearch.co.nz/research/social/
documents/ (accessed June 2006).
McFadyen, R.E. 2004. Biological control: managing risks or
strangling progress. In: Sindel, B.M. and Johnson, S.B.
(eds) Proceedings of the 14th Australian Weeds Conference. Weed Society of New South Wales, Sydney, Australia, pp. 7881.
Roper, J., Zorn, T. and Weaver, C.K. (2004) Science dialogues. The communicative properties of science and
technology dialogue. Ministry of Research, Science and
Technology Report, Wellington, New Zealand. Available
as pdf at: www.morst.govt.nz/uploadedfiles/Documents/
(accessed January 2006).
381
XII International Symposium on Biological Control of Weeds
Sheppard, A.W., Hill, R., DeClerck-Floate, R.A., McClay,
A., Olckers, T., Quimby Jr., P.C. and Zimmerman, H.G.
(2003) A global review of riskbenefitcost analysis for
the introduction of classical biological control agents
against weeds: a crisis in the making? Biocontrol News
and Information 24(4), 91N108N.
Stanley, M.C. and Fowler, S.V. (2004) Conflicts of interest
associated with the biological control of weeds. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M.,
Morin, L. and Scott, J.K. (eds) Proceedings of the XI International Symposium on Biological Control of Weeds.
CSIRO Entomology, Canberra, Australia, pp. 322340.
Wilsdon, J. and Willis, R. (2004) See-through science: why
public engagement needs to move upstream. HenDI Systems, Demos, London. Available at: www.demos.co.uk/
catalogue/paddlingupstream/ (accessed January 2006).
Winstanley, A., Tipene-Matua, B., Kilvington, M., Allen, W.
and Du Plessis, R. (2005) From dialogue to engagement?
Learning beyond cases. Ministry of Research, Science and
Technology Report, Wellington, New Zealand. Available
as pdf at: www.morst.govt.nz/uploadedfiles/Documents/
(accessed January 2006).
382
Avoiding tears before bedtime
Appendix:
Feedback on biological control of
weeds and third alternative ideas
Overall, participants were supportive of biological control of weeds. They liked the idea of natural control but
were not so keen that this involves introducing exotic
organisms. Many were keen to minimize the use of
chemicals. All want to have a choice of weed control
tools, including new and better tools, and more information about integrated weed management.
Scientists need to communicate with their communities at a much earlier stage. There is no universal
agreement about which plants are weeds, and people
want to have more say in what targets are tackled for
biological control. Some in the nursery industry would
be prepared to sacrifice some of the plants they sell in
order to control closely related weeds, and beekeeping
is not necessarily compromised by biological control
since the weeds do not disappear completely. There
needs to be increased effort to improve public awareness of the seriousness of weeds.
Scientists need to have adequate funding to do communication, and many scientists would benefit from
more training in communication skills - the public would
prefer to hear from them than public relations people,
and the media has its own separate agenda. People find
terminology like biological control a bit daunting, and
it may be better to find more friendly descriptors.
A lot of dissatisfaction was expressed with tradi
tional consultation processes. Undertaking communication only when it is time to apply to release a new
agent is not satisfactory. Many stakeholders want to be
involved in true dialogue. Scientists need to explore
different points of view and concerns and take time to
address them. People want cultural, spiritual and economic values and traditional knowledge to be taken
into account, not just scientific values.
People want more information about all aspects
of biological control, including expectations. Even if
100% guarantees cannot be made about effectiveness,
people want 100% guarantees about safety. It will
never be possible to do this, but many will be prepared
to accept the risks once they have the opportunity to
consider historical safety, the thoroughness of regulatory procedures and the disciplines commitment to
best practice.
Follow-up must be done on all biological control
agents, and scientists need to provide more assurances
that this is occurring. There is a paucity of published
studies worldwide showing what happens to weed populations after control of any kind. Obtaining sufficient
resources to undertake adequate follow-up is always
problematic, and scientists will need to undertake further dialogue with funders about this and find quicker
and smarter ways of assessing impact.
Scientists need to tell the bad news as well as the
good news and acknowledge negative aspects of biological control and any mistakes or failures (which
we need to try to learn from). This will lead to more
trust and buy-in. We need to find better ways of evaluating risk and dealing with uncertainty and make
more of an effort to celebrate success so people know
biological control can and does work and what the
benefits are. Success should be defined at the start of
projects.
383
Field release of the rust fungus Puccinia
spegazzinii to control Mikania micrantha in
India: protocols and raising awareness
K.V. Sankaran,1 K.C. Puzari,2 C.A. Ellison,3
P.S. Kumar4 and U. Dev5
Summary
Mikania weed, Mikania micrantha H.B.K., a perennial plant of neotropical origin, is a major threat to
natural and plantation forests and agricultural systems in Asia and the Pacific. In India, it is a serious
weed in the south-eastern and north-eastern states. The efficacy of herbicides to control mikania weed
is short lived, and manual weeding is labour intensive and expensive. In this context, the rust fungus
Puccinia spegazzinii de Toni, from Trinidad, shown to be highly specific and damaging to Mikania,
was assessed for its control. Following a consultation process with the Ministry of Agriculture, Government of India and other local stakeholders, the rust was imported in 2004 into the quarantine facility at the National Bureau of Plant Genetic Resources in New Delhi. After additional host-specificity
testing, field release was permitted by the Government of India in 2005. The rust was first released
in tea gardens in Assam (north-east India) in October 2005 but did not establish, most likely due to
the presence of a biotype of the weed that was partially resistant to the rust pathotype used. In Kerala
(south-west India), releases of the rust were initially made in agricultural systems in August 2006,
followed by forest sites. These releases are now considered to be successful; the rust has spread and
is persisting. This is the first instance where a fungal pathogen has been used as a biocontrol agent
against an invasive alien plant in continental Asia. An awareness-raising campaign on the merits of
biological control of invasive alien weeds, targeting the general public, farmers, policy makers, forest
officials and the scientific community, was undertaken. The range of methods, including engaging the
media, publications and demonstrations are discussed.
Keywords: invasive alien species, plantation crops, classical biological control, Kerala,
Assam.
Introduction
Mikania micrantha H.B.K. (Asteraceae), a native of
tropical and subtropical zones throughout the Americas, is a perennial, fast-growing invasive plant, capable
of smothering agroforesty and natural forest ecosystems. It also invades many crops within home gardens
and plantation production systems in the tropical moist
Kerala Forest Research Institute, Peechi 680 653, Kerala, India.
Assam Agricultural University, Jorhat 785 013, India.
3
CABI Europe-UK, Bakeham Lane, Egham, Surrey TW20 9TY, UK.
4
Project Directorate of Biological Control, PB No. 2491, H.A. Farm
Post, Bellary Road, Bangalore 560 024, India.
5
National Bureau of Plant Genetic Resources, New Delhi 110 012, India.
Corresponding author: K.V. Sankaran <[email protected]>.
CAB International 2008
1
2
forest zones of Asia and the Pacific (Waterhouse, 1994;
Global Invasive Species Database, 2002). The preferred habitat for growth of mikania weed is open areas
with moist soil. It occupies marginal lands, pastures,
roadsides, uncultivated areas, degraded forests, plantations and agricultural systems. M. micrantha was introduced into the north-eastern part of India during the
Second World War for camouflage of airfields and was
later used as a ground cover for tea plantations (Parker,
1972). Thence, it has dramatically increased its range
within India, spreading to over ten states especially in
the north-east and south-west (Sankaran et al., 2001).
Mechanical control methods of mikania like sickle
weeding, uprooting and digging are labour intensive,
expensive and not effective in the longer term. Chemical control based on herbicides, such as glyphosate,
and 2,4-D compounds, is practiced in several countries,
384
Field release of the rust fungus Puccinia spegazzinii to control Mikania micrantha in India
but the efficacy is short term, and vigorous re-growth is
observed after a few months of application (Sankaran
and Pandalai, 2004). However, the weed was considered to be an ideal candidate for classical biological
control using co-evolved natural enemies, since it is
rarely a weed in its native range, where natural enemies
limit its abundance (Cock et al., 2000).
Under a UK-Department for International Development (DfID)-funded project, fungal pathogens were
assessed for their biological control potential of M.
micrantha in India. No local pathogens were found to
be suitable in India. However, the rust fungus Puccinia
spegazzinii de Toni (Evans and Ellison, 2005) was selected from the broad range of coevolved fungal pathogens recorded from the neotropical, native range of the
plant (Barreto and Evans, 1995), as a suitable candidate
for introduction into India. This rust pathogen causes
stem, petiole and leaf infections on M. micrantha, and
11 isolates from six countries were evaluated in the
CABI Europe-UK quarantine glasshouse. The rust was
found to demonstrate intra-species specificity, each pathotype infecting only a selected number of genotypes
of its host (Ellison et al., 2004).
However, a pathotype from Trinidad (IMI 393067)
proved to be virulent against a wide range of Indian
populations of the weed, infecting all those tested from
the Western Ghats, and hence was selected for further assessment. This pathotype was screened against
65 non-target species and found to be highly specific
(infecting a limited number of species in the genus
Mikania), damaging (infection often leading to plant
death) and has a broad environmental tolerance (Ellison et al., 2008). After consultation with Indian stakeholders, permission was sought to import and release
the pathogen in mikania weed-affected areas in the
south-west (Kerala) and north-east (Assam) regions of
India. This paper focuses on the processes involved in
this and engagement with the public concerning this
novel approach to weed control.
Materials and methods
Importation of P. spegazzinii into India
Protocols: India has a relatively long history of importing classical biological control agents for the control of invasive alien weeds; however, all the natural
enemies have, thus far, been arthropods (Singh, 2001).
P. spegazzinii was the first pathogen considered by the
Indian authorities for classical biological weed control, and in fact, it was a first for mainland Asia. This
necessitated consultations between biological control
scientists, the Ministry of Agriculture, Government
of India, the Indian Council of Agricultural Research,
policy makers and other stakeholders to develop and
refine the protocols. A dossier on the rust, produced
by CABI Europe-UK for The Project Directorate for
Biological Control, Bangalore (the nodal point for im-
port of biological control agents into India), following
the Food and Agriculture Organization (FAO) Code of
Conduct (FAO, 1996; Ellison and Murphy, 2001), was
submitted to the Indian Ministry of Agriculture. This
document also included permission from the Ministry
of Agriculture, Land and Marine Resources of Trinidad
and Tobago (where the rust isolate originated) for the
use of their genetic resources, following the Convention on Biodiversity (http://www.biodiv.org/). Permission to import the rust into quarantine facility at the
National Bureau of Plant Genetic Resources (NBPGR),
New Delhi was granted in September 2002.
P. spegazzinii can only survive in living plants; once
the infected plant parts are dried, the teliospores are
rendered non-viable. In addition, the firmly embedded
teliospores do not readily survive scrapping from the
host tissue. Thus, the rust had to be shipped to India on
the living host plant. However, because high humidity
in the shipment box would cause the embedded teliospores to sporulate (produce the infective basidiospore
stage), the rust had to be shipped during the period
post-inoculation but before the teliospores were fully
viable (i.e. 210 days after inoculation). The shipment
was hand-carried to avoid potential delays that could
occur if it was sent as a cargo item. The rust was successfully established in the quarantine facility at the
NBPGR in September 2004.
Additional host specificity tests: Following consultation with systematic botanists, mycologists and plant
pathologists at Indian Council of Agricultural Research,
an additional host-specificity testing list was drawn up,
consisting of 74 test species/varieties of plants. Of these,
25 plants were closely related to the genus Mikania
(members of Asteraceae), and the rest were economically important plants collected from different parts of
India. The inoculation procedure is detailed in Evans and
Ellison (2005) and assessment procedure in Ellison et
al. (2004). For each test plant species or variety, eight
replicate plants were used.
The screening was completed by April 2005 and a
supplementary dossier (Kumar and Rabindra, 2005)
was submitted to the Ministry of Agriculture with the
application for field release of the rust. In June 2005,
the Plant Protection Advisor to the Government of India from the Ministry of Agriculture gave the permit
for release of P. spegazzinii in four identified areas, two
each in Kerala and Assam.
Mikania plants inoculated with P. spegazzinii prepared in the quarantine facility of the National Bureau
of Plant Genetic Resources were hand carried in polystyrene boxes to Assam Agricultural University in July
and September 2005 and to Kerala Forest Research Institute in November 2005 and established in purposebuilt facilities. Around 100 rust-infected plants were
produced ready for field release at selected sites. The
culture conditions under which the field inoculum
plants were produced were found to be critical. Plants
have to be pest free, therefore they are sprayed with
385
XII International Symposium on Biological Control of Weeds
a general-purpose insecticide at least 1 week before
inoculation; young, to ensure that they are growing at
their maximum rate; and with abundant meristematic
tissue since this is the most rust-susceptible part of
the plant.
Field release of P. spegazzinii in
Assam and Kerala
The release strategy in Assam and Kerala (Table 1)
involved placing large earthenware pots containing
rust-infected M. micrantha plants in strategic positions
in infestations of mikania weed. Positions were chosen
in humid, slightly shady places, in a dense stand of the
weed. Shoots of the plants in the surrounding vegetation were heavily sprayed with a fine mist of water before putting the rust-infected plants in position. Also, as
much as possible, the shoots of the surrounding vegetation were pulled underneath the infected leaves, petioles and stems of the source plants. The release sites
were regularly monitored, and progress of the rust infection recorded. Specific methodology at the different
release sites is detailed below.
Table 1.
Site
Assam: At site one, Experimental Garden for Plantation Crops (EGPC), the rust source plants were set in
the ground by excavating a hole 20 cm in diameter and
30 cm deep for each pot, separated by at least 1 m, so
the initial field infection could potentially be recorded
separately for each inoculum pot. At site two, Cinnamara Tea Estate (CTE), the pots containing the rustsource plants were hung in a dense stand of mikania
weed at the level of the tea table (flat top to rows of
tea bushes where leaves are plucked) at about 75 cm
height, in a shady place suspended from a bamboo pole
with rope (Table 1). The mikania leaves were sparse at
ground level due to shading by the tea bushes. All the
release sites were sprayed with water twice a day for
15 days, except on rainy days.
Kerala: Each pot containing three rust-infected M.
micrantha plants were placed on the soil surface within
a defined 2 2 m quadrat separated from the next quadrat by 3 m; this potentially would allow the spread of
each rust infection to be recorded separately for more
than one generation. The total number of leaves, petioles and stems infected by the rust was determined for
each site, at each date.
Details of the Puccinia spegazzinii de Toni releases in the field in Assam and Kerala (India) during 20052006.
Site details
Release dates
Average temperature
and relative humiditya
Inoculum source
Ecosystem type
Experimental
garden for plantation crops, Assam Agricultural
University
Cinnamora Tea
Estate
a. Early October 2005
b. April 2006
c. June 2006
a. 2030C, 8590%
b. Not recorded
c. Not recorded
Six pots each release separated
by at least 1 m
Mikania weed
monoculture
a. Early November
2005
b. April 2006
c. June 2006
a. 2030C, 8590%
Two groups
of three pots
separated by 4
m each release
Kerala
Site 1
Echippara,
Trichur Forest
Division
Tea plantation
(Camellia sinensis [L.]
O. Kuntze) heavily
infested with mikania
weed
a. 24 August 2006
b. 25 September 2006
a. & b. 23.329.6C,
70100%
a. 11 pots
b. Three pots
(10 m from
release a.)
Site 2
25 September 2006
2328.7C, 80100%
Three pots
Agricultural system
with mixed cropping
of coconut (Cocos nucifera L.) and areca
nut (Areca catechu L.).
On the banks of a
perennial stream with
dense canopy
Degraded moist
deciduous forest
a. 19 September 2006
See Table 3
Six pots each
release
Assam
Site 1
Site 2
Site 3
Palappilly,
Thrissur Forest
Division
Peechi, Thrissur
Forest
Division-Kerala
Forest Research
Institute (KFRI)
campus
b. Not recorded
c. Not recorded
b. 30 October 2006
During 15 day inoculation period.
386
Degraded moist
deciduous forest
Field release of the rust fungus Puccinia spegazzinii to control Mikania micrantha in India
Raising awareness: An important part of the mikania
weed classical biological control programme involved
an awareness-raising campaign on the benefits of biological control amongst the communities where the rust
was planned to be released, as well as the government
policy makers, forest officials and scientists. At Kerala Forest Research Institute, the opinion of farmers
concerning the use of host-specific natural enemies to
control invasive alien weeds rather than chemicals was
initially sought via farmer questionnaires and meetings. This was followed by demonstrations and exhibitions aimed at the agricultural and forestry extension
services and students from universities and schools. A
pre-rust-release workshop was held at Kerala Forest
Research Institute and a post-release workshop in Assam Agricultural University to educate all stakeholders
on the usefulness of biocontrol agents in controlling
invasive weeds.
The media was also engaged: local newspapers in
Kerala published articles on the release of the rust depicting the rust as a welcome solution to the weed
problem; CABI Europe submitted press releases in
the UK and India, resulting in popular articles being
published in the press and radio interviews; Kerala
Forest Research Institute in collaboration with the
Audiovisual Research Centre, University of Calicut,
produced two documentary films aimed at the general public. The first Weeds: the Biological Invaders
was telecast all over India through the National Television Network. The second, focusing on biological
control of weeds, is currently being edited prior to
broadcast.
Publications have included a popular-style book
aimed at policy makers in the developing world; Invasive Alien Plants: Problems and Solutions (in press,
CABI-Europe); and local-language brochure for Kerala farmers on the sustainable management of invasive
alien weeds.
Results and discussion
Importation of P. spegazzinii into India
Additional host specificity tests: None of the 74 plant
species inoculated was infected by P. spegazzinii,
showing that the pathogen was highly host specific.
Mild chlorotic flecks were observed on a few top
leaves of four cultivars of sunflower, but the leaves
recovered from the symptoms and there was no
sporulation of the rust. The non-susceptibility of sunflower was also ascertained through histopathological studies (Ellison et al., 2008). All the inoculated
sunflower plants showed normal growth and flowering, confirming the host specificity results from CABI
Europe-UK and establishing that P. spegazzinii can
safely be used as a classical biological control agent in
India.
Field release of P. spegazzinii in
Assam and Kerala
Assam: At site one, the Experimental Garden for Plantation Crops, infection on field M. micrantha plants
was observed 12 days after the release of the rust. Rust
pustules were observed on leaves, petioles and stems
of the surrounding mikania vegetation. However, the
pustules were smaller than those normally observed on
fully susceptible plants. The number of leaves infected
ranged from 8 to 33, and maximum number of pustules
developed on leaves ranged from 5 to 32. The number
of stems and petioles infected was low (one to three).
Following the first release, rainfall was continuous for
14 days, after which the common mikania leaf spot
pathogen, Cercospora mikaniacola F. Stevens, became
abundant on leaves. This sudden high level of Cercospora caused early senescence of the rust-affected
leaves, hence slowing the spread of the rust infection.
However, stem infections were not affected by Cercospora, so rust infection was able to progress in the field
with the inoculum released from the infected stems.
The disease progression was noted until January 2006,
after which no progress was observed due to the nonconducive environmental conditions (high temperature
and low humidity). Release of the fungus in April 2006
also resulted in good infection in the field, but the disease spread only a short distance from the inoculum
source (30 cm in 9 weeks). Unfortunately, dry conditions following the June release prevented the rust from
spreading.
At site 2, CTE, the rust infected the M. micrantha
plants surrounding the inoculum source and spread further than at site one (1 m). As with site one, the pustules
were small. However, progress of the rust infection was
curtailed as the environmental conditions became nonconducive for natural spread of the rust. A similar result was found for the 2006 inoculations as at site one.
The first inoculation at both sites was undertaken late
in the wet season, when the conditions suitable for rust
infection were already deteriorating. However, infection and spread was still achieved, but (at the chosen
release sites) the rust was unable to survive the dry season. The 2006 inoculation did not lead to a significant
level of infection and spread of the rust. This relatively
disappointing result has been attributed mainly to the
presence of a semi-resistant biotype of mikania weed
present at the release sites, demonstrated by the small
pustule size.
The evaluation of the pathogenicity of pathotypes of
the rust against biotypes of the weed was reported by
Ellison et al. (2004). This showed the presence of biotypes of mikania in Assam that were semi-resistant to
the rust pathotype from Trinidad and that the Peruvian
pathotype should be released in Assam as well. However, it was decided to proceed initially with importing the Trinidad pathotype into India, which was fully
screened and is fully pathogenic to all biotypes tested
387
XII International Symposium on Biological Control of Weeds
in Kerala (the region targeted during the first phase of
the project, when original rust selection work was undertaken) and most of those from Assam.
Kerala: At all sites, the initial symptoms of the disease
on the field M. micrantha plants were noticed a week
after release of the rust. The results are summarized in
Table 2. Similar results were observed at all plots inoculated in August and September although with varying
degrees of disease severity and a maximum distance of
spread of 1.5 m away from the rust-source plant. However, by October, the environmental conditions became
non-conducive for spread of the rust, and levels of infection gradually declined (Table 3). Inoculations carried out in late October led to low levels of infection.
By December, no rust infection was observed in the
field.
The results, in general, indicate good spread of the
rust from the source plant to field population of mikania weed in Kerala and Assam. Even though the field
inoculations were carried out late in the wet season in
Kerala, there was still also good spread within the field
population until late October. Laboratory studies with
the rust have shown that in the Western Ghats, optimum conditions conducive to rust infection (see Introduction) are likely to occur from June to September,
although the temperature can go above optimum (up
to 33C) during August and September. However, beyond this period (until April or May), the maximum
atmospheric temperature rises to over 40C, and minimum relative humidity goes down to 30%. The conditions are more or less similar in Assam where the hotter
conditions may extend until June. During this summer
period, mikania weed tends to die back in open areas,
Table 2.
The result of a farmer survey showed that over 90%
were willing to try the biocontrol agent in their farm.
The Government of India adjudged the film Weeds:
the Biological Invaders as the best documentary film
on humanity, environment and human rights, in 2004.
Communication activities were undertaken to create
awareness among the general public, forest officials,
scientists and policy makers. The Tea Research Institute
at Valparai, Tamil Nadu and plantation owners within
Kerala approached KFRI on the possible use of the
biocontrol agent to control mikania weed in their farms
and plantations. Overall, this campaign showed that it
is possible to cultivate a positive thinking on the use of
biocontrol agent against invasive weeds in India.
Conclusions
In Kerala in 2007, the aim is to significantly increase
the frequency and quantity of inoculations of the rust
Site 1
Dates of field release
24 August 2006
Days after release
No. of leaves infected
No. of pustules per leaf
No. of petioles infected
No. of stems infected
25
67
13
7
1
54
27
12
0
0
75
10
12
0
0
Site 2
25 September
2006
100
1
1
0
0
22
43
13
0
0
40
0
0
0
0
Site 3
25 September
2006
20
24
12
2
0
70
1
1
0
0
19 September
2006
30 October
2006
20
82
110
19
4
17
6
15
0
0
37
0
0
0
0
Air temperature and relative humidity at Peechi (Kerala, India) during AugustDecember 2006.
Month
August
September
October
November
December
Raising awareness of the use of CBC
Field infection of Puccinia spegazzinii de Toni on Mikania micrantha H.B.K. in Kerala, India.
Progress of the disease
in the field
Table 3.
although in areas with perennial standing water and
along permanent streams, plants continue to grow and
maintain leaves. Hence, over most of the mikaniainfested areas, the rust will not be able to perpetuate.
Evidence from the native range of M. micrantha and
from glasshouse studies suggest that the rust will survive in living stems as cankers in open areas and on
all aerial parts of plants surviving by permanent water.
These rust refuges could act as the inoculum source
to initiate the rust epidemic as the rains begin and the
mikania weed starts to reinvade.
Relative humidity %a
Air Temperature Ca
Minimum
Maximum
Minimum
Maximum
21.724.8 (23.3)
22.025.1 (23.1)
22.125.8 (23.4)
22.125.1 (23.6)
18.325.4 (22.5)
24.433.2 (29.6)
24.132.8 (28.7)
26.344.0 (35.8)
29.142.0 (37.7)
33.540.4 (36.6)
52.095.7 (70.1)
60.098.7 (79.7)
38.962.6 (51.2)
45.059.0 (50.8)
31.354.6 (41.1)
100 (100)
100 (100)
90100 (99.3)
88100 (99.0)
76100 (88.6)
Mean value in parentheses.
388
36
0
0
0
0
Field release of the rust fungus Puccinia spegazzinii to control Mikania micrantha in India
in the field during the season most favourable to its
spread (June to August), in order to build up the rust
concentrations. As with the first releases in 2006, the
selected areas will be those which encourage optimum
rust propagation, e.g. cooler sites under shade or along
the banks of perennial streams. It is suggested that once
there is a critical concentration of the rust in an area,
the infection will enter an epidemic phase.
Work is continuing in Assam to identify release sites
with populations of mikania weed that are fully susceptible to the rust, where new releases can be made. In addition, the screening of the pathotype of P. spegazzinii,
collected in Peru against a few selected plant species
closely related to M. micrantha at CABI, has suggested
that its selectivity, outside of its host species, is identical to the Trinidad strain. The Peruvian pathotype was
subsequently (2006) imported into quarantine at National Bureau of Plant Genetic Resources, New Delhi,
and additional confirmatory host-specificity screening
is near completion. Permission to release this isolate in
the field in Assam is being sought.
Awareness-raising activities will continue and will
be combined with a rust-distribution programme by
farmers and foresters, supported by the extension services, once optimum rust release strategies have been
established.
Acknowledgements
The authors are grateful to Dr J.K. Sharma, former Director and Dr R. Gnanaharan, Director, Kerala Forest
Research Institute for kind support and encouragement.
We thank Dr S.T. Murphy and Dr H.C. Evans from
CABI for reviewing the manuscript. We also thank the
officials of the Forest Department of Kerala and Cinnamora Tea Estate in Assam for their co-operation and
help, without which this study would not have been
possible. This publication is an output from a research
project funded by the United Kingdom Department for
International Development (DfID) for the benefit of developing countries (R8228 Crop Protection Research
Programme). The views expressed are not necessarily
those of DfID. The rust is held in the UK under DEFRA
licence no. PHL 182/4869.
References
Barreto, R.W. and Evans, H.C. (1995) The mycobiota of the
weed Mikania micrantha in southern Brazil with particular reference to fungal pathogens for biological control.
Mycological Research 99, 343352.
Cock, M.J.W., Ellison, C.A., Evans, H.C. and Ooi, P.A.C.
(2000) Can failure be turned into success for biological
control of mile-a-minute weed (Mikania micrantha)? In:
Spencer, N.R. (ed.) Proceedings of the X International
Symposium on Biological Control of Weeds. Bozeman,
MT, USA, pp. 155167.
Ellison, C.A. and Murphy, S.T. (2001) Dossier on: Puccinia
spegazzinii de Toni (Basdiomycetes: Uredinales) a potential biological control for Mikania micrantha Kunth ex
H.B.K. (Asteraceae) in India. CABI Europe-UK, unpublished report submitted to Government of India. 50 p.
Ellison, C.A., Evans, H.C. and Ineson, J. (2004) The significance of intraspecies pathogenicity in the selection
of a rust pathotype for the classical biological control of
Mikania micrantha (mile-a-minute weed) in Southeast
Asia. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lons
dale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings
of the XI International Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia,
pp. 102107.
Ellison, C.A., Evans, H.C., Djeddour, D.H. and Thomas, S.E.
(2008) Biology and host range of the rust fungus Puccinia
spegazzinii: A new classical biological control agent for
the invasive, alien weed Mikania micrantha in Asia. Biological Control 45, 133145.
Evans, H.C. and Ellison, C.A. (2005) The biology and taxonomy of rust fungi associated with the neotropical vine
Mikania micrantha, a major invasive weed in Asia. Mycologia 97, 935947.
FAO (1996) International Standards for Phytosanitary Measures. Code of Conduct for the Import and Release of Exotic Biological Control Agents. Rome, Italy, Secretariat of
the International Plant Protection Convention. 21 p.
Global Invasive Species Database (2002) Mikania micrantha
(land plant). Available at: http://www.issg.org/database/
species/ Ecology.asp.
Kumar, P.S. and Rabindra, R.J. (2005) Supplementary dossier
on: Puccinia spegazzinii (Basidiomycetes: Uredinales)
a potential biological control agent for Mikania micrantha H.B.K. (Asteraceae ) in India - Project Directorate of
Biological Control, Bangalore, India. Unpublished report
submitted to Government of India. 23 p.
Parker, C. (1972) The Mikania problem. PANS 18, 312315.
Sankaran, K.V., Muraleedharan, P.K. and Anitha, V. (2001)
Integrated management of the alien invasive weed Mikania micrantha in the Western Ghats. KFRI Report
No.202, Kerala Forest Research Institute, Peechi, India.
51 p.
Sankaran, K.V. and Pandalai, R.C. (2004) Field trials for controlling mikania infestation in forest plantations and natural forests in Kerala. KFRI Report No.265, Kerala Forest
Research Institute, India. 52 p.
Singh, S.P. (2001) Biological control of invasive weeds in
India. In: Sankaran, K.V., Murphy, S.T. and Evans, H.C.
(eds) Proceedings of the Workshop on Alien Weeds in
Moist Tropical Zones: Banes and Benefits. Kerala Forest
Research Institute, India and CABI Bioscience UK Centre
(Ascot), UK, pp. 1119.
Waterhouse, D.F. (1994) Biological Control of Weeds: Southeast Asian Prospects. ACIAR, Canberra, Australia. 302 p.
389
What every biocontrol researcher
should know about the public
K.D. Warner,1 J.N. McNeil2 and C. Getz3
Summary
Classical biological control is a public-interest science. This places a special responsibility on researchers and practitioners to communicate to the public about their activities and the benefits these
provide to them. More than many other forms of science practice, biological control requires understanding public perceptions of their work and a coordinated effort to communicate with the public.
Publicly addressing the risks of classical biological control introductions can foster a public consensus on appropriate risk-management strategies.
Keywords: public-interest science, risk perception, risk management, public outreach.
Introduction
The practice of classical and conservation biological
control is a public-interest science, done on behalf of
the public and generally with public funds. The kind
of knowledge produced by classical and conservation
biological control work is of a public good character,
meaning that it is non-rival and non-excludable; in
other words, it is a pure common resource. Biological control research does not result in commodifiable
knowledge (e.g. patents), and this trait distinguishes
this form of scientific activity from many others. Consequently, as products of this science are not amenable
to private property right protections, its practitioners
- usually employees of public agencies or publicly
funded universities - rely upon public funds to do their
work. Thus, practitioners in the field of biological control have a special need to understand the public and
cultivate public support for their work.
Declining public funding threatens to undermine the
institutional capacity for biological control. The discovery of some nontarget effects has led some ecologists to assert that biological control is inherently risky
and that much more precaution is necessary (Howarth,
Environmental Studies Institute, Santa Clara University, California
95053, USA.
2
Deparment of Biology, University of Western Ontario, London, Ontario, Canada, N6A 5B8.
3
Department of Environmental Science, Policy and Management, University of California, Berkeley, CA 94720 USA.
Corresponding author: K.D. Warner <[email protected]>.
CAB International 2008
1
1991; Lockwood, 1996). Thus, over the past two decades, critics, practitioners and regulators have publicly
debated norms and policies that might apply to biological control. Several countries have implemented new
regulations, prompting what some have described as
an emerging regulatory crisis (Sheppard et al., 2003).
Biological control researchers have long recognized
the importance of cultivating public trust and support,
which are critically necessary for policy support and
public funding (van Lenteren, 2004, 2006). Members
of the International Organization for Biological Control (IOBC) are advancing persuasive arguments that,
while no pest-management strategy is risk free, biological control is often the safest and most cost-effective
approach (van Lenteren, 2004; Delfosse, 2004, 2005).
These kinds of initiatives are essential to sustain public
funding and policy support for this scientific practice.
The relationships between scientists, scientific
knowledge and the public are critical - yet contested issues in the modern world and central to making progress toward a more sustainable relationship between
humanity and the biosphere. In this paper, we report
our research into the ways that public attitudes, public
communication and public initiatives affect the practice of biological control. Many social scientists have
analysed efforts to improve science communication
and policy, and thus we begin by placing our work in
this broader context. We will first review the obstacles
and challenges illustrated by prior studies that underscore the importance of the public communication of
science. We then report original data from field work
in Canada and California on the relationship between
390
What every biocontrol researcher should know about the public
several publics and biological control efforts. This paper concludes with several recommendations to guide
public initiatives on behalf of biological control.
field of Science, Technology and Society (STS) has addressed many of these kinds of questions for decades
(Gregory and Miller, 1998).
Communication, science
and the publics
What do we know about what the
public thinks about biocontrol?
Scientists often avoid speaking to lay, public audiences
and to the media, as their statements are often mis
quoted, taken out of context, clipped and distorted, sensationalized or even ridiculed (Hayes and Grossman,
2006). While these are valid reasons, our modern world
consequently suffers greatly from the broad problem of
scientific illiteracy.
However, the low public acceptance of transgenic
seeds/genetically modified organisms in the United
Kingdom should serve as a cautionary case study for
anyone with simplistic ideas about the problem of
public resistance to a novel technology. Both the
transnational corporations and government agencies involved in the development and regulation of transgenic
organisms have behaved as though the public merely
lacked information and that increasing the available
knowledge would remedy the situation. During the
1990s, as more information was conveyed to the British
public, public skepticism markedly increased. While the
promoters of this technology did recognize the problem
of scientific illiteracy, they proceeded to attack what
they perceived as emotional responses, the public misperception of risk and irresponsible media coverage.
One social scientist described their approach as an exercise in creating public alienation (Wynne, 2001).
Regardless of what one thinks of these technologies, the public outreach strategy for genetically modified organisms (GMOs) in the UK was a disaster. Many
science communication scholars could have predicted
this outcome because prior research in this field has
consistently revealed that the critical issue is not scientific illiteracy of the public but rather the uneven levels of public trust in expert scientists and the political
regulatory institutions (Rampton and Stauber, 2002).
Studies have demonstrated that more knowledge communicated to the public about a controversial scientific
practice can very easily result in amplifying rather than
diminishing public fears (Gregory and Miller, 1998).
At least in advanced industrial societies, most people
generally trust, and thus filter out, scientific knowledge
and technologies as a matter of routine. Querying the
general public about their opinions regarding a scientific practice about which they know little may provide
them with specific information that they find disturbing. Consequently, while science practitioners may use
new findings to increase the level of public knowledge,
building trust may be a more effective strategy. Therefore, any effort to communicate expert knowledge to
the public must also seriously address the need to establish credibility and foster trust. The interdisciplinary
In North America, two social science surveys have assessed public knowledge about biological control, one
in California and the other in Canada. The ash whitefly, Siphonius phillyrea (Haliday) (Homoptera: Aleyrodidae) was introduced into Californias urban landscape, causing millions of dollars of damage through
defoliation. The California Department of Food and
Agricultures Biological Control Program introduced a
parasitoid, Encarsia ianaron (Walker) (Hymenoptera:
Aphelinidae) which established and provided a highly
successful control effort (Pickett et al., 1996) that provided between 219 and 298 million US dollars in benefits to the public. However, shortly afterward, the budget of Californias overall Biological Control Program
was cut, resulting in half of the permanent scientific
staff being let go. This program was vulnerable to such
vagaries of California State funding because it did not
have a dedicated revenue stream, nor did it have sufficiently powerful political allies.
To assess potential public support for such a revenue
stream, Jetter and Paine (2004) surveyed consumers
about their economic preferences for three strategies
(chemical pesticide, biorational insecticide or introduced natural enemy) in controlling an invasive pest
of urban forest landscapes. They provided respondents
a booklet with background information on these pestmanagement options and asked urban homeowners to
report their relative willingness to pay for them. The
findings suggested that social and financial support by
urban residents could be tapped to fund the introduction of classical biological control agents for landscape
pests.
Under the auspices of the Canadian Biological Control Network, several Canadian researchers conducted
a Canada-wide telephone survey in 2005 to determine
public perception of biological control as an alternative
to the use of traditional pesticides and GMOs (McNeil,
personal communication). Here the thrust was to assess
the perceived risks of these three pest-management options. Although the data from this survey are still being
analysed, initial findings indicate that Canadians generally consider biological control to be safer than conventional agrochemical pesticides in agriculture. The
findings also indicate that Canadians, especially those
of middle age, would like more information about pestmanagement strategies used in their food production, a
finding consistent with other surveys about how much
information consumers would like about the conditions
of their food production (Eilenberg and Hokkanen,
2006).
391
XII International Symposium on Biological Control of Weeds
A fundamental limitation of these types of surveys
is that they assess public opinion, not actual behavior.
Neither of these surveys query who among the public
would actively participate in deliberations about the
importation of biological control agents or funding for
such initiatives if the opportunity arose. However, in
the USA and Canada, there is no formal process for soliciting public input into biological control importation
decisions (Mason et al., 2005). Thus, other participants
in the administrative policy process must represent the
public interest.
Identifying clients for
biological control
Warner and Getz have been conducting a study since
2004 of the social and economic factors in California
serving as obstacles and opportunities for further implementing biological control in agriculture. Our work
reveals the critical role of two institutions that recognize the value of biological control: county agricultural
commissioners and growers groups. Agricultural commissioners are appointed by locally elected county officials, and thus they are responsive to local residents.
Commodity board research directors are hired by organizations of farmers, and they understand their role to
be supporting the kind of research that will serve the
needs of their farmers. Interviews indicate that both
play a critical intermediary role for county residents/
taxpayers and large groups of growers, which are two
groups interested in pest management. The dynamic
works like this: when an urban landowner or a farmer
of a specific commodity has a problem with a new pest,
s/he contacts the agricultural commissioner or commodity board research director, who, in turn, contacts
a university specialist or state program researcher. The
commissioners and research directors function as key
intermediaries, or knowledge brokers, from land managers to the knowledgeable expert, who then, if appropriate, conduct research and share information in turn
with these intermediaries. This arrangement works well,
as long as the research institutions are properly funded,
but as the section above narrates, this is not presently
the case. Scientific knowledge is circulating through
this system, but because political knowledge about the
benefits does not, financial resources are erratic.
The evolution of environmental regulations in California has made conventional insecticides increasingly
more difficult to use. Consequently, both urban landowners and farmers are generally open to considering
biological control if it proves effective and affordable.
However, even these groups, who are able to identify
the value of university and state biocontrol programs,
know very little of the budgetary constraints faced by
the scientists carrying out the research. An opinion survey of land owners and farmers would likely indicate
support for these researchers (and any other expert
knowledge that could help them) but does not automatically follow that this publics interest will be transmitted to government funders.
Our interviews with agricultural commissioners and
commodity board research directors indicate that they
are quite aware of, and concerned about, the diminishing institutional capacity of the University of California
and the state Biological Control Program. Their professional responsibilities include helping (urban and
agricultural) land managers control pests and ensure
they conform to environmental regulations. Research
directors are particularly concerned that the number
of scientists conducting practical research in biological control has declined significantly over the past few
decades. One noted that she could provide funding for
any genuine biological control proposal that had the
potential to advance knowledge of that crops farming
systems but that the number of researchers in the field
has diminished significantly. Another, representing a
major crop in the state, said that there was only one
scientist in California that could help him with one of
his major pests.
The agricultural commissioners who are often the
first officials to receive a phone call from a distressed
landowner have legislatively mandated responsibilities
for protecting their county from noxious weeds and insects and also for enforcing state pesticide laws. Thus,
they too experience the tension of having to coordinate
pest-management efforts but within the limitation of
existing laws. Consequently, they are among the most
active consumers of the research knowledge and the
biological control agents provided by California Department of Food and Agriculture (CDFA)s Biological Control Program. They advocate for funding this
program, but they are somewhat constrained as their
own county activities depend on the State Secretary of
Agriculture for funding.
One particularly noteworthy institutional vehicle for
building public support for funding has been the California Weed Management Area Support Program (California Department of Food and Agriculture, 2006).
With relatively modest state funding, this program has
fostered local networks of concerned landowners and
agencies to focus attention on noxious weeds. It has
leveraged US $5.4 million of state money to attract
over US $7 million of additional funds, but more important has been its ability to provide a vehicle for local
landowners to coordinate their efforts and educate the
public. Weed-management areas (WMAs) provide the
social infrastructure to cooperate in a meaningful way
with the states Biological Control Program. The current WMAs rest on a long history of coordinated pest
management in California (Baker, 1988; Warner, 2007).
They are essential for coordinating widely dispersed
weed-management activities but also for activating existing social networks to advocate for continued funding. Members of the public who have benefited from
coordinated weed eradication are much more likely
392
What every biocontrol researcher should know about the public
to provide the political support for biological control.
One landowner who has benefited by such a program is
much more likely to take action in support of biological control than one million consumers who express a
favorable opinion on a mass survey.
Public initiatives on behalf of
biological control
From the observations above, it seems that biocontrol
is a terrific pest-management strategy but that it faces
worrying trends. Science funding and policy in the industrial countries now operate in a new, more challenging political context. No scientists can count on stable
funding or policy work. This is particularly true of a
public-interest science and one that must now confront
the controversies of non-target impacts. We therefore
recommend that the biocontrol community develop
strategic alliances with several potential publics. Developing coordinated science communication policy
requires more work but few financial resources.
Biological control has five publics, with different
communication needs: client communities (land managers and farmers); funders (interested in pest management and invasive species); regulators; potential
scientific allies (public-interest ecologists interested in
pest-management alternatives); and contrarians (scientists who are philosophically opposed to biological control). Communicating with these five requires
a carefully targeted message. An example of this kind
of linking of research and public communication can
be found in the Ecological Society of Americas Sustainable Biosphere Initiative (SBI) presented by Lubchenco et al. (1991), described by Lubchenco (1998)
and analysed by FitzSimmons (2004). The SBI has
developed research briefs for policy makers through
their Issues In Ecology publications, and this kind of
effort could be copied by the biocontrol community.
A biocontrol science communication working group,
perhaps coordinated by the IOBC, could engage policy
makers by presenting the value that this science could
make to managing invasive species and pests.
Conclusions
Conclusions and recommendations in this paper include:
1. The public-interest character of biological control
requires ongoing initiatives to cultivate public and
governmental support. The work of IOBC global
and regional sections is to be commended. We recommend that IOBC cultivate the help of social scientists, especially those who study STS and science
communication, to strategize a coordinated and sustained effort to engage the public in its many forms.
2. Identify and partner with institutions most likely
to benefit from your work, and encourage them to
represent the public value of biological control in
the public sphere. The commodity organizations and
WMAs are excellent examples of how this kind of
partnering can bear fruit.
3. Strategic analysis suggests that favouring outreach
efforts which cultivate partners and clients who can
serve as credible messengers to public policy makers and funders is likely to be more fruitful than efforts to outreach to generic public masses.
4. Recognize that most publics evaluate your work
not on the basis of scientific knowledge or its merit
but rather on trust. Build trust through collaboration with credible partners, and be transparent about
risks, risk-management efforts, peer review and the
value of client participation in this work.
5. Do this as a group, through institutions and networks,
to justify biological control for its public good features (not because of original scientific discovery).
Emphasizing the public good features of biocontrol
can convert passive acceptance into active advocates for funding and supportive policy. Partners can
serve as credible messengers, conferring legitimacy
to your work and increasing the likelihood of funding for providing pesticide alternatives and strategies for controlling invasive species.
Acknowledgements
The authors gratefully acknowledge support from
the California Department of Food and Agriculture,
Santa Clara Universitys Center for Science, Technology and Society and Food and Agribusiness Institute,
the US National Science Foundation, and the Canadian
Biological Control Network. Dustin Mulvaney offered
helpful comments on an earlier version of this article.
References
Baker, B. (1988) Pest control in the public interest: crop protection in California. UCLA Journal of Environmental
Law 8 (1), 3171.
California Department of Food and Agriculture (2006) Noxious Weed Management Area Support Program Final
Report. CDFA Integrated Pest Control Branch, Sacramento. Available at: <http://www.cdfa.ca.gov/phpps/ipc/
weedmgtareas/wma_sb1740_final.pdf>.
Delfosse, E.S. (2004) Introduction, In: Coombs, E.M., Clark,
J.K., Piper, G.L. and Cofrancesco Jr., A.F. (eds) Biological Control of Invasive Plants in the United States. Oregon State University Press, Corvallis, pp. 111.
Delfosse, E.S. (2005) Risk and ethics in biological control.
Biological Control 35, 319329.
Eilenberg, J. and Hokkanen H.M.T. (eds) (2006) An Ecological and Societal Approach to Biological Control. Springer,
Dordrecht, The Netherlands. 322 p.
FitzSimmons, M. (2004) Engaging ecologies, In: Cloke, P.,
Crang P. and Goodwin, M. (eds) Envisioning Human Geographies. Edward Arnold, London, pp. 3047.
393
XII International Symposium on Biological Control of Weeds
Gregory, J. and Miller, S. (1998) Science in Public: Communication, Culture and Credibility. Basic Books, Cambridge. 294 p.
Hayes, R. and Grossman, D. (2006) A Scientists Guide to
Talking with the Media: Practical Advice from the Union
of Concerned Scientists. Rutgers University Press, Piscataway, New Jersey. 220 p.
Howarth, F.G. (1991) Environmental impacts of classical
biological control. Annual Review of Entomology 36,
34570.
Jetter, K. and Paine, T.D. (2004) Consumer preferences and
willingness to pay for biological control in the urban landscape. Biological Control 30, 312322.
Lockwood, J.A. (1996) The ethics of biological control: understanding the moral implications of our most powerful
ecological technology. Agriculture and Human Values 13
(1), 219.
Lubchenco, J. (1998) Entering the century of the environment: a new social contract for science. Science 279
(5350), 491497.
Lubchenco, J., Olson, A.M., Brubaker, L.B., Carpenter, S.R.,
Holland, M.M., Hubbell, S.P., Levin, S.A., MacMahon,
J.A., Matson, P.A., Melillo, J.M., Mooney, H.A., Peterson, C.H., Pulliam, H.R., Real, L.A., Regal, P.J. and
Risse, P.G. (1991) The Sustainable Biosphere Initiative:
an ecological research agenda: a report from the Ecological Society of America. Ecology 72, 371412.
Mason, P.G., Flanders, R.G. and Arrendondo-Bernal, H.A.
(2005) How can legislation facilitate the use of biological
control of arthropods in North America? In: Hoddle, M.S.
(ed.) Proceedings of the Second International Symposium
on the Biological Control of Arthropods. Davos, Switzerland, pp. 701713.
Pickett, C.H., Ball, J.C., Casanave, K.C., Klonsky, K., Jetter,
K., Bezark, L.G. and Schoenig, S.E. (1996) Establishment
of the ash whitefly parasitoid Encarsia inaron (Walker)
and its economic benefit to ornamental trees in California.
Biological Control 6, 260272.
Rampton, S. and Stauber, J. (2002) Trust Us Were Experts:
How Industry Manipulates Science and Gambles with
Your Future. Tarcher, New York. 368 p.
Sheppard, A.W., Hill, R., DeClerck-Floate, F., McClay,
A., Olckers, T., Quimy Jr. P.C. and Zimmermann, H.G.
(2003) A global review of riskbenefitcost analysis for
the introduction of classical biological control agents
against weeds: a crisis in the making? Biocontrol News
and Information 24 (4), 91N108N.
van Lenteren, J. (2004) Biological control: sound, safe and
sustainable. Presidential Address at the General Assembly
of the International Organization of Biological Control,
Brisbane, Australia.
van Lenteren, J. (ed.) (2006) IOBC Internet Book of Biological Control. Wageningen, The Netherlands. Available at:
<www.iobc-global.org>.
Warner, K.D. (2007) Agroecology in Action: Extending Alternative Agriculture Through Social Networks. MIT Press,
Cambridge. 273 p.
Wynne, B. (2001) Creating public alienation: expert cultures
of risk and ethics on GMOs. Science as Culture 10 (4),
445481.
394
Abstracts: Theme 5 Regulations and Public Awareness
Is the Code of Best Practices helping to make
biological control of weeds less risky?
J. Balciunas1 and E.M. Coombs2
1
USDA-ARS Exotic and Invasive Weed Lab, 800 Buchanan St, Albany, CA 94710, USA
2
Oregon Department of Agriculture, 635 Capital St. NE, Salem, OR 97310, USA
Although practitioners know that biological control is one of the safest approaches to managing invasive species, they also realize that this tool must be used wisely and appropriately. The sub-discipline of
biological control of weeds was the first to acknowledge the need for universal standards to both guide
those practicing weed biocontrol and to allow outside observers to better discriminate good biological
control of weed practices from those that are ill-advised. In 1999, at the close of the Xth Biological Control of Weeds Symposium in Bozeman, Montana, the delegates overwhelmingly adopted a resolution
requesting that practitioners adhere to the 12 guidelines of the Code. We review the 12 guidelines and
how they apply to both groups that are involved in biological control: (1) the scientists performing the
research necessary to find, release and establish the biocontrol agents and (2) those who re-distribute
established agents. We also discuss the results of a recent survey we conducted to assess the impact
the Code was having on biological control of weeds in north-western USA. Finally, we provide some
examples of how the Code has been implemented by various agencies in this region and ponder on
why it seems to be having less impact elsewhere.
The new quarantine facility, St. Paul, MN, USA
R.L. Becker,1 D.W. Ragsdale,2 D. Sreenivasam,3 J. Heil,3 Z. Wu,4 M. Hanks,4
E.J.S. Katovich1 and L.C. Skinner5
Department of Agronomy and Plant Genetics, University of Minnesota, St. Paul, MN, USA
2
Department of Entomology, University of Minnesota, St. Paul, MN, USA
3
Minnesota Department of Agriculture, St. Paul, MN, USA (Retired)
4
Agricultural Resources Management and Development Division,
Minnesota Department of Agriculture, St. Paul, MN, USA
5
Minnesota Department of Natural Resources, St. Paul, MN, USA
The University of Minnesota Agricultural Experiment Station and the Minnesota Department of Agriculture combined resources to construct a state-of-the-art quarantine facility, which began operation
in 2003. Drs David Ragsdale, University of Minnesota, and Dharma Sreenivasam, Minnesota Department of Agriculture, were instrumental in obtaining the funds and commitment for this biosafety level
2 (BL2) facility. BL3 capabilities are being added with potential initial targets including Asian soybean
rust (Phakopsora pachyrhizi) research. The first invasive plant biological control effort at this facility
is an ongoing cooperative effort with Centre for Agriculture and Biosciences, International (CABI)
Bioscience, Delmont, Switzerland, to complete host-specificity screening for Ceutorhynchus spp. for
biological control of garlic mustard (Alliaria petiolata). Current research at this facility involves studies in taxonomy, genetics, life history, host specificity, behavior, control efficacy, experimental release
and post-release evaluation of plant and insect biological control agents.
CAB International 2008
395
XII International Symposium on Biological Control of Weeds
Biological control of weeds at the USDA-ARS-SABCL
in Argentina: history and current program
J.A. Briano
USDA-ARS-South American Biological Control Laboratory, Bolivar 1559,
B1686EFA Hurlingham, Buenos Aires, Argentina
The South American Biological Control Laboratory (SABCL) has a long and successful history in the
biological control of weeds. Since its establishment in Argentina in 1962, the SABCL has worked with
29 target weeds and more than 110 biocontrol candidates; 15 were field released in many countries
around the world, while nine are still in quarantine for further testing. Most of the weeds investigated at
SABCL are invasive species in the USA, Australia, South Africa and other countries. The first SABCL
projects were alligator weed (Alternanthera philoxeroides) and water hyacinth (Eichhornia crassipes).
A total of 13 natural enemies were studied against these two weeds from 1962 to 1980, nine of which
were field released. Currently, 56% of the SABCL scientific staff (n = 16) is assigned to weed research
on the following targets: water hyacinth, alligator weed, fanwort (Cabomba spp.), Brazilian peppertree
(Schinus terebenthifolius), balloon vine (Cardiospermum grandiflorum), pompom weed (Campuloclinium macrocephalum), Barbados gooseberry (Pereskia aculeata), Brazilian waterweed (Egeria densa),
water primrose (Ludwigia hexapetala) and Lippia (Phyla canescens). Research on this weed program
is funded by the US Department of Agriculture, the Commonwealth Scientific and Industrial Research
Organisation (CSIRO), Australia, and the Plant Protection Research Institute (PPRI), South Africa.
A quarter of a century of contributions from the
FDWSRU in biological control of weeds
W.L. Bruckart, D.K. Berner and D.G. Luster
USDA-ARS, Foreign Disease-Weed Science Research Unit, 1301 Ditto Ave.,
Ft. Detrick, MD 21702, USA
Evaluation of foreign plant pathogens for biological control of weeds was initiated at the United States
Department of Agriculture, Agricultural Research Service (ARS) in the mid 1970s. Justification for
locating this research effort at the Foreign Disease-Weed Science Research Unit (FDWSRU), Ft. Detrick, is a containment greenhouse facility that enables evaluation of exotic pathogens of crop plants and
weeds. Since transfer to ARS, three foreign weed pathogens evaluated in containment have been introduced into the USA under permit from federal and state regulatory organizations. These pathogens,
all rust fungi, are: Puccinia chondrillina, Puccinia carduorum and Puccinia jaceae var. solstitialis.
The program at FDWSRU has since expanded to include 2.5 research scientists with full technical
support. A number of new projects have been initiated, including rust fungi and facultative saprophytes
on Salsola tragus (two pathogens), Acroptilon repens (two pathogens) and Crupina vulgaris (two
pathogens). A new thrust into the use of floral smut fungi on Silybum marianum and Carduus thistles is
being pursued as well. Several new pathogens also have been discovered in Greece, Hungary, Russia,
Tunisia, Turkey and the USA. This paper is a review of developments, accomplishments and current
and anticipated research at the FDWSRU.
396
Abstracts: Theme 5 Regulations and Public Awareness
Protocol for projects on classical biological
control of weeds with insects
G. Campobasso and G. Terragitti
USDA-ARS-European Biological Control Laboratory Rome Station, Via Colle Trugli 9, 00132 Rome, Italy
A protocol for projects on classical biological control with insects was developed by European Biological Control Laboratory (EBCL) staff to help biocontrol workers and biocontrol consumers. The
general problem of finding, evaluating and eventually introducing a biological control agent is a complex process formed by unequal parts: politics, administration and science. All process includes eight
phases where several agencies at different levels, federal, state and University, are involved. First
phase involves information on targets; second phase, selection of potential agents; third phase, overseas explorations, collections and research; fourth phase, importation into USA for further study and
evaluation; fifth phase, release and establishment in the USA and elsewhere; sixth phase, evaluation
in USA of agents released involving State, University Region, and County authority; seventh phase,
evaluation in USA, ARS State agencies methodology of development and rearing of selected agents
to be released; and eighth phase, evaluation study in USA ARS State biological and ecological aspects
such as biology, economic impact and competition study.
Weed biological control evaluation process in the
United States - past and present
A.F. Cofrancesco, Jr
US Army Corps of Engineers, Engineering Research and Development Center,
3909 Halls Ferry Road, Vicksburg, Mississippi 39180, USA
Starting in 1957, the United States Department of Agriculture (USDA) implemented outside agency
reviews to evaluate the introduction of weed biological control agents. A group of five agencies were
identified as the Subcommittee on Biological Control of Weeds and ascertained if targeted plants were
weeds and identified non-target test plants that should be examined for host specificity. During the
1960s, the group expanded and began to exchange information between the USA and Canada. In
1971, the subcommittee became the Working Group on Biological Control of Weeds (WGBCW) with
the addition of four agencies. Also, information exchange began with Mexico. In 1987, the group
was revised, becoming the Technical Advisory Group (TAG) of Biological Control Agents of Weeds.
It is currently composed of 17 organizations and Canada and Mexico officials and provides guidance and recommendations to researchers and regulators. From 1987 through 2005, the TAG reviewed
153 petitions, with 105 petitions requesting the release of biological control agents. Recent examinations have identified that 57% of the first-release petitions received a favourable recommendation.
Through the re-evaluation process, eventually over 75% of the agents that were initially requested for
release received favourable release recommendations.
397
XII International Symposium on Biological Control of Weeds
Biocontrol capacity of ARS research group in
Central Asia and surrounding areas
R.V. Jashenko1 and C.J. DeLoach2
Tethys Scientific Society, Institute of Zoology, 93 Al-Farabi St., Almaty, 050060, Kazakhstan
2
United States Department of Agriculture, Agricultural Research Service, Grassland,
Soil and Water Research Laboratory, 808 E. Blackland Road, Temple, TX 76502, USA
The Kazakhstan biocontrol research group was organized as an ARS cooperator in 1994. Now, the
station consists of five entomologists, two botanists, one soil scientist, one Geographic Information
System (GIS) specialist and several technicians and equipped by field and laboratory equipment. The
capacity of biocontrol research is based on the native distribution of many Central Asian plants that
are weeds in western USA and Canada: 36 weed species are native to Central Asia such as perennial pepperweed (Lepidium latifolium), Russian thistle (Salsola spp.), Russian knapweed (Acroptilon
repens), yellow starthistle (Centaurea solstitialis), medusahead (Taeniatherum caput - medusae) and
other weeds, as well as for several serious introduced insect pests. The Almaty, Kazakhstan station is
well situated to conduct explorations for control agents for many weeds. The close proximity of these
weeds to the Almaty station allows for inexpensive, season-long studies of field ecology, behaviour,
host-range observations in the field, and no-cage formal testing which cannot be done in the USA and
which provide the most realistic evaluation of these critical factors. Good relations between Kazakh
stan and Russia and other Central Asian countries and a common language (Russian) and cultural
similarities allows open travel and free scientific exchanges with these countries unavailable to most
western scientists.
USDA-ARS Australian Biological Control Laboratory
M.F. Purcell, A.D. Wright, J. Makinson, R. Zonneveld, B. Brown,
D. Mira and G.W. Fichera
CSIRO Entomology, USDA-ARS-OIRP, Australian Biological Control Laboratory,
120 Meiers Rd. Indooroopilly, Queensland, Australia 4068
The staff of the United States Department of Agriculture (USDA), Agricultural Research Service
(ARS), Australian Biological Control Laboratory (ABCL) actively search the natural areas of Australia
and Southeast Asia for insects and other organisms that feed on pest insects and plant species that are
invasive in the USA. Based in Brisbane, Queensland, the ABCL is operated by the USDA-ARS Office
of International Research Programs (OIRP) through a cooperative agreement with the Commonwealth
Scientific and Industrial Research Organization (CSIRO). Many invasive weeds in the USA such as
the broad-leaved paperbark tree, Melaleuca quinquenervia; Old World climbing fern, Lygodium microphyllum; hydrilla, Hydrilla verticillata; and Australian pine, Casuarina spp. are native to Australia.
However, the native range of many of the weed species continues northward into tropical and subtropical Southeast Asia. With excellent collaborators in this region, ABCL has the capability to find the most
promising biological control agents. Research conducted at ABCL includes determination of the native
distribution of a weed species, exploration for natural enemies, molecular typing of herbivores, ecology of the agents and their weed hosts, field host-range surveys and ultimately preliminary host-range
screening of candidate agents. In collaboration with US-based ARS scientists, agents are selected for
further quarantine studies and possible release in the USA.
398
Abstracts: Theme 5 Regulations and Public Awareness
Status of biological control in Australia,
policy and regulatory influences
J.K. Scott
CSIRO Entomology, Private Bag 5, PO Wembley, W.A. 6913, Australia
Currently, biological control of weeds in Australia is experiencing a reduction in the number of newly
nominated targets and a reduction in the number of agents released from ten per year a decade ago to
two per year since 2000. At the same time, there has been a reduction in scientific activity as measured
by publications and loss and non-replacement of experienced staff specializing in biological control.
This is despite increasing recognition of the threats posed by invasive weeds where biological control
is often the only suitable long-term solution. There are, however, indications that the situation is improving. This presentation will examine the status of biological control in Australia and the influence of
legislation, regulation, infrastructure, national committees and funding policy on future developments.
The future of biological control will also be examined in the context of a developing environment of
acceptance of alien invasive species (new ecosystems).
399
This page intentionally left blank
Theme 6:
Evolutionary Processes
Session Chair: Ruth Hufbauer
401
This page intentionally left blank
Keynote Presenter
The primacy of evolution in
biological control
G. Roderick1 and M. Navajas2
Summary
Evolutionary biology underlies much of the theory and practice of classical biological control; yet,
its importance remains largely unappreciated. Procedures of classical biological control, including
agent selection, quarantine, pre- and post-release studies, establishment, agenttarget interactions,
non-target effects and even risk analysis, all involve, to a greater or lesser extent, evolutionary issues
and all can likely be improved by a better understanding of evolutionary processes. In this paper, we
examine these processes, particularly adaptation and genetic variability. We also point to promising
emerging areas in evolutionary biology, population genetics and related fields that can better inform
biological control. These include DNA barcoding, whole genome sequencing, distributed databases
and new computational approaches. Despite their promise, evolutionary studies in the context of biological control can be difficult: Evolutionary studies typically are studied over multiple generations
and often the ideal experimental protocols are logistically complex. To address these problems, we
draw parallels to invasion biology and emphasize the need for long-term, follow-up studies, even
when biological control is not successful.
Keywords: adaptation, genetic variation, evolution, selection.
Introduction
A central assumption of classical biological control
is that predators, parasites, pathogens and herbivores
will maintain their affinity for the target pest(s) while
adapting to exploit their new habitats (see Simberloff
and Stiling, 1996). Yet, adapting to non-target hosts is
not desirable and frequently not acceptable. Adaptation is an evolutionary process, and the extent to which
adaptation actually occurs will determine not only the
success of establishment and control of the target species but also the extent of non-target effects. To predict potential non-target effects, researchers often use
information concerning the phylogenetic relatedness
of targets and non-targets (see Meyer et al., 2008, this
University of California, Environmental Science Policy and Management (ESPM), 137 Mulford Hall MC 3114, Berkeley, CA 94720,
USA.
2
Institut national de la recherche agronomique, Centre de Biologie et
Gestion des PopulationsINRA, CBGP, UMR 1062, Campus International de Baillarguet, CS 30 016, 34988 Montferrier sur Lez cedex,
France.
Corresponding author: G. Roderick <[email protected]>.
CAB International 2008
1
proceedings). Evolutionary processes in biological
control are not limited to non-target effects. In fact,
evolutionary processes can inform most procedures
in classical biological control, from agent selection,
to quarantine and pre-release studies, to establishment
and other agenttarget interactions (Table 1). Evolutionary predictions can also assist in costbenefit and
risk analyses. However, few biological control practitioners would consider conducting evolutionary studies
to improve a biological control programme. Why is this
the case?
There are a number of reasons why the importance
of evolutionary processes in classical biological control has not been recognized. Evolution can be defined
as a genetic change from one generation to the next.
As such, many studies of evolutionary processes can
take a long time or longer than can be accommodated
within a funding cycle for a typical biological control
programme. Fortunately, some processes, such as the
intensity of natural or artificial selection or heritability,
can be estimated within a generation. A second difficulty in understanding the role of evolution is the difficulty of conducting experimental tests of the phenomenon. For example, to assess whether an organism has
403
XII International Symposium on Biological Control of Weeds
Table 1.
Evolutionary considerations in the procedures associated with classical biological control and the themes in this
proceedings that relate to these procedures.
Procedure
Agent selection
ISBCW Themea
5
Evolutionary concepts
Phylogenetic hypotheses
Quarantine
Population bottlenecks
Pre-release studies
Identification and diagnostics
Establishment
Genetic variation, effective
population size, adaptation
Post-release studies
Identification and diagnostics
Performance
6,7
Selection, adaptation
Agent-target interactions
6,7
Selection, adaptation,
co-evolution
Non-target effects
3,4,8
Selection, adaptation
Risk analysis
3,4,8
Selection, adaptation
Types of studies
Using relatedness in choice of agents and
predicted host use
Avoiding population bottlenecks through
sampling design and out-crossing
DNA-based species identification (DNA
barcoding); determining population origins; tracking individuals in experiments
Potentially maximizing genetic variation
through maintaining large effective population sizes and out-crossing; understanding role of adaptation to new conditions
DNA-based species identification (DNA
barcoding); determining population origins; tracking individuals in experiments
Evolutionary response to selection;
estimating heritability for relevant traits,
such as host use
Evolutionary response to selection;
estimating heritability for relevant traits,
such as host use
Estimating and predicting response
selection, or lack thereof, for physiological
tolerance or use of novel hosts
Predicting response to selection
ISBCW Themes: 1 ecology and modelling, 3 benefitriskcost analysis, 4 regulations and public awareness, 5 target and agent selection,
6 pre-release, specificity and efficacy testing, 7 release activities, 8 management specifics.
adapted to a novel host in a new habitat, ideally, one
would like to conduct a reciprocal transplant involving both the source populations and introduced populations. With a few exceptions (e.g. Hufbauer, 2002),
such experiments are typically not feasible. Nevertheless, useful information can be gained from less elaborate designs than reciprocal transplants, and much can
be learned from so-called natural experiments. Finally,
even when manipulations are possible, some field conditions are not easily replicated under laboratory or
controlled settings. There is no easy solution to this
problem. However, as discussed below, each classical
biological control programme is an experiment in itself
and follow-up studies or long-term monitoring can be
used to analyze what works and what does not.
Adaptation
Adaptation by biological control agents to new habitats
not only can increase establishment and success but also
can lead to undesired non-target effects (Hufbauer and
Roderick, 2005). Adaptation is an evolutionary process
caused by natural or artificial selection and is relatively
simple to measure in the laboratory or field. One way
to measure selection is by estimating the selection differential, S, which is the difference of the average phenotype of organisms after and before a selective event,
such as the performance on a new host plant compared
to an existing host. Estimates of selection differentials
can be compared across studies and systems. To estimate
whether a trait has a genetic basis, one can measure
also the response to selection, R, which is the difference in the average phenotype in the generation after
selection and the previous generation. The ratio of R/S
is the heritability, h2, which is also an estimate of the
proportion of phenotypic variation that has a genetic
basis. There are other ways to measure heritability and
other genetic parameters using slightly more elaborate
experimental designs, such as by following related cohorts relative to their parents (parentoffspring regression) or by splitting families to estimate maternal effects
(1/2-sib design).
Heritability can be used to measure and predict evolutionary responses in the field. For example, Cotter
and Edwards (2006) raised 1/2-sib families of the moth,
Helicoverpa armigera (Hbner), on resistant and susceptible chickpeas and found that host use was highly
heritable, but the response also depended on the larval
stage (instar) and host resistance. Even when breeding
studies are not possible, one can test for evolutionary
change. For example, Zangerl and Berenbaum (2005)
used herbarium specimens over a span of 150 years
to show that the phytochemistry of the invasive wild
parsnip changed after introduction of the parsnip web-
404
The primacy of evolution in biological control
worm. Indeed, many studies of herbivorous insects and
plants have demonstrated evolutionary responses associated with colonization as a result of biological invasions (Mller-Scharer, 2006; Strauss et al., 2006), and
the same approaches can be used in studies of biological control. Likely, the most opportune time to conduct
such studies is in the course of rearing for quarantine
or pre-release testing. However, one can also estimate
evolutionary responses over longer periods, such as associated with long-term monitoring.
life-history traits, including reproductive rate and sex
ratio at low population densities, can protect populations from Allee effects (Fauverguee et al., 2007).
Genetic variability
1. Theory and empirical evidence from studies of biological invasions suggest that adaptation to old and
new hosts and to new local environmental conditions should be common. However, few data are
available to test this notion in biological control
programmes.
2. Lack of genetic variation has been shown to have
little impact on the success of introduced species
and is not likely to be limiting in many biological
control programs. This prediction is also very testable and very relevant to developing strategies for
agent sampling, quarantine studies and release strategies.
3. Micro- and macro-organisms used for classical
biological control may differ fundamentally in the
extent to which adaptive change is important. For example, with their increased reproductive rate relative
to their hosts, microorganisms adapt more quickly
to novel hosts compared to macro-organisms. Studies to date of plant pathogens appear to support this
notion (Roderick and Navajas, 2003).
Genetic variability is the raw material on which selection acts and so a reduction of genetic variability,
such as in a population bottleneck, can result in a reduction in fitness as a result of inbreeding and a reduction in the ability of organisms to respond to new
environmental conditions. In theory, issues related to
population size and accompanying genetic variability
are important in all aspects of biological control, from
the initial collections and quarantine populations, to the
individuals released and the potential for adaptation and
non-target effects. Surprisingly, although introduced
populations typically have lower genetic variability
than their source populations (but see, Kolbe et al.,
2004; Marrs et al., 2008), there is less evidence to show
that genetic variation limits population growth in introduced populations (Roderick and Navajas, 2003;
Hufbauer and Roderick, 2005). Several factors may
explain this apparent conundrum. First, it may be that
we have not observed the species or populations that
were not successful as a result of low genetic variability, perhaps because they died out before observations
were possible. Second, theoretical studies show that, although founding populations do lose alleles, particularly the rarer alleles, if populations can rebound quickly
after introduction, the loss of overall genetic variability
(measured as heterozygosity) can be minimized (Nei
et al., 1975). Finally, a series of explanations have been
proposed to explain how small founding populations
may recover genetic variation through genetic mechanisms, such as through conversion of epistatic variation
(see Carson, 1990), greater effects of sex-linked genes
(Whitlock and Wade, 1995), founder-flush phenomena where genetic drift is weaker in growing populations (Slatkin, 1996) and multiple introductions (Kolbe
et al., 2004). These effects have not yet been studied in
biological control situations.
The effects of low population size may be both ecological and genetic. For example, experimental studies have
shown that the probability of population establishment
increases with release size (Grevstad, 1999). The Allee
effect, which is a decline in population growth associ
ated with low population size (Stephens and Sutherland,
1999), has been often evoked to explain this. However,
recent experimental manipulation of initial densities of
an invading parasitoid have shown that a number of
What can one predict?
Taken together, studies of adaptation and genetic
variability in small, introduced populations can be used
to make several predictions about changes that might
be expected in biological control agents after introduction to a new environment.
Methods and results
The methods of evolutionary biology include observations, manipulations, breeding studies and inferences
based on genetic variability. Several texts in evolutionary biology and population genetics discuss these
methodologies in detail (Table 2). In this paper, we focus on four advances that have been made in related
fields, not necessarily directly connected to assessment
of evolutionary change, but from which future studies of the role of evolution in biological control will
clearly benefit.
DNA barcoding
Many applications in biology, including biological
control, require accurate and timely species identification (Navajas and Roderick, 2008). As species are generally thought to be interbreeding units separated from
other such units, individuals within species will be
more alike genetically than individuals of different species. This is the basis for an emerging tool called DNA
barcoding, in which a small section of DNA can be used
for species diagnostics (Savolainen et al., 2005). For
many taxa, this approach works extremely well, e.g. the
405
XII International Symposium on Biological Control of Weeds
Table 2.
An introduction to the literature in evolutionary biology and population genetics relevant to classical biological
control.
Topic
Information
Evolutionary biology textbook College text in evolutionary
biology
Population genetics textbooks Approachable population genetics
texts geared toward applications
Phylogenetic methods
Advancing methods in phylogenetics, coalescence, and parameters
estimated from these analyses
Population genetic methods
Advancing methods in population
biology and population genetics
and parameters estimated from
these methods
Molecular genetic markers and Description of molecular marktheir uses
ers and their uses for evolutionary
biology, population biology and
diagnostics
Overviews
Reviews of evolutionary and population genetic concepts in biological
control, biological invasions, and
the use of historical collections
References
Futuyma, 2005
Falconer and MacKay, 1996; Hartl, 2000;
Conner, 2004; Lowe et al., 2004
Emerson et al., 2001; Rosenberg and Nordborg,
2002; Baldauf, 2003; Holder and Lewis, 2003;
Hall, 2007
Beaumont and Rannala, 2004; Manel et al., 2005;
Excoffier and Heckel, 2006; Noor and Feder, 2006
Avise, 2004; Roderick, 2004; Schltterer, 2004;
Armstrong and Ball, 2005; Savolainen et al., 2005;
Navajas and Roderick, 2008
Roderick, 1992; Hopper et al., 1993; Ehler, 1998;
Fagan et al., 2002; Roderick and Navajas, 2003;
Suarez and Tsutsui, 2004; Hufbauer and Roderick,
2005; Strauss et al., 2006; Sax et al., 2007;
Vellend et al., 2007
use of mitochondrial cytochrome oxidase I in insects,
although there are some complications, as for example
with individuals of hybrid origin, species that have only
recently diverged from one another and larger groups
of taxa for which single genetic markers cannot reliably distinguish species, e.g. plants. Nevertheless, DNA
barcoding can be an effective diagnostic tool for many
needs in biological control, such as linking unknown
larvae with adults, associating males and females of
the same species, identifying parasitoid species in
their hosts and making rapid taxonomic identifications.
While genetic identification of species is not new, the
novel focus on DNA barcoding worldwide is resulting in new collections and catalogues of species and
extensive DNA databases with vouchered specimens
deposited in museums, including species of interest to
biological control. These specimens and their associated
ecological and genetic data can provide baseline data
for long-term studies that are critical to understanding
the role of evolution in biological control.
of other species. These genomes provide the opportunity to identify many genes of functional significance
(Gomez-Zurita and Galian, 2005) and to develop a
wealth of molecular markers for population genetic
studies (Bouck and Vision, 2007) as well as to resolve
evolutionary relationships between taxa (Savard et al.,
2006). Because biological control agents are typically
not model organisms, biological control may not benefit as much from the genome boom as other biological
disciplines. Nevertheless, we should expect the identification of homologous genes from model organisms
that have relevance for biological control. For example,
a DNA sequencing project of the first arthropod herbi
vore, the mite Tetranychus urticae, is nearly completed
(Grbic et al., 2007), and data from this project should
lead to a better understanding of the genetics of host
plant interactions. It is likely that the same loci can be
examined in other herbivorous arthropods.
Whole genome sequencing
Recent progress has been made in databasing the
information from specimen collection events, making it possible for anyone to use these data through a
distributed database over the Internet. The core idea
is that each collection or natural history museum curates and owns its own specimens and is a provider,
each deciding what to make publicly available. Then,
if the different collections use similar fields in their
data structure, a researcher can query data across all
the participating providers simultaneously. The underlying software infrastructure supporting these tools is
the collection management system already in use for
Advances in DNA sequencing technology and corresponding bioinformatics tools of DNA fragment assembly and annotation have made it possible to obtain
nearly entire DNA genome sequences for a handful of
so-called model organisms, including human, mouse, the
puffer fish Takifugu rubripes Temminck and Schlegel,
the nematode worm Caenorhabditis elegans Maupas,
the wild mustard Arabidopsis thaliana (L.) Heynh., the
fruit fly Drosophila melanogaster Meigen, the honeybee Apis mellifera Linnaeus and a growing number
Distributed databases
406
The primacy of evolution in biological control
many museums worldwide (for a working example, see
the Berkeley Natural History Museums, BNHM, biodiversity science initiative, http://bscit.berkeley.edu/).
These systems are designed to be modular and flexible
such that individual collections can easily link to and
interact with other resources both within each museum and offsite, such as collections at other museums.
Many styles of interaction can be supported, including
machine-friendly methods, e.g. URL-based queries,
formatted data dumps and DiGIR, Darwin Core, XML
protocols and more human-friendly methods, such as
linked interactive keys. Geospatial visualization, or
mapping the data, can also be performed, e.g. using
BerkeleyMapper and DiGIRMap. Many specimens
also include a wealth of other information, including
host use and habitat variables that can also be linked to
such databases (Suarez and Tsutsui, 2004). For biological control, distributed databases can provide worldwide information on potential agents, including indigenous and introduced ranges. Coupled with climate or
other ecological data, distributional data can be used to
predict future spread and physiological tolerance. Target and non-target organisms can be investigated in a
similar way.
New computational tools
Computational tools for both population genetics
and phylogenetics are advancing rapidly. In addition
to bioinformatics through which enormous quantities
of genetic data can now be processed and summarized
(noted above), new approaches are emerging for both
phylogenetic reconstruction (Baldauf, 2003; Noor and
Feder, 2006) and population genetics (Emerson et al.,
2001; Manel et al., 2005; Excoffier and Heckel, 2006).
One promising avenue has been the use of Bayesian
statistical methods to test a diversity of hypotheses in
both areas (Holder and Lewis, 2003; Beaumont and
Rannala, 2004). New phylogenetic approaches should
improve biological control through more accurate assessments of species relationships for agent selection.
Knowledge of host and habitat use of close relatives
can be used to minimize non-target effects. Advances
in population genetics are providing novel ways to
trace the history of populations and to infer and predict
demographic parameters useful for biological control
programmes, such as population growth, geographical
spread and habitat suitability.
Discussion
One goal of this paper is to raise awareness of evolutionary biology and related fields so that new approaches can be used to better inform biological control programmes. A number of researchers have argued
that recognizing general principles in the field of biological control can help move the discipline to more
of a predictive science from one of a set of individual
case studies (see discussions in Kareiva, 1996; Simberloff and Stiling, 1996; Holt and Hochberg, 1997;
Hochberg and Gotelli, 2005). Certainly, evolutionary
principles can help in this regard, as noted above for
choosing agents and in better understanding the roles
of adaptation and genetic variation. New methodologies provide additional data for study and new ways to
examine existing data. Yet, many desirable manipulations are difficult, if even feasible. In this paper, biological control can likely benefit from the discipline of
invasion biology, which also focuses on introductions
and colonizations, establishment and spread (see Fagan
et al., 2002). Particularly important to both fields is a
better understanding of successes and failures, which
necessarily demands follow-up studies or long-term
monitoring.
Relevance of invasion biology
Classical biological control can be thought of as a
series of semi-replicated field manipulations in which
organisms are transplanted into new environments. It is
well recognized that this process shares much with biological invasions (McEvoy et al., 2008, this proceedings; Warner et al., 2008, this proceedings) in that the
introduced organisms typically find abundant resources
and few predators, parasites or competitors (Ehler,
1998; Roderick and Navajas, 2003; Mller-Scharer,
2006). Both disciplines offer the possibility to follow
long-term effects of introductions/colonizations. To
make use of these studies, it is important to follow the
introductions/colonizations from the beginning. In this
regard, classical biological control has a distinct advantage in that introductions are well planned. In addition,
to understand what factors determine which introductions will be successful, it is important to monitor not
only well-known successes, but also failures. Unfortu
nately, after-the-fact studies are typically difficult to
fund, although one might argue that they are the most
important (McFadyen, 2008, this proceedings). Certainly, long-term evolutionary responses associated
with classical biological control or invasion biology
can only be understood when historical data or specimens are available for comparison.
Suggestions for further study
To understand better evolutionary processes in
classical biological control, we can make several recommendations that are relevant to agent selection,
quarantine, pre- and post-release studies, establishment, agentstarget interactions, non-target effects and
risk analysis. The following are the suggestions for further study:
1. Involve researchers from several disciplines in biological control efforts. This review has focussed on
407
XII International Symposium on Biological Control of Weeds
2.
3.
4.
5.
evolutionary biology where the recommendation
is particularly relevant, but biological control programmes would also benefit from those studying
risk and social science.
Preserve and database pre-release samples and associated ecological data. Pre-release data collections
with vouchers are critical both for future ecological
and genetic studies. We recommend that the specimens be accessioned to a well-maintained national
or regional collection or natural history museum,
which has a mission of long-term storage. For some
questions, specimens also can provide information
on habitat variables, including competitors, predators and pathogens (Suarez and Tsutsui, 2004). This
is a requirement of some countries, e.g. Australia,
when introducing a potential biological control agent
for quarantine studies.
Conduct genetic studies in the context of rearing in
quarantine and pre-release testing. If rearing studies
are being conducted and if it is possible to have replicates within the rearing design, by keeping track
of parentage, it is possible to conduct simple studies
of heritability for host use and for other traits. Such
data is invaluable for predicting host use and could
be obtained with little extra work.
Emphasize long-term monitoring. Genetic markers
have been proven to be useful to monitor the fate
of released individuals (Navajas et al., 2001). Longterm monitoring of classical biological efforts will
benefit many sub-disciplines, not just evolutionary
biology. Risk assessment is one area in which realworld data are lacking, e.g. McFadyen (2008, this
proceedings). Such monitoring is not easy, is often
difficult to fund and will likely take a coordinated
effort of researchers.
Publish on all biological control introductions, including those that do not work for whatever reason.
By publishing results of unsuccessful introductions,
we can compare what does work with what does not
and hence assess variables underlying success. Admittedly, publishing on unsuccessful introductions
will take some change in editorial policies in respected journals. Towards the same goal, researchers
can present research on unsuccessful introductions
in a broader hypothetical framework, such as invasion biology, so that negative results can be more
interesting and relevant.
Acknowledgements
We thank the ISBCW Organizational Committee, Mic
Julien, Ren Sforza, Marie-Claude Bon, Brian Rector
and Janine Vitou, for their hard work and insightful
comments and suggestions. This work was supported by
INRA France, US Department of Agriculture, the Fulbright/Franco-American Commission and the FranceBerkeley Fund.
References
Armstrong, K.F. and Ball, S.L. (2005) DNA barcodes for bio
security: invasive species identification. Philosophical
Transactions of the Royal Society: Biological Sciences
360, 18181823.
Avise, J.C. (2004) Molecular Markers, Natural History, and
Evolution, 2nd edn. Chapman and Hall, London, 541 pp.
Baldauf, S.L. (2003) Phylogeny for the faint of heart: a tutorial. Trends in Genetics 19, 345351.
Beaumont, M.A. and Rannala, B. (2004) The Bayesian revolution in genetics. Nature Reviews Genetics 5, 251261.
Bouck, A. and Vision, T. (2007) The molecular ecologists
guide to expressed sequence tags. Molecular Ecology 16,
907924.
Carson, H.L. (1990) Increased genetic variance after a population bottleneck. Trends in Ecology and Evolution 5,
228230.
Conner, J.K. (2004) Primer in Ecological Genetics. Sinauer,
Sunderland, MA, 304 pp.
Cotter, S.C. and Edwards, O.R. (2006) Quantitative genetics
of preference and performance on chickpeas in the noctuid moth, Helicoverpa armigera. Heredity 96, 396402.
Ehler, L.E. (1998) Invasion biology and biological control.
Biological Control 13, 127133.
Emerson, B.C., Paradis, E. and Thbaud, C. (2001) Revealing the demographic histories of species using DNA sequences. Trends in Ecology & Evolution 16, 707716.
Excoffier, L. and Heckel, G. (2006) Computer programs for
population genetics data analysis: a survival guide. Nature
Reviews Genetics 7, 745758.
Fagan, W.F., Lewis, M.A., Neubert, M.G. and van den
Driesse, P. (2002) Invasion theory and biological control.
Ecological Letters 5, 148157.
Falconer, D.S. and MacKay, T.F.C. (1996) Introduction to
Quantitative Genetics. Longman, Harlow, England, 464
pp.
Fauverguee, X., Malausa, J.C., Giuge, L. and Corchamp, P.
(2007) Invading parasitoids suffer no Allee Effect: a manipulative field experiment. Ecology 88, 23922403.
Futuyma, D.J. (2005) Evolution. Sinauer, Sunderland, MA,
543 pp.
Gomez-Zurita, J. and Galian, J. (2005) Current knowledge on
genes and genomes of phytophagous beetles (Coleoptera :
Chrysomeloidea, Curculionoidea): a review. European
Journal of Entomology 102, 577597.
Grbic, M., Khila, A., Lee, K., Bjelica, A., Grbic, V., Whistlecraft, J., Verdon, L., Navajas, M. and Nagy, L. (2007)
Mity model: Tetranychus urticae, a candidate for chelicerate model organism. Bioessays 29, 489496.
Grevstad, F.S. (1999) Experimental invasions using biological control introductions: the influence of release size on
the chance of population establishment. Biological Invasions 1, 313323.
Hall, B.G. (2007) Phylogenetic trees made easy: a how-to
manual, 3rd edn. Sinauer, Sunderland, MA, 203 pp.
Hartl, D.L. (2000) A Primer of Population Genetics, 3rd edn.
Sinauer, Sunderland, MA, 180 pp.
Hochberg, M.E. and Gotelli, N.J. (2005) An invasions special
issue. Trends in Ecology & Evolution 20, 211269.
Holder, M. and Lewis, P.O. (2003) Phylogeny estimation: traditional and bayesian approaches. Nature Reviews Genetics 4, 275284.
408
The primacy of evolution in biological control
Holt, R.D. and Hochberg, M.E. (1997) When is biological
control evolutionarily stable (or is it)? Ecology 78, 1673
1683.
Hopper, K.R., Roush, R.T. and Powell, W. (1993) Management of genetics of biological control introductions. Annual Review of Entomology 38, 2751.
Hufbauer, R.A. (2002) Evidence for nonadaptive evolution in
parasitoid virulence following a biological control introduction. Ecological Applications 12, 6678.
Hufbauer, R.A. and Roderick, G.K. (2005) Microevolution in
biological control: mechanisms, patterns, and processes.
Biological Control 35, 227239.
Kareiva, P. (1996) Contributions of ecology to biological
control. Ecology 77, 19631964.
Kolbe, J.J., Glor, R.E., Schettino, L.R., Lara, A.C., Larson,
A. and Losos, J.B. (2004) Genetic variation increases during biological invasion by a Cuban lizard. Nature 431,
177181.
Lowe, A., Harris, S. and Ashton, P. (2004) Ecological Genetics: Design, Analysis, and Application. Wiley, New York,
344 pp.
Manel, S., Gaggiotti, O.E. and Waples, R.S. (2005) Assignment methods: matching biological questions with appropriate techniques. Trends in Ecology & Evolution 20,
136142.
Marrs, R.A., Sforza, R. and Hufbauer, R.A. (2008) When invasion increases population genetic structure: a study with
Centaurea diffusa. Biological Invasions 10, 561572.
Mller-Scharer, H. (2006) Evolution in invasive plants: implications for biological control. Trends in Ecology &
Evolution 19, 417422.
Navajas, M. and Roderick, G.K. (2008) Molecular diagnosis. In: Capinera, J. (ed.) Encyclopedia of Entomology.
Springer, Dordrecht, pp. 191196.
Navajas, M., Thistlewood, H., Lagnel, J., Marshall, D., Tsagkarakou, A. and Pasteur, N. (2001) Field releases of the
predatory mite Neoseiulus fallacis (Acari : Phytoseiidae) in
Canada, monitored by pyrethroid resistance and allozyme
markers. Biological Control 20, 191198.
Nei, M., Maruyama, T. and Chakraborty, R. (1975) The bottle
neck effect and genetic variability in populations. Evolution 29, 110.
Noor, M.A.F. and Feder, J.L. (2006) Speciation genetics: evolving approaches. Nature Reviews Genetics 7, 851861.
Roderick, G.K. (1992) Post-colonization evolution of natural
enemies. In: Kauffman, W.C. and Nechols, J.E. (eds) Thomas Say Proceedings: Selection Criteria and Ecological
Consequences of Importing Natural Enemies. Entomological Society of America, Lanham, MD, pp. 7186.
Roderick, G.K. (2004) Tracing the origin of pests and natural enemies: genetic and statistical approaches. In: Ehler,
L.E., Sforza, R. and Mateille, T. (eds) Genetics, Evolution,
and Biological Control. CABI Publishing, New York,
pp. 97109.
Roderick, G.K. and Navajas, M. (2003) Genes in novel environments: Genetics and evolution in biological control.
Nature Reviews Genetics 4, 889899.
Rosenberg, N.A. and Nordborg, M. (2002) Genealogical
trees, coalescent theory and the analysis of genetic polymorphisms. Nature Reviews Genetics 3, 380390.
Savard, J., Tautz, D., Richards, S., Weinstock, G.M., Gibbs,
R.A., Werren, J.H., Tettelin, H. and Lercher, M.J. (2006)
Phylogenomic analysis reveals bees and wasps (Hymenoptera) at the base of the radiation of Holometabolous
insects. Genome Research 16, 13341338.
Savolainen, V., Cowan, R.S., Vogler, A.P., Roderick, G.K.
and Lane, R. (2005) Towards writing the encyclopaedia
of life: an introduction to DNA barcoding. Philosophical
Transactions of the Royal Society: Biological Sciences
360, 18051811.
Sax, D.F., Stachowicz, J.J., Brown, J.H., Bruno, J.F., Dawson, M.N., Gaines, S.D., Grosberg, R.K., Hastings, A.,
Holt, R.D., Mayfield, M.M., OConnor, M.I. and Rice,
W.R. (2007) Ecological and evolutionary insights from
species invasions. Trends in Ecology and Evolution 22,
465471.
Schltterer, C. (2004) The evolution of molecular markers
just a matter of fashion? Nature Reviews Genetics 5,
6369.
Simberloff, D. and Stiling, P. (1996) How risky is biological
control? Ecology 77, 19651974.
Slatkin, M. (1996) In defense of founder flush theories of
evolution. American Naturalist 147, 493505.
Stephens, P.A. and Sutherland, W.J. (1999) Consequences of
the Allee effect for behaviour, ecology and conservation.
Trends in Ecology and Evolution 14, 401405.
Strauss, S.Y., Lau, J.A. and Carroll, S.P. (2006) Evolutionary responses of natives to introduced species: what do
introductions tell us about natural communities. Ecology
Letters 9, 357374.
Suarez, A.V. and Tsutsui, N.D. (2004) The value of museum collections for research and society. BioScience 54,
6674.
Vellend, M., Harmon, L.J., Lockwood, J.A., Mayfield, M.M.,
Hughes, A.R., Wares, J.P. and Sax, D.F. (2007) Effects of
exotic species on evolutionary diversification. Trends in
Ecology & Evolution 22, 481488.
Whitlock, M.C. and Wade, M.J. (1995) Speciation: founder
events and their effects on X-linked and autosomal genes.
American Naturalist 145, 676685.
Zangerl, A.R. and Berenbaum, M.R. (2005) Increase in toxicity of an invasive weed after reassociation with its coevolved herbivore. Proceedings of the National Academy
of Sciences USA 102, 1552915532.
409
Does phylogeny explain the host-choice
behaviour of potential biological control
agents for Brassicaceae weeds?
H.L. Hinz1, M. Schwarzlnder2 and J. Gaskin3
Summary
Four invasive Brassicaceae are currently being studied at CABI Europe-Switzerland for biological
control. A phylogenetic approach to host testing has so far been hampered by the fact that the evolutionary relationships of taxa within the Brassicaceae were unclear. Recently, a new phylogeny of the
Brassicaceae, largely based on molecular studies, has been proposed. This presents a unique opportunity to relate host-range test results for some of our Brassicaceae agents to the new phylogeny. The
host range of Ceutorhynchus scrobicollis Nerensheimer & Wagner, a root-crown mining weevil investigated as a potential agent for garlic mustard, Alliaria petiolata (Bieb.) Cavara & Grande, appeared
to closely follow the new classification (significant linear relationship between phylogenetic distance
and host-range test results). However, for Ceutorhynchus cardariae Korotyaev, a gall-inducing
weevil considered as biocontrol agent for hoary cress, Lepidium draba L., phylogenetic distance of
the test species to the target weed did not explain a significant amount of the variation in host preference or suitability. These results question the general applicability of the centrifugal phylogenetic
method, where it is assumed that species more closely related to the target are at greater risk of attack
than species more distantly related. The importance of other factors, specifically secondary metabolite
profiles and morphological characteristics for the host-choice behaviour of C. cardariae are currently
being investigated.
Keywords: Alliaria petiolata, Lepidium draba, centrifugal phylogenetic method.
Introduction
To date, no biological control agents have been released
against weeds in the mustard family (Brassicaceae).
The main reason for this is the familys large number
of economically important crop species and its many
genera indigenous to North America. Four invasive Brassicaceae are currently being studied at CABI EuropeSwitzerland for biological control. They are garlic
mustard, Alliaria petiolata (Bieb.) Cavara & Grande,
hoary cress, Lepidium draba L., perennial pepperweed,
Lepidium latifolium L. (Hinz et al., 2008, this proceedings), and dyers woad, Isatis tinctoria L. (Cortat et al.,
2008, this proceedings).
CABI Europe-Switzerland, Rue des Grillons 1, 2800 Delmont,
Switzerland.
2
University of Idaho, Department of Plant, Soil and Entomological Sciences, Moscow, ID 83844-2339, USA.
3
USDA, ARS, NPARL, 1500 N. Central Ave., Sidney, MT 59270, USA.
Corresponding author: H.L. Hinz <[email protected]>.
CAB International 2008
1
In biological control of weeds, it is generally assumed that species closely related to the target are at
greater risk of attack than species more distantly related.
However, a phylogenetic approach to host testing has
so far been hampered by the fact that the evolutionary
relationships of taxa within the Brassicaceae were unclear. The subdivision of the Brassicaceae at the tribal
and subtribal levels has been a controversial aspect in
the systematics of the family (Appel and Al-Shehbaz,
2003). Appel and Al-Shehbaz (2003) concluded that in
the absence of comprehensive, family-wide molecular
data it is not regarded advisable to propose or recommend any classification system. Recently, Al-Shehbaz
et al. (2006) and Bailey et al. (2006) proposed the longawaited new tribal alignment of the Brassicaceae based
on molecular studies and careful evaluation of morphology and generic circumscriptions. This presented
a unique opportunity to see whether host-range test
results for some of our Brassicaceae agents correlated
with the new phylogeny of the Brassicaceae. We used
host-specificity test results of two currently studied
potential biocontrol agents, viz., Ceutorhynchus scro-
410
Does phylogeny explain the host-choice behaviour of potential biological control agents for Brassicaceae weeds?
bicollis Nerensheimer & Wagner (Coleoptera, Curculionidae) on A. petiolata and Ceutorhynchus cardariae
Korotyaev on L. draba, and correlated them with the
genetic distance of test plant species to the respective
target weed.
Methods and materials
The study species
A. petiolata is a strict biennial European herb introduced into North America that, by 2000, had spread to
34 states and four Canadian provinces (Blossey et al.,
2001). A. petiolata is one of the few introduced herbaceous species that invades and dominates the understory of forested areas in North America. In 1998, a
biological control program was started, and four weevils were prioritized as potential agents. One of these is
the root-crown mining weevil, C. scrobicollis. Adults
of this species aestivate in summer. Oviposition on
A. petiolata rosettes begins in mid-September and lasts
until the beginning of April of the following year (Gerber et al., 2007). Eggs are laid into petioles and leaves,
and the growing point and larvae mine through petioles
towards the roots and develop mainly in root crowns,
occasionally also in shoot bases. Mature larvae leave
host plants in spring to pupate in the soil, and the next
generation of adults emerges in May and June.
L. draba [=Cardaria draba (L.) Desv.], with its
two subspecies, L. draba spp. draba and L. draba spp.
chalapense, and its close relative Lepidium appelianum Al-Shehbaz [=Cardaria pubescens (C.A. Mey.)
Jarm.], are perennial mustards (Brassicaceae) of European origin that were introduced in the USA in the late
19th century (Lyons, 1998). Since then, they spread
throughout the western and the northeastern states and
are now declared noxious weeds in 16 states and three
Canadian provinces (Rice, 2005). Because they are difficult to control sustainably using mechanical or chemical methods, a consortium was established in Spring
2001 to investigate the scope for classical biological
control. Five insect species are currently being studied
for the biological control of L. draba: four weevils and
one flea beetle. One of these is the gall-inducing weevil
C. cardariae Korotyaev. Females of C. cardariae lay
their eggs in the leaf midribs, petioles and developing
shoots of L. draba from early spring until mid-June
(Hinz et al., 2007). Oviposition induces the formation
of galls, in which the larvae mine and develop. Mature
larvae leave the plants to pupate in the soil, and the next
generation of adults emerges from May to July.
Host-specificity tests
A. petiolataC. scrobicollis: Between 1999 and 2006,
sequential no-choice oviposition tests were conducted
with C. scrobicollis. A mated pair of C. scrobicollis
was placed into a transparent plastic cylinder (11 cm
diameter, 15 cm high) and alternately offered cut leaves
of A. petiolata, then those of a test-plant species. After
3 to 4 days, the plant material was removed, checked
for feeding marks, dissected for eggs and the weevils
provided with fresh plant material. Each exposure period was treated as one replicate. A replicate for test
plants was only regarded as valid when the female laid
at least one egg into the test plant or into an A. petiolata
plant following a test plant.
Plant species accepted for oviposition were subsequently exposed in no-choice oviposition and development tests. Two to three females and one to two males
were released onto individually potted, gauze-covered
rosettes of A. petiolata or onto test species. To verify
that females were fertile, one pair was offered a cut
leaf of garlic mustard in a cylinder for 2 to 3 days after
which plant material was dissected for eggs. Only females that laid eggs on these leaves were used for the
tests. For each no-choice experiment, two to 13 garlic
mustard plants were infested concurrently as controls.
After 2 to 4 weeks, weevils were retrieved and plants
re-covered with gauze bags. In late spring of the following year, all plants were searched regularly for
emerging adults until emergence ceased.
L. drabaC. cardariae: Between 2003 and 2006, nochoice oviposition and development tests were conducted with C. cardariae. One to five females and one
to four males, depending on plant size, were placed
onto individually potted, gauze-covered test plants or
L. draba (control plants). All females were tested for
egg-laying before use in tests (see above). Each time a
series of test plants was infested, two to four L. draba
plants were infested concurrently as controls. After 8
to 12 days, weevils were retrieved from the plants, and
feeding, oviposition and gall formation were recorded.
To ensure that females had a chance to feed and oviposit
on L. draba in between tests, all weevils were placed
into cylinders for a couple of days and provided with cut
plant material of L. draba before beginning a new series
of tests. About 1 month after infestation, plants were rechecked for gall development. After about 12 weeks,
all plants were checked for adult emergence.
Molecular and statistical analyses
Leaf material of plant species within the family Brassicaceae used in tests with C. scrobicollis and
C. cardariae that were not included in the phylogenetic
tree of Beilstein et al. (2006) was collected and subjected to molecular analysis. Genomic DNA was isolated
using a cetyl trimethylammonium bromide method.
Polymerase chain reaction (PCR) amplification of the
chloroplast ndhF region was done with same primers and conditions as in Beilstein et al. (2006). PCR
products were purified using QIAquick PCR Purification kit (Qiagen) before sequencing in a Beckman CEQ
2000XL automated sequencer using standard protocols
including the LFR-1 method of injection time and volt-
411
XII International Symposium on Biological Control of Weeds
age. Sequences were aligned manually using SeAl
(Rambaut, 1996). Maximum parsimony (MP) analysis
was performed and uncorrected (p) distances of the
data set were determined using PAUP* v. 4.0b8 (Swofford, 2000). For estimation of the most parsimonious
phylogenetic trees, the heuristic MP search employed
500 random taxon addition sequences and the treebisection-reconnection branch-swapping algorithm. All
characters were weighted equally. A 10,000-replicate,
fast stepwise-addition, bootstrap analysis was conducted
to assess clade support. The phylogenetic analysis was
included to illustrate evolutionary relationships of the
plant taxa, while the distance measurements were correlated with host-specificity measurements.
To relate results of host-specificity tests (for C.
scrobicollis, the number of eggs laid and the number
of offspring produced per female; for C. cardariae, the
number of galls induced and offspring produced per
female) with the genetic distances generated for test
plants of each target weed species, we used simple lin-
Figure 1.
ear regression analysis. The two target weeds were not
included in the analyses, as the aim was to test whether
plants more closely related to the respective target weed
would be preferred by female weevils for oviposition
and/or would be more suitable for weevil development.
When data on the genetic distance for a test-plant species was not available, we used congeners of known genetic distance from the target to extrapolate the missing
value to the precision of two decimal places.
Results
A. petiolataC. scrobicollis
Of the 28 plant species and varieties for which
data on both oviposition-test results and genetic distance were available, 18 were accepted for oviposition by C. scrobicollis females (Fig. 1A). As expected,
females of C. scrobicollis laid more eggs on plants more
closely related to A. petiolata (r2 = 0.298, F1,27 = 11.02,
Relationship between genetic distance of test-plant species to the
control (i.e. target weed) Alliaria petiolata and (A) the number of
eggs laid and (B) the number of offspring produced per Ceutorhynchus scrobicollis female on the respective test-plant species.
412
Does phylogeny explain the host-choice behaviour of potential biological control agents for Brassicaceae weeds?
P = 0.003) than on less closely related plants. The two
plant species that received the most eggs, the European
species, Peltaria alliacea Jacq. and Thlaspi arvense L.,
are in the same tribe as A. petiolata (Thlaspideae), according to the new molecular phylogeny (Al-Shehbaz
et al., 2006; Fig. 2). No plant species outside the family
Brassicaceae supported normal oviposition behaviour
of C. scrobicollis (Gerber et al., 2005).
Of the 21 plant species and varieties for which data
on both results of no-choice development tests and
genetic distance were available, adults emerged from
Figure 2.
five species other than A. petiolata, viz. the European
species, Nasturtium officinale R.Br., P. alliacea, and
T. arvense, and the North American species, Rorippa
sinuata (Nutt.) A.S. Hitchc. In addition, a single adult
in a single replicate emerged from the commercially
grown cabbage variety Brassica oleracea sabauda
Paradisler. However, none of the other six B. oleracea varieties we offered were attacked (Gerber et al.,
2005), and moreover, no attack occurred under multiplechoice cage conditions. This suggests that B. oleracea
and its various cultivars are outside the fundamental
Phylogeny of Alliaria petiolata and test-plant species used in oviposition and development tests with Ceutorhynchus scrobicollis. The figure is a strict consensus
of the ten most parsimonious trees, each 615 steps in length, derived from 1965
aligned bases of the chloroplast gene ndhF. Bootstrap values (>50%) are shown
above branches. Taxa that supported adult development are shown in bold font.
An asterisk indicates that the DNA sequence was provided by M. Beilstein.
413
XII International Symposium on Biological Control of Weeds
host range of C. scrobicollis. We therefore consider the
emergence of a single adult an artefact of experimental
conditions. Females of C. scrobicollis produced more
offspring on plants more closely related to A. petiolata
(r2 = 0.237, F1,20 = 5.91, P = 0.025; Figs. 1B and 2) than
on less closely related species.
It is worth noting that the North American species,
Noccaea fendleri (A. Gray) Holub (see Fig. 2), was
formerly known as Thlaspi montanum L. (Thlaspidae),
until all North American species of Thlaspi were recently assigned to the tribe Noccaeeae, genus Noccaea,
to form a monophyletic group separate from the European species of Thlaspi (Koch and Al-Shehbaz, 2004;
also see Fig. 2). Host-specificity results with C. scrobicollis supported this new classification.
L. drabaC. cardariae
Of the 58 plant species for which data on both
host-specificity test results and genetic distance were
available, galls were induced on 11 species and adults
Figure 3.
emerged from ten (Fig. 3A, B), including the second
target weed L. appelianum. In contrast to C. scrobicollis on A. petiolata, neither the number of galls induced
nor the number of offspring produced per C. cardariae
female was correlated with genetic distance of testplant species from the target weed L. draba (number
of galls induced, r2 = 0.025, F1,57 = 1.46, P = 0.232;
number of offspring produced, r2 = 0.020, F1,57 = 1.15,
P = 0.287). Plant species that supported gall induction and the development of adults included the fairly
distantly related species Caulanthus anceps Payson
and C. inflatus S. Wats. (the genetic distance for both
of which was extrapolated from C. crassicaulis, see
Methods and materials for details), Stanleya pinnata (Pursh) Britt. and Lobularia maritima (L.) Desv.
(Fig. 4). While only one adult each emerged from
each of the latter two species, a similar number of
adults emerged from C. anceps and C. inflatus as from
L. draba control plants (Hinz et al., 2007). In contrast,
S. viridiflora Nutt. did not support gall induction and
was minimally fed on by C. cardariae.
Relationship between genetic distance of test-plant species
to the control (i.e. target weed) Lepidium draba and (A) the
number of galls induced and (B) the number of offspring produced per Ceutorhynchus cardariae female on the respective
test-plant species.
414
Does phylogeny explain the host-choice behaviour of potential biological control agents for Brassicaceae weeds?
Figure 4.
Phylogeny of Lepidium draba and test-plant species (used in oviposition and development tests with Ceutorhynchus cardariae). The figure is a strict consensus
of the 54 most parsimonious trees, each 999 steps in length, derived from 2017
aligned bases of the chloroplast gene ndhF. Bootstrap values (>50%) are shown
above branches. Taxa that supported adult development are shown in bold font.
An asterisk indicates that the DNA sequence was provided by H. Beilstein.
Discussion
In three multiple-choice, field-cage tests established
between 2004 and 2006 with C. cardariae, in which
several test species were exposed that had supported
development under no-choice conditions, only the
three target weeds, i.e. the two subspecies L. draba spp.
draba and L. draba spp. chalapense and L. appelianum
were attacked, indicating a very narrow host range for
the weevil under multiple-choice conditions. However,
plants that supported development under no-choice
conditions constitute the physiological host range of a
species, which in turn appears to be an effective criterion for identifying species potentially at risk of attack,
as there is no example of an insect agent attacking a
plant outside its physiological host range after release
(Pemberton, 2000; van Klinken and Edwards, 2002).
In North America alone, the family Brassicaceae
is represented by approximately 600 species in more
than 35 endemic genera. Also in North America, there
415
XII International Symposium on Biological Control of Weeds
are at least 123 Brassicaceae species within 23 genera
that are considered economically important. Our hostspecificity test results showed that none of the most
widespread and important economic Brassicaceae species are at risk of attack by either biological control
candidate. In addition, if commercially grown Brassicaceae would be part of the normal host range of any
of our agents tested, the species would have been recorded as a pest in the European literature, which is not
the case (Schwarz et al., 1990). However, native North
American Brassicaceae have never been previously exposed to the insects studied here, and the large number
of plant species makes it difficult, if not impossible, to
test them all. We were therefore hoping to use the new
Brassicaceae phylogeny to select plant species for additional host-specificity tests and to extrapolate the risk
of non-target attack on more distantly related genera to
the respective target weeds.
The host range of C. scrobicollis, investigated for
the biological control of A. petiolata, only appears to
include species closely related to the target. In contrast,
C. cardariae on L. draba appears to have a disjunct
host range with some distantly related plants supporting development to a similar degree as the control under no-choice conditions. These results are at odds with
the centrifugal phylogenetic method (Wapshere, 1974),
where it is generally assumed that species closely related to the target are at greater risk of attack than species
more distantly related. In the absence of detailed, explicit phylogenies, the centrifugal phylogenetic method
has usually been based on traditional taxonomic classifications, which has been questioned (Briese and
Walker, 2002; Kelch and McClay, 2003). However, our
study was based on the latest molecular phylogeny of
the Brassicaceae.
For Longitarsus jacobaeae Waterhouse, which was
studied for the biological control of Senecio jacobaea
L., neither adult feeding nor the fundamental larval
host range of L. jacobaeae were clearly predicted by
the phylogeny of the genus Senecio (U. Schaffner,
unpublished data). In contrast, it was found that leaf
dry matter content of Senecio species explained a significant amount of the variability in the amount of leaf
area eaten by adult beetles (U. Schaffner, unpublished
data). Another study found that the functional composition of herbivore assemblages on 18 shrubs was correlated with respective leaf structural traits (Peeters,
2002). More specifically, leaf trichome density and leaf
surface waxes have been shown to influence host suitability or preference of insect herbivores (Eigenbrode
and Espelie, 1995; Levin, 1973, and Edwards, 1982,
in Peeters, 2002). Finally, secondary plant compounds
play an important role in host finding, host acceptance
and host suitability of insect herbivores (e.g. Renwick,
1989; Rask et al., 2000). We are currently investigating some of these potential factors to better explain the
host acceptance and suitability patterns observed for
C. cardariae. Because host acceptance and suitability
of herbivores might be influenced by a combination
of many different secondary compounds and physical
characteristics (see above), we will be using a relatively
new technique, i.e. metabolomics that will allow us to
identify and quantify all metabolites (primary and secondary) of an organism simultaneously (Bezemer and
van Dam, 2005 and refs therein).
In conclusion, it is not our intention to question
the importance of phylogenetic relatedness to understanding host-choice behaviour of herbivorous insects.
Indeed, since the general adoption of the centrifugal
phylogenetic method for host-range testing of weed
biological control candidates (Wapshere, 1974), there
has not been a single significant case of an agent that
attacked a non-target that was completely unanticipated
or unpredicted (Pemberton, 2000). We do, however,
propose that other factors influence the host choice behaviour of insect herbviores more than commonly considered. As already suggested previously (e.g. Keller,
1999; Withers; 1999; Briese, 2005), these factors need
to be better understood and should be included in the
selection of test plant species and in the interpretation
of host-specificity test results.
Acknowledgements
We thank Ghislaine Cortat, Bethany Muffley, Carole
Rapo, Christian Lechenne and Florence Willemin for
technical assistance. We would also like to thank Esther
Gerber for providing her data. We are grateful to Urs
Schaffner for advice in data analyses. Mark Beilstein
generously provided sequence data nexus files. Financial support for these projects came from the Idaho State
Department of Agriculture through the University of
Idaho, the Wyoming Biological Control Steering Committee, the Montana Weed Trust Fund through Montana
State University, the USDI-BLM, USDA-APHIS-PPQ,
USDA-ARS, the Strategic Environmental Research
Development Programme (SERDP) through Cornell
University, USDA Forest Service through Cornell University, US Fish and Wildlife Service, Minnesota Department of Natural Resources, Wisconsin Department
of Natural Resources and the Illinois Natural History
Survey.
References
Al-Shehbaz, I. A., Beilstein, M.A. and Kellogg, E.A. (2006)
Systematics and phylogeny of the Brassicaceae. Plant
Systematics and Evolution 259, 89120.
Appel, O. and Al-Shehbaz, I.A. (2003) Cruciferae. In: Kubitzki, K., and Bayer, C. (eds) The Families and Genera of
Vascular Plants. Springer, Berlin, pp. 75174.
Bailey, C.D., Koch, M.A., Mayer, M., Mummenhoff, K.,
OKane Jr. S.L., Warwick, S.I., Windham, M.D. and
Al-Shehbaz, I.A. (2006) Toward a global phylongeny of
the Brassicaceae. Molecular Biology and Evolution 23,
21422160.
416
Does phylogeny explain the host-choice behaviour of potential biological control agents for Brassicaceae weeds?
Beilstein, M.A., Al-Shehbaz, I.A. and Kellogg, E.A. (2006)
Brassicaceae phylogeny and trichome evolution. American Journal of Botany 93, 607619.
Bezemer, T.M. and van Dam, N.M. (2005) Linking above
ground and belowground interactions via induced plant
defenses. Trends in Ecology and Evolution 20, 617624.
Blossey, B., Nuzzo, V., Hinz, H.L. and Gerber, E. (2001)
Developing biological control of Alliaria petiolata (M.
Bieb.) Cavara and Grande (garlic mustard). Natural Areas
Journal 21, 357367.
Briese, D. T. (2005) Translating host-specificity test results
into the real world: The need to harmonize the yin and
yang of current testing procedures. Biological Control 35,
208214.
Briese, D.T. and Walker, A. (2002) A new perspective on the
selection of test plants for evaluating the host-specificity
of weed biological control agents: the case of Deuterocampta quadrijuga, a potential insect control agent of
Heliotropium amplexicaule. Biological Control 25, 273
287.
Eigenbrode, S.D. and Espelie, K.E. (1995) Effects of plant
epicuticular lipids on insect herbivores. Annual Review of
Entomology 40, 171194.
Gerber, E., Hinz, H.L. and Cortat, G. (2005). Biological control of garlic mustard, alliaria petiolata. Annual Report
2004. Unpublished Report. CABI Europe-Switzerland,
Delmont, Switzerland, p. 42.
Gerber, E., Hinz, H.L. and Blossey, B. (2007) Impact of the
belowground herbivore and potential biological control
agent, Ceutorhynchus scrobicollis, on Alliaria petiolata
performance. Biological Control 42, 355364.
Hinz, H.L., Cortat, G., Muffley, B. and Tostado, C. (2007)
Biological control of whitetops, Lepidium draba and
L. appelianum. Annual Report 2006. Unpublished Report. CABI Europe-Switzerland, Delmont, Switzerland,
p. 32.
Kelch, D.G. and McClay, A. (2003) Putting the phylogeny
into the centrifugal phylogenetic method. In: Cullen, J.M.,
Briese, D.T., Kriticos, D.J., Londsale, W.M., Morin, L.
and Scott, J.K. (eds) Proceedings of the XI International
Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 287291.
Keller, M.A. (1999) Understanding host selection behaviour:
the key to more effective host specificity testing. In: Withers, T.M., Barton Browne, L. and Stanley, J. (eds) Host
Specificity Testing in Australasia: Towards Improved As-
says for Biological Control. CRC for Tropical Pest Management, Brisbane, Australia, pp. 8492.
Koch, M. and Al-Shehbaz, I.A. (2004) Taxonomic and phylogenetic evaluation of the American Thlaspi species:
identity and relationship to the Eurasian genus Noccaea
(Brassicaceae). Systematic Botany 29(2), 375384.
Lyons, K. E. (1998) Cardaria draba (L.) Desv. Heart-podded
hoary cress, Cardaria chalepensis (L.) Hand-Maz. Lenspodded hoary cress and Cardaria pubescens (C.A. Meyer)
Jarmolenko Globe-podded hoary cress. Elemental Stewardship Abstract. The Nature Conservancy, Virginia, USA.
Peeters, P.J. (2002) Correlations between leaf structural traits
and the densities of herbivorous insect guilds. Biological
Journal of the Linnean Society 77, 4365.
Pemberton, R.W. (2000) Predictable risk to native plants in
weed biological control. Oecologia 125, 489494.
Rambaut, A. (1996) SeAl sequence alignment editor. Avail
able at: http://evolve.zoo.ox.ac.uk/software.html?id5seal.
Rask L., Andrasson E., Ekbom B., Eriksson S., Pontoppidan
B. and Meijer J. (2000) Myrosinase: gene family evolution and herbivore defense in Brassicaceae. Plant Molecular Biology 42, 93113.
Renwick, J. A. A. (1989). Chemical ecology of oviposition in
phytophagous insects. Experientia 45, 223228.
Rice, P. (2005) Invaders database system. Available at: http://
invader.dbs.umt.edu/ (accessed May 2005).
Schwarz, A., Etter, J., Knzler, R., Potter, C. and Rauchenstein, H.R. (1990) Pflanzenschutz im Integrierten Gemsebau. Verlag LmZ Landwirtschafltiche Lehrmittelzentrale,
Zollikofen, Schweiz.
Swofford, D.L. (2000) PAUP* Phylogenetic Analysis Using
Parsimony (* and Other Methods). Sinauer Associates,
Sunderland, MA.
Wapshere, A.J. (1974) Host specificity of phytophagous organisms and the evolutionary centres of plant genera or
sub-genera. Entomophaga 19(3), 301309.
Withers, T.M. (1999) Towards an integrated approach to
predicting risk to non-target species. In: Withers, T.M.,
Barton Browne, L. and Stanley, J. (eds) Host specificity
testing in Australasia: towards improved assays for biological control. CRC for Tropical Pest Management, Brisbane, Australia, pp. 9398.
van Klinken, R.D. and Edwards, O.R. (2002) Is hostspecificity of weed biological control agents likely to
evolve rapidly following establishment? Ecology Letters
5, 590596.
417
Population structure of an inadvertently
introduced biological control agent of
toadflaxes: Brachypterolus pulicarius
in North America
R.A. Hufbauer1 and D.K. MacKinnon1,2
Summary
Phytophagous insect species that use more than one host plant often harbour distinct populations
specialized on different host species. Brachypterolus pulicarius L. (Kateridae) is considered to be a
biological control agent of Dalmatian and yellow toadflax, Linaria dalmatica (L.) P. Mill. and Linaria
vulgaris P. Mill., in North America. We evaluated the population structure of this beetle using microsatellite loci to determine whether beetles collected from the two hosts differed. We found no significant microsatellite variation attributable to host plant. These results corroborate previous ecological
results showing that beetles collected from both hosts have similar preferences: They prefer yellow
toadflax and generally perform better on it as well.
Keywords: Linaria, plantinsect interactions, microsatellite.
Introduction
Many phytophagous insect species that feed on more
than one host are comprised of genetically differentiated populations that are specialized on individual
host plants or even genotypes of host plants (Fox and
Morrow, 1981; Mopper and Strauss, 1998). In biological control, such specialization has the potential to increase efficacy and perhaps safety when there is a good
match between the target weed, which is the host plant,
and the biological control agent (Goolsby et al., 2004,
2006). Alternatively, specialization may inhibit successful control if there is not a good match (Lym et al.,
1996; Lym and Carlson, 2002) or if the plant has the
upper hand in the evolutionary arms race with its enemies so that it has defences specific to genotypes of
enemies attacking it (Kaltz et al., 1999).
Colorado State University, Department of Bioagricultural Science and
Pest Management and Graduate Degree Program in Ecology, Fort Collins, CO 80523-1177, USA.
2
Current address: USDA PPQ APHIS 2301 Research Blvd, Suite 108,
Fort Collins, CO 80526, USA.
Corresponding author: R.A. Hufbauer <[email protected]>.
CAB International 2008
1
Brachypterolus pulicarius L. (Kateridae) is a flowerfeeding beetle that attacks Dalmatian and yellow toadflax, Linaria dalmatica (L.) P. Mill., spp. dalmatica
and Linaria vulgaris P. Mill., (Scrophulariaceae).
It was inadvertently introduced into North America
along with its host plants. These two toadflax species
were initially brought to North America from Eurasia for their ornamental and medicinal properties but
have since become invasive weeds. B. pulicarius, now
considered to be a biological control agent, is able to
reduce the seed set of both hosts dramatically under controlled conditions (McClay, 1992; Grubb et al., 2002).
It is common at high densities on yellow toadflax and
is thought to contribute to successful biological control of that weed (MacKinnon et al., 2005). It is not
always present on Dalmatian toadflax, however, and
when it is present, it is often found only at low densities (MacKinnon et al., 2005, 2007). MacKinnon et al.
(2005, 2007) have investigated the preference and performance of B. pulicarius on both hosts to determine
whether this species is comprised of populations specialized on each host. They found that beetles collected
from both hosts generally prefer yellow toadflax and
also perform better on yellow toadflax. In this paper,
we follow up on previous work by evaluating whether
418
Population structure of an inadvertently introduced biological control agent of toadflaxes
the molecular population structure of B. pulicarius in
North America provides evidence for genetic differentiation on the two hosts.
Methods and materials
We collected B. pulicarius from six infestations of
Dalmatian toadflax and eight of yellow toadflax in
western North America (Table 1). Additionally, we obtained samples from two European sites, one of each
host plant (Table 1). Beetles were stored in 95% ethanol in a -20 or -80C freezer before DNA extraction.
We extracted genomic DNA using DNeasyTM tissue
kits (Qiagen). To develop microsatellite markers, genomic DNA was digested with ApoI and BstYI. The
genomic library was constructed by ligating sizeselected fragments (7001600 bp) into pUC 19 that had
been digested with BamHI and EcoRI and gel-purified.
Vectors were transformed into competent Escherichia
coli cells (Life Technologies Max Efficiency DH5a)
and plated on LuriaBertani/ampicillin agar. Colonies
were lifted with Magna Graph membranes (Osmonics,
Inc.) and screened with a 33P-labeled probe of pooled
oligonucleotides (AC and AG dimers and all possible
trimers excluding homopolymers). Positive colonies
were polymerase chain reaction (PCR) amplified with
M13 forward or reverse primers and sequenced with a
BigDye Terminator Cycle Sequencing Kit and an ABI
377 automated sequencer (PE Applied Biosystems).
Primers were designed for 30 loci (GenBank accession
nos. EU078572EU078591), only four of which pro-
Table 1.
duced reliable and relatively easy to score amplification products (Table 2).
Microsatellite loci were amplified using PCR Express thermocyclers (Hybaid) in 10 l reactions containing 1 l genomic DNA, 1 PCR buffer (20 mM
TrisHCL, pH 8.4, 50 mM KCl), 2 mM MgCl2, 0.2 mM
each deoxyribonucleotide triphosphate, 2 pmol of each
primer, 0.5 units Taq polymerase (Life Technologies)
and 0.1 l TaqStart antibody (Clonetech). Amplification cycle conditions consisted of approximately 1 min
at 90C and then 35 cycles of 50 s at 95C, 1 min at
annealing temperature (Table 2), 1.5 min at 72C and
then a final extension step for 45 min at 72C. Reactions were held at 04C before separation in an ABI
3100 capillary instrument.
We ran basic statistics to evaluate HardyWeinberg
equilibrium and linkage disequilibrium on the microsatellite data using GenePop on the Web (http://genepop.
curtin.edu.au/; Raymond and Rousset, 1995). We looked
for evidence that a bottleneck in population size has
reduced the number of rare alleles using both the stepwise and infinite allele mutation models in the program Bottleneck (Cornuet and Luikart, 1996; Piry et
al.,1999). We evaluated population structure between
regions (North America and Europe) and within North
America due to sampling location and host plant, with
analyses of molecular variance (AMOVA) in Arlequin
version 3.01 (Excoffier et al., 2005). We evaluated
whether there is a relationship between geographic
and genetic distance (isolation by distance) by implementing a Mantel test in IBDWS (http://www.bio.sdsu.
ollection locations, host plant, approximate GPS coordinates, and sample sizes (n)
C
for Brachypterolus pulicarius.
Plant species location
Dalmatian toadflax
Christina Lake, BC, Canada
Kamloops, BC, Canada
Cheyenne, WY
Gillette, WY
Steamboat Springs, CO
Necedeh, WI
Macedoniab
Yellow toadflax
Donald, BC, Canada
Edmonton, AB, Canada
Rosalind, AB, Canada
Afton, WY
Poudre Park, CO
Estes Park, CO
Steamboat Springs, CO
Buckhorn Park, WI
Rhine Valley, Germany2
Coordinatesa
49.0300N, 118.1200W
50.4200N, 120.2300W
41.0612N, 104.5322W
44.1734N, 105.3019W
40.2927N, 106.4903W
44.0100N, 90.0400W
22
23
28
23
45
24
11
51.2900N, 117.0900W
53.2902N, 113.2951W
52.4700N, 112.2600W
42.4501N, 110.5717W
40.4200N, 105.1600W
40.2306N, 105.3233W
40.2957N, 106.4915W
43.5600N, 89.5900W
26
17
19
11
12
37
17
21
9
Coordinates without minutes (indicated as 00) are approximate.
European samples were provided without precise locality data.
419
XII International Symposium on Biological Control of Weeds
Table 2.
haracteristics of four microsatellite loci isolated from Brachypterolus pulicarius including locus name (clone
C
number) and GenBank accession number, primer sequences, PCR annealing temperature (Ta), repeat motif in
cloned allele, the size of the sequenced allele, the number of alleles found, the size range of the amplified alleles, ob
served heterozygosity (HO) and expected heterozygosity (HE). The top primer was dye-labeled for visualization.
Locus
(accession)
Primer sequences
(53)
AACTGACCAGCGTTAAATGATAAT
AGAGTGAATATTGTCCCTTCTCAA
ATTATCAGCTCCACAGAAAACACC
ATATAAGTTCACGTTCGGGGTTTG
TGAGGCCAACTAAACTTCAGA
GACTCGAGGGCAGATACAATC
ACTGCCAAACCAAGTCCAAAACT
GTTGGTTGCTTTCTCGGC
(EU078586)
16
(EU078591)
19
(EU078590)
37
(EU078589)
Ta (C)
Repeat of
cloned allele
Size (bp)
Number
of alleles
Size range
(bp)
HO
HE
60
(GA)13
304
317325
0.15
0.31
55
(AG)15
263
235263
0.26
0.31
55
(GT)8
139
11
125157
0.20
0.37
65
(CT)22
331
15
335379
0.08
0.35
edu/pub/andy/IBD.html; Bohonak, 2002; Jensen et al.,
2005).
Results and discussion
None of the four loci were in HardyWeinberg equilibrium: All showed a significant deficit of heterozygotes (P < 0.0001 across all loci and populations,
Table 2). This indicates that at least one evolutionary
process (selection, drift, migration and non-random
mating) is influencing the populations sampled. While
this is not in itself surprising, particularly for an introduced species, it is notable that the pattern is consistent
across locations, given that it takes only a single generation for HardyWeinberg equilibrium to be established. In addition, two of the loci (16 and 19) were in
linkage disequilibrium (P < 0.0001), suggesting physical linkage, selection or non-random mating.
Despite the deficit of heterozygotes, the program
Bottleneck generally revealed an L-shaped distribution
of alleles for all loci and population combinations. This
suggests that a bottleneck in population size associated
with the introduction of the beetles into North America
did not last long enough to lead to the loss of substantial numbers of rare alleles or that multiple introductions occurred. Bottleneck should ideally be used with
20 or more loci, however; thus, the lack of evidence for
a bottleneck should be considered a preliminary find-
ing. With only four loci, the power of those tests was
low.
The lack of HardyWeinberg equilibrium breaks the
basic assumptions of AMOVA. While AMOVA is fairly
robust to such issues, the following analyses should be
interpreted cautiously. The AMOVA comparing the native and introduced range showed significant differentiation between North American and European sample
locations (Table 3). This differentiation suggests that
the inadvertent introduction of B. pulicarius into North
America was from somewhere other than the collection
sites we tested from the Rhine Valley and Macedonia.
Alternatively, differences in the selective regime or
other evolutionary processes could have led to differentiation of North American and European samples.
The AMOVA examining the variation within North
American samples showed that most of the variation
was within populations, as is typical for microsatellite
loci (Table 3). There were also significant differences
among collection locations, revealing significant population structuring. Despite this, the Mantel test found no
relationship between geographic and genetic distance
across North America (r = 0.13, P = 0.18), suggesting
either high mobility or insufficient time for a balance
between drift and migration to establish, both of which
are likely in this system.
Finally, AMOVA provided no evidence that beetles
from the two host plants were genetically differentiated
420
Population structure of an inadvertently introduced biological control agent of toadflaxes
Table 3.
Analyses of molecular variance (AMOVA).
Source of variation
df
Sums of squares
Variance components
Percent of variation
Partitioning of variation across European and North American samples, by region and location within region
Among regions
1
10.64
0.122
16.90*
Among locations within
13
15.86
0.014
1.96*
regions
Within locations
641
375.39
0.586
81.14**
Partitioning of variation within North America due to host plant at the collection location and among collection locations
Among host plants
1
1.27
-0.004
-0.45
Among locations within
12
28.06
0.033
3.43**
hosts
Within locations
614
564.41
0.919
97.02**
* P<0.01, ** P<0.001.
(Table 3). This finding fits well with previous results
from this system, which showed that beetles from both
hosts tend to prefer and perform better on yellow toadflax than Dalmatian toadflax (MacKinnon et al., 2005,
2007). Thus, it appears that B. pulicarius is not comprised of distinct groups specialized on alternate hosts
(Fox and Morrow, 1981) but, rather, is a yellow toadflax specialist, and use of Dalmatian toadflax is merely
incidental. This likely will prevent B. pulicarius from
contributing in a meaningful way to the control of Dalmatian toadflax.
Acknowledgements
We thank Rosemarie De Clerck-Floate and Alec McClay for Canadian samples, Andrew Norton for Wisconsin samples and Robert Nowierski for European
samples. Comments from Hariet Hinz and Paul Hatcher
improved the manuscript. Many thanks to Steve Bogdanowicz for cloning the microsatellite loci.
References
Bohonak A.J. (2002) IBD (isolation by distance): a program
for analyses of isolation by distance. Journal of Heredity
93, 153154.
Cornuet J.M. and Luikart G. (1996) Description and power
analysis of two tests for detecting recent population
bottlenecks from allele frequency data. Genetics 144,
20012014.
Excoffier L., Laval G. and Schneider S. (2005) Arlequin ver.
3.0: an integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1,
4750.
Fox L.R. and Morrow P.A. (1981) Specialization: species
property or local phenomenon? Science 211, 887891.
Goolsby O.A., Zonneveld R. and Bourne A. (2004) Prerelease assessment of impact on biomass production of
an invasive weed, Lygodium microphyllum (Lygodiaceae:
Pteridophyta), by a potential biological control agent,
Floracarus perrepae (Acariformes:Eriophyidae). Environmental Entomology 33, 9971002.
Goolsby J.A., De Barro P.J., Makinson J.R., Pemberton
R.W., Hartley D.M. and Frohlich D.R. (2006) Matching
the origin of an invasive weed for selection of a herbivore
haplotype for a biological control programme. Molecular
Ecology 15, 287297.
Grubb R.T., Nowierski R.M. and Sheley R.L. (2002) Effects
of Brachypterolus pulicarius (L.) (Coleoptera:Nitidulidae) on growth and seed production of Dalmatian toadflax, Linaria genistifolia ssp. dalmatica (L.) Maire and
Petitmengin (Scrophulariaceae). Biological Control 23,
107114.
Jensen J.L., Bohonak A.J. and Kelley S.T. (2005) Isolation by
distance, web service. Bmc Genetics 6, 13.
Kaltz O., Gandon S., Michalakis Y. and Shykoff J.A. (1999)
Local maladaptation in the anther-smut fungus Microbotryum violaceum to its host plant Silene latifolia: Evidence from a cross-inoculation experiment. Evolution 53,
395407.
Lym R.G. and Carlson R.B. (2002) Effect of leafy spurge
(Euphorbia esula) genotype on feeding damage and reproduction of Aphthona spp.: implications for biological
weed control. Biological Control 23, 127133.
Lym R.G., Nissen S.J., Rowe M.L., Lee D.J. and Masters
R.A. (1996) Leafy spurge (Euphorbia esula) genotype
affects gall midge (Spurgia esulae) establishment. Weed
Science 44, 629633.
MacKinnon D.K., Hufbauer R.A. and Norton A.P. (2005)
Host-plant preference of Brachypterolus pulicarius, an
inadvertently introduced biological control insect of toadflaxes. Entomologia Experimentalis et Applicata 116,
183189.
MacKinnon D.K., Hufbauer R.A. and Norton A.P. (2007)
Evaluating host use of an accidentally introduced herbivore on two invasive toadflaxes. Biological Control 41,
184189.
McClay A.S. (1992) Effects of Brachypterolus pulicarius (L)
(Coleoptera, Nitidulidae) on flowering and seed production of common toadflax. Canadian Entomologist 124,
631636.
Mopper S. and Strauss S.Y. (1998) Genetic Structure and Local Adaptation in Natural Insect Populations: Effects of
Ecology, Life History, and Behavior. Chapman and Hall,
New York, NY, 449 pp.
Piry S., Luikart G. and Cornuet J.M. (1999) BOTTLENECK:
A computer program for detecting recent reductions in
the effective population size using allele frequency data.
Journal of Heredity 90, 502503.
Raymond M. and Rousset F. (1995) An exact test for population differentiation. Evolution 49, 12801283.
421
Genetic analysis of native and
introduced populations of Taeniatherum
caput-medusae (Poaceae): implications
for biological control
S.J. Novak1,2 and R. Sforza3
Summary
Genetic analysis of both native and introduced populations of invasive species can be used to exam
ine population origins and spread. Accurate delineation of an invasive species source populations
can contribute to the search for specific and effective biological control agents. Medusahead, Taeniatherum caput-medusae (L.) Nevski, a primarily self-pollinating Eurasian annual grass that was
introduced in the western USA in the late 1800s, is now widely distributed in California, Idaho,
Nevada, Oregon, Utah and Washington. The goal of our current research is to assess introduction
dynamics and range expansion of this grass in the western USA, and to identify source populations in
the native range to facilitate the search for potential biocontrol agents. Across introduced populations,
nine multilocus genotypes were detected, and we suggest a minimum of seven separate introduction
events of T. caput-medusae in the western USA. Although range expansion appears to have occurred
primarily on a local level, several introduced populations appear to be composed of admixtures of
introduced genotypes. None of the native populations analysed to date possess the exact multilocus
genotypes detected in introduced populations. We have recently begun screening Eurasian popula
tions using intersimple sequence repeat (ISSR) genetic markers to determine whether this polymerase
chain reactionbased technique can provide a higher degree of resolution for the identification of
source populations.
Keywords: invasive grass, multilocus genotypes, multiple introductions.
Introduction
Experimental analyses of both native and introduced
populations of invasive species can be used to assess
various ecological, genetic and evolutionary aspects of
invasion and the invasion process (Hierro et al., 2005;
Novak, 2007). For instance, comparison of the level
and structure of genetic diversity within and among
native and introduced populations can be used to de
termine whether the distribution of a species in its new
range stems from single or multiple introduction events
Current address: CSIRO European Laboratory, Campus International
de Baillarguet, 34988 Montferrier-sur-Lez, France.
2
Permanent address: Boise State University, Department of Biology,
1910 University Dr., Boise, ID, 83725-1515, USA.
3
USDA-ARS, European Biological Control Laboratory, Campus Inter
national de Baillarguet, 34988 Montferrier-sur-Lez, France.
Corresponding author: S.J. Novak <[email protected]>.
CAB International 2008
1
(Novak and Mack, 2001, 2005; Lavergne and Molof
sky, 2007). Additionally, genetic analysis of native and
introduced populations can be used to examine popula
tion origins and spread (Roderick and Navajas, 2003).
Accurate delineation of an invasive species source
populations (or regions) can contribute to the search for
biological control agents. Indeed, the identification of
areas of origin may reduce the economic cost of pros
pecting for agents and may result in the development
of more specific and effective biological control agents
(Goolsby et al., 2006a).
Taeniatherum caput-medusae (L.) Nevski, a mem
ber of the tribe Triticeae in the grass family, is consi
dered a noxious weed in many western US states (e.g.
Colorado, California, Oregon, Nevada and Utah). The
grass was first collected in the USA in Roseburg, Or
egon in 1887 (Fig. 1), and its collection history is well
documented (McKell et al., 1962; Young, 1992). The
grass has now invaded millions of hectares of semi-arid
422
Genetic analysis of native and introduced populations of Taeniatherum caput-medusae (Poaceae)
Figure 1.
Chronology of the spread of Taeniatherum caput-medusae in the western USA. This chronology was reconstructed using published accounts (see text), herbarium specimens and historical
rangeland in the western USA (Young, 1992; Miller et
al., 1999). It primary occurs in areas disturbed by over
grazing and fire in the 25100 cm annual precipitation
zone, and it can become the dominant plant species at
certain sites (Hironaka, 1961; Dahl and Tisdale, 1975;
Young, 1992). Ominously, the species has probably not
yet reached its ecological limit. If its ecological require
ments approximate those of Bromus tectorum L., it has
the potential to spread widely in the Great Basin of the
USA and beyond. Different methods have been tried to
control T. caput-medusae (burning, grazing, competi
tion and herbicides), and all have generally resulted in
failure (Horton, 1991).
The native range of T. caput-medusae includes
much of Eurasia, where three distinct subspecies have
been recognized (Frederiksen, 1986), but only T. caputmedusae ssp. asperum is believed to have been intro
duced into the USA (Young, 1992). Recently, foreign
exploration was carried out for identifying candidates for
biological control, and several plant pathogens were de
scribed, including the fungi, Ustilago phrygica Magnus
and Tilletia bornmuelleri Magnus (Siegwart et al., 2003;
Widmer and Sforza, 2004). A preliminary host range
screening with U. phrygica, a systematic smut fungi that
was collected in Turkey and attacks T. caput-medusae,
has been conducted (Sforza et al., 2004).
Because natural enemy pressure can vary across
genotypes (Evans and Gomez, 2003), populations and re
gions, the enemy release hypothesis is best tested by com
paring introduced populations with native populations
423
XII International Symposium on Biological Control of Weeds
that were the source of the invasion. When implement
ing biological control programs to reverse the release
from natural enemies, knowledge of source populations
or regions would increase the probability of finding
highly specialized enemies (Goolsby et al., 2006b). In
addition, evaluating the efficacy of candidate biologi
cal control agents should be done on the full range of
genotypic diversity present within the introduced range
(Gaskin et al., 2005), through the evaluation of specific
polymerase chain reaction (PCR) primers when screen
ing populations (Marrs et al., 2006).
The overall goals of our current research are to as
sess introduction dynamics and range expansion of
T. caput-medusae in the western USA and to identify
source populations in the species native range to fa
cilitate the search for potential biocontrol agents. This
research specifically addresses the following questions.
(1) How many multilocus genotypes occur in western
US populations of T. caput-medusae, and what is their
geographic distribution? (2) What does the distribution
of these genotypes indicate about range expansion of
this species in its new territory? (3) How many geno
types occur in native range populations, and what is
their geographic distribution? (4) Can source popula
tions for the invasion of the grass in the western USA
be identified?
Methods and materials
pulations were brought back to the United States De
partment of Agriculture Agricultural Research Service
European Biological Control Laboratory (USDA-ARSEBCL) in Montpellier, France, on an official authoriza
tion (04LR011) granted by the French government. Seeds
were stored in a quarantine greenhouse at the EBCL until
further use.
Enzyme electrophoresis
The level and structure of genetic diversity within
and among populations of T. caput-medusae in its
invasive range in the western USA is based on the
analysis of 1663 individuals from 45 populations. In
the laboratory, one seed from each individual in a po
pulation was germinated on moistened filter paper in
a Petri dish and harvested approximately 7 days after
germination. Enzyme electrophoresis was conducted
generally following the methods of Soltis et al. (1983),
with modifications described by Novak et al. (1991).
The 15 enzymes employed in this study were resolved
with enzyme electrophoresis using four buffer systems
(1, 6, 8 and 9), and these 15 enzymes were genetically
encoded by 29 loci. Because T. caput-medusae is a dip
loid with low genetic diversity, the genetic basis of all
allozyme variation observed was easily inferred based
on known subunit structure and compartmentalization
of these enzymes (Weeden and Wendel, 1989).
Sampling of plant material
ISSR analysis
One objective of our sampling has been to collect
plant material from populations across the entire geo
graphic distribution of T. caput-medusae, in both its
invasive and native ranges. Another objective has been
to obtain population samples at or near localities where
the plant was first collected, or reported, during its in
vasion of the western USA (Fig. 1). Samples have been
collected from a total of 45 populations in the states of
California, Idaho, Nevada, Oregon, Utah and Washing
ton, with several early collection localities represented.
Two groups of native range samples have been
included in this study. The first group consisted of
23 populations, with 22 of these populations being ac
cessions obtained from the USDA Plant Introduction
Laboratory in Pullman, WA: 12 populations from Tur
key, seven from Afghanistan, two from Iran and one
from Kazakhstan. The personnel of the USDA Plant
Introduction Laboratory variously classified these ac
cessions to each of the three subspecies of T. caputmedusae. Only one of these 23 populations originated
in Europe: Sterea Hellas, Greece. The second group
consisted of 49 populations collected in August or Sep
tember, 2002 and 2003, from across the grasses native
range in Eurasia. For most populations, intact spikes
were collected from 30 to 40 individual plants along
a transect at approximately 1-m intervals and placed
in separate envelopes. Seeds from the 49 Eurasian po
Five of the 49 Eurasian populations of T. caputmedusae mentioned above were selected for a prelimi
nary analysis using intersimple sequence repeat (ISSR)
genetic markers: one population from Morocco, Spain,
France, Greece (Crete) and Turkey. For each popula
tion, five seeds were randomly selected from each of
three plants located 5, 18 and 25 m along the transect
from which they were sampled. In the EBCL quaran
tine greenhouse, seeds were germinated in Petri dishes
with distilled water at 25C, 80% relative humidity and
16:8 h light/dark. Ten days after germination, leaves
were removed from the plants and frozen at -20C.
Total genomic DNA was extracted from frozen leaf
material using DNeasy Plant Mini Kits according to
the manufacturers instructions (Qiagen Inc., Valen
cia, CA). After extraction, DNA was amplified with
the PCR using six ISSR primers decribed by Wolfe
et al. (1998). Primer names and sequences are provi
ded in Table 1. DNA amplifications were performed
in 20 l final reaction volumes containing 1 U of Taq
DNA polymerase (Qiagen Inc.), 1 buffer (Qiagen),
1 mM MgCl2, 0.2 mM of each deoxyribonucleotide
triphosphates, 0.5 M of a single primer and 2 l of the
template DNA. Amplifications were performed using
the GeneAmp PCR System 9700 (Applied Biosys
tems, Forest City, CA) as follows: 94C for 3 min, then
35 cycles at 94C for 30s, 45C for 45 s and and 72C
424
Genetic analysis of native and introduced populations of Taeniatherum caput-medusae (Poaceae)
Table 1.
I dentity and nucleotide sequences of the ISSR primers used in this preliminary analysis
of Taeniatherum caput-medusae from its native range. ISSR primers used in this study
were described by Wolfe et al. (1998). The utility of each primer, based on the criteria
described in the text, is indicated: Y yes, N no.
Primer
ISSR-17898A
ISSR-17898B
ISSR-17899A
ISSR-17899B
ISSR-814
ISSR-HB15
Primer sequence
(CA)7AC
(CA)7GT
(CA)7AG
(CA)7GG
(CT)8TG
GTGGTGGTGGC
for 1 1/2 min. A final extension was performed at 72C
for 7 min. Amplified products were electrophoresed on
1.0% agarose gels. Gels were stained with ethidium
bromide, and DNA fragments were visualized with a
UV transilluminator and photographed.
Data analysis
Allozyme multilocus genotypes were identified
from enzyme electrophoresis data, and these genotypes
were used to assess introduction dynamics and spread
of T. caput-medusae in the western USA and to identify
source populations of the grass in its native range. Al
lozyme multilocus genotypes are defined as the com
posite genotype over all loci examined and therefore
are designated based on the identity of alleles at each
scored enzyme locus. Populations were defined as ge
netically polymorphic if they contained two or more
multilocus genotypes. As part of our preliminary anal
ysis of native populations of T. caput-medusae using
ISSR genetic markers, bands were not scored; however,
we did qualitatively assess each primer to determine
its utility for future analysis. Specifically, primers were
evaluated based on whether they (1) did not generate
bands in control reactions, (2) generated clear, distinct,
darkly stained bands and (3) were polymorphic among
test populations.
Results
Multilocus genotypes in the
introduced range
Multilocus genotypes are named based on the
populations in which they were first found. A total of
nine multilocus genotypes were detected among all
45 populations: seven homozygous multilocus geno
types and two genotypes with one or two heterozygous
loci (unpublished data, not shown). The seven homo
zygous genotypes were first detected in Roseburg, OR,
Steptoe Butte, WA, Rattlesnake Station, ID, Ladd Can
yon, OR, Pullman, WA, Malloy Prairie, WA and Salt
Creek, UT. Five different multilocus genotypes were
Utility
Y
N
Y
Y
Y
N
observed in the Palouse region of eastern Washington.
The multilocus genotypes detected in Pullman, Malloy
Prairie and Salt Creek appear to be restricted to just a
single population. Heterozygous multilocus genotypes
were found in two different populations: White Bird,
ID contained two heterozygous genotypes and one of
these genotypes was also detected at Emigrant Hill, OR.
The level of polymorphisms within introduced popula
tions was low: Only 17 of 45 populations (37.8%) con
tained two or more multilocus genotypes. Furthermore,
of these 17 polymorphic populations, only three con
tained three or more multilocus genotypes.
Multilocus genotypes in the native range
Two distinct categories of multilocus genotypes
were detected within the 23 Eurasian populations. The
first group of genotypes were quite distinct and diffe
red from those detected in the western USA at multiple
loci. These genotypes were present in seven populati
ons from Turkey, and all populations from Afghanistan,
Iran and Kazakhstan. Based on their different enzyme
banding patterns, plus their larger seed size, these pop
ulations probably all consisted of T. caput-medusae
ssp. crinitum. The enzyme multilocus genotypes de
tected in the remaining populations in Turkey and the
one from Greece were similar to those observed in the
western USA, and these populations were tentatively
assigned to T. caput-medusae ssp. asperum. However,
none of the multilocus genotypes present in these
23 native populations was an exact match to those pre
viously detected in the introduced range.
Preliminary ISSR analysis
Of the six ISSR primers that were screened in our
preliminary analysis of native populations of T. caputmedusae, four met our selection criteria and will prove
useful in future studies of both native and introduced
populations (Table 1). Samples from the five countries
(populations) display different DNA banding patterns
with primer ISSR-17899A (Fig. 2), although no variabil
ity was detected among individuals within populations.
425
XII International Symposium on Biological Control of Weeds
Figure 2.
Photograph of DNA banding patterns obtained using ISSR primer 17899A. Note the different banding
patterns for populations of Taeniatherum caput-medusae from different countries. Contents of the lanes
on this gel are as follows: 1 1 kb ladder; 24 France; 5 and 6, Greece; 7 and 8, Morocco; 911, Spain;
1214, Turkey; 15 PCR control; 16 extraction control.
Discussion
Introduction dynamics and spread
in the western USA
The level of genetic diversity observed across and
within western US populations of T. caput-medusae
is lower than the mean value reported for other selfpollinating plant species (Hamrick and Godt, 1990) but
similar to that of other invasive plants that exhibit a uni
parental mode of reproduction such as selfing (Novak
et al., 1991). Yet, despite its lack of genetic diversity,
at least at the loci examined in this study, this species
is now invasive over much of the semi-arid portions of
the western USA.
The occurrence and geographic distribution of mul
tilocus genotypes can provide insights into introduction
dynamics and spread of invasive species (Novak and
Mack, 2001, 2005). Multilocus genotype results for
western US populations of T. caput-medusae are con
sistent with the pattern often associated with multiple
introductions. Based on just the number of homozygous
multilocus genotypes detected across all populations,
we suggest a minimum of seven independent founder
events. This conclusion is bolstered by the observation
that four of the localities where these genotypes were
detected are at or near early collection sites of the plant:
Roseburg (1887), Steptoe Butte (1901), Rattlesnake
Station (1930) and Ladd Canyon (1944).
The detection of five different multilocus genotypes
in eastern Washington suggests that multiple introduc
tions can occur within a relatively small geographic
area. Because the plant was not collected in Utah until
1988, the detection of a unique multilocus genotype in
Salt Creek, Utah may be evidence for a relatively re
cent introduction event. If so, these data indicate that
introduction of this grass is ongoing. Our results for
T. caput-medusae join a growing body of information
indicating that, for invasive plant species, multiple in
troduction may be the rule rather than the exception
(Novak and Mack, 2005) and may contribute to inva
siveness (Allendorf and Lundquist, 2003; Lavergne
and Molofsky, 2007; Novak, 2007).
The low level of polymorphisms observed within
introduced populations of T. caput-medusae indicates
that gene flow or dispersal among these populations is
low. Thus, we conclude that spread or range expansion
of the species has occurred mostly at the local or re
gional level and certainly has not been widespread.
However, the detection of populations that appear to be
admixtures of different introduced genotypes, as seen at
several locations in the western USA (data not shown),
suggests that intermixing of genotypes can take place
if multiple introductions have occurred within the same
region. Moreover, the detection of heterozygous mul
tilocus genotypes suggests that plants with different
genotypes and potentially originating from different
introduction events have recently mated. Such hybrid
ization events have been suggested to contribute to
increased invasiveness in introduced species (Ellstrand
and Schierenbeck, 2000).
Source populations
Although the multilocus genotyes observed in po
pulations from Greece and several from Turkey are
similar to those of the western USA, no exact matc
hes were found among native populations. Thus, our
allozyme analysis did not reveal the source populations
(or regions) for the introduction of T. caput-medusae
in the USA, but the data clearly excludes many of the
southwest and central Asian locations from serving as
426
Genetic analysis of native and introduced populations of Taeniatherum caput-medusae (Poaceae)
source populations. These results probably stem from
insufficient sampling in the native range, especially
Europe: Only one of 23 native populations was from
Europe, and the other 22 populations were collected
from southwest or central Asia.
Additional analysis of European populations will be
required before source populations, and introduction
dynamics of the invasion of T. caput-medusae in the
western USA can be more confidently described. To
this end, our future plans include allozyme analysis
of the 49 additional Eurasian populations, which have
already been collected. In addition, our preliminary
analysis of native range populations of the grass using
ISSR genetic markers is very promising and will hope
fully provide us with PCR-based markers that posses
ses high degrees of polymorphism and resolution for
identifying source populations.
Prospects for biological control
Foreign exploration for the identification of pos
sible biological control agents has already led to the
identification of several promising plant pathogens
(Siegwart et al., 2003; Widmer and Sforza, 2004). The
genetic analyses described in this paper are meant to
compliment this effort, and results of these analyses
reveal the likelihood for biological control of T. caputmedusae. Introduced populations of the grass are ge
netically depauperate; thus, we would anticipate fast
population build-up and spread of highly adapted bio
control agents (Mller-Schrer et al., 2004). However,
multiple introductions, the occurrence of some intro
duced populations that are genetic admixtures and the
detection of low level of outcrossing within a few pop
ulations means that several biological control agents
from different portions of the native range may be
required (Burdon and Marshall, 1981). Thus, the ac
curate identification of source populations is needed to
augment the exploration for biological control agents
already taking place and may result in the development
of more specific and effective biological control agents
for this highly destructive invasive species.
Acknowledgements
We gratefully thank Marie-Claude Bon and Corinne
Hurard at the EBCL for their assistance with the ISSR
analysis. This work would not have been conducted
without the support of Walker Jones and Mic Julien.
The allozyme work described here was done in collabo
ration with Dean Marsh, Joseph Rausch, Lynell Dienes,
Kelly Burden, Kevin Hansen and Matt Score at Boise
State University. Funding for this work was provided
by the EBCL, the M.J. Murdock Charitable Trust, the
Merck-AAAS Undergraduate Science Research Pro
gram and the Faculty Research Grant Program and De
partment of Biology at Boise State University.
References
Allendorf, F.W. and Lundquist, L.L. (2003) Introduction:
population biology, evolution, and control of invasive
species. Conservation Biology 17, 2430.
Burdon, J.J. and Marshall, D.R. (1981) Biological control
and the reproductive mode of weeds. Journal of Applied
Ecology 18, 649658.
Dahl, B.E. and Tisdale, E.W. (1975) Environmental factors
related to medusahead distribution. Journal of Range
Management 28, 463468.
Ellstrand, N.C. and Schierenbeck, K.A. (2000) Hybridization
as a stimulus for the evolution of invasiveness in plants?
Proceedings of the National Academy of Sciences of the
United States of America 97, 70437050.
Evans, K.J. and Gomez, D.R. (2004) Genetic markers in rust
fungi and their application to weed biocontrol. In: Ehler,
L.E., Sforza, R. and Mateille, T. (eds) Genetics, Evolution
and Biological Control. CABI Publishing, Wallingford,
UK, pp. 7395.
Frederiksen, S. (1986) Revision of Taeniatherum (Poaceae).
Nordic Journal of Botany 6, 389397.
Gaskin, J.F., Zhang, D.Y. and Bon, M.C. (2005) Invasion
of Lepidium draba (Brassicaceae) in the western United
States: distributions and origins of chloroplast DNA hap
lotypes. Molecular Ecology 14, 23312341.
Goolsby, J.A., van Klinken, R.D. and Palmer, W.A. (2006a)
Maximising the contribution of native-range studies to
wards the identification and prioritization of weed bio
control agents. Australian Journal of Entomology 45,
276286.
Goolsby, J.A., De Barro, P.J., Makinson, J.R., Pemberton,
R.W., Hartley, D.M and Frohlich, D.R. (2006b) Matching
the origin of an invasive weed for selection of a herbivore
haplotype for a biological control programme. Molecular
Ecology 15, 287297.
Hamrick, J.H. and Godt, M.J. (1990) Allozyme diversity in
plant species. In: Brown, A.D.H., Clegg, M.T., Kahler,
A.L. and Weir, B.S. (eds) Plant Population Genetics,
Breeding, and Germplasm Resources. Sinauer, Sunder
land, MA, pp. 4363.
Hierro, J.L., Maron, J.L. and Callaway, R.M. (2005) A bio
geographical approach to plant invasions: the importance
of studying exotics in their introduced and native range.
Journal of Ecology 93, 515.
Hironaka, M. (1961) The relative rate of root development of
cheatgrass and medusahead. Journal of Range Management 14, 263267.
Horton, W.H. (1991) Medusahead: importance, distribution
and control. In: James, L.F., Evans, J.O., Ralphs, M.H.
and Child, R.D. (eds) Noxious Range Weeds. Westview
Press, Boulder, CO, pp. 394398.
Lavergne, S. and Molofsky, J. (2007) Increased genetic varia
tion and evolutionary potential drive the success of an in
vasion. Proceedings of the National Academy of Sciences
of the United States of America 104, 38833888.
McKell, C.M., Robison, J.P. and Major, J. (1962) Ecotypic
variation in medusahead, an introduced annual grass.
Ecology 43, 686698.
Marrs, R.A., Hufbauer, R.A., Bogdanowicz, S.M. and
Sforza, R. (2006) Nine polymorphic microsatellite mark
ers in Centaurea stoebe L. (subspecies C. s. stoebe and
427
XII International Symposium on Biological Control of Weeds
C. s. micranthos (S. G. Gmelin ex Gugler) Hayek) and
C. diffusa Lam. (Asteraceae). Molecular Ecology Notes
6, 897899.
Miller, H.C., Clausnitzer, D. and Borman, M.M. (1999)
Medusahead. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State University Press, Corvallis, OR, pp. 271281.
Mller-Schrer, H, Schaffner, U. and Steinger, T. (2004) Evo
lution in invasive plants: implications for biological con
trol. Trends in Ecology and Evolution 19, 417422.
Novak, S.J., Mack, R.N. and Soltis, D.E. (1991) Genetic vari
ation in Bromus tectorum (Poaceae): population differen
tiation in its North American range. American Journal of
Botany 78, 11501161.
Novak, S.J. and Mack, R.N. (2001) Tracing plant introduc
tion and spread: genetic evidence from Bromus tectorum
(cheatgrass). Bioscience 51, 114122.
Novak, S.J. and Mack, R.N. (2005) Genetic bottlenecks in
alien species: influence of mating systems and introduc
tion dynamics. In: Sax, D.F., Stachowicz, J.J. and Gaines
S.D. (eds) Species invasions: insights into ecology, evolution and biogeography. Sinauer, Sunderland, MA,
pp. 210228.
Novak, S.J. (2007) The role of evolution in the invasion proc
ess. Proceedings of the National Academy of Sciences of
the United States of America 104, 36713672.
Roderick, G.K. and Navajas, M. (2003) Genes in new en
vironments: genetics and evolution in biological control.
Nature Reviews Genetics 4, 889899.
Sforza, R, Eken, C, Hayat, R. and Widmer, T.L. (2004) First
evaluation of Ustilago phrygica for the biological control
of Taeniatherum caput-medusae (Triticeae). In: Proceedings of the XIIeme colloque International sur la biologie
des mauvaises herbes, Dijon, France, 31 aot 2 sep
tembre 2004, pp. 407412.
Siegwart, M., Bon, M., Widmer, T., Crespy, N. and Sforza,
R. (2003) First report of Fusarium arthrosporioides on
medusahead (Taeniatherum caput-medusae) and prelimi
nary tests for host-specificity. Plant Pathology, 52, 416.
Soltis, D.E., Haufler, C.H., Darrow, D.C. and Gastony, G.L.
(1983) Starch gel electrophoresis of ferns: a compilation
of grinding buffers, gel and electrode buffers, and staining
schedules. American Fern Journal 73, 927.
Weeden, N.F. and Wendel, J.F. (1989) Genetics and plant
isozymes. In: Soltis, D.E. and Soltis, P.S. (eds) Isozymes in Plant Biology. Dioscorides Press, Portland, OR,
pp. 4672.
Widmer T.L. and Sforza R. (2004) Exploration for plant
pathogens against Taeniatherum caput-medusae (medusa
head). In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lons
dale, W.M., Morin, L. and Scott, J.K. (eds) Proceedings
of the XI International Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia,
pp. 193197.
Wolfe, A.D., Xiang, Q.-Y. and Kephardt, S.R. (1998) as
sessing hybridization in natural populations of Penstemon (Scrophulariaceae) using hypervariable intersimple
sequence repeat (ISSR) bands. Molecular Ecology 7,
11071125.
Young, J.A. (1992) Ecology and management of medusa
head (Taeniatherum caput-medusae ssp. asperum [Simk.]
Melderis). Great Basin Naturalist 52, 245252.
428
The use of surrogate herbivores for the
pre-release efficacy screening of biological
control agents of Lepidium draba
K.P. Puliafico,1 M. Schwarzlnder,1 H.L. Hinz2 and B.L. Harmon1
Summary
Pre-release efficacy assessment has been suggested as a primary selection criterion for potential biological control species to insure the success and safety of biological weed control. Pre-release efficacy
of candidate agents is commonly assessed by exposing insects to target plants in the garden or greenhouse (native range) or in quarantine (introduced range) before or in parallel with host-specificity
testing. Conducting pre-release impact experiments for several candidate agents simultaneously may
be difficult because potential agents are either scarce or may require development of novel culturing procedures. We propose an alternative approach to pre-release efficacy assessment that utilizes
oligophagous or polyphagous insect herbivores from the introduced range as surrogates for biological
control agents to assess the impact of specific feeding niches on the target weed to direct the search for
effective candidate agents. Surrogate herbivores can be cosmopolitan or indigenous insects collected
directly from the invasive plant and confamilial species. Insect pest species are particularly well
suited to act as surrogates because they are seasonally abundant or easily reared. Based on previous
surveys, we identified surrogate herbivores attacking different above-ground plant organs of the biological control target hoary cress (Lepidium draba L., Brassicaceae). We tested the density-dependent
impact of four oligophagous herbivore species on L. draba to demonstrate the applicability of this
novel efficacy assessment technique. We found that the endophagous stem miner, Ceutorhynchus
americanus Buchanan (Coleoptera: Curculionidae), had the highest per-capita effect on hoary cress
growth, suggesting candidate agents within this niche should be prioritized.
Keywords: herbivore niche, agent selection, generalist insects.
Introduction
The biological control of weeds has been a successful,
economical and environmentally sound management
tool for curbing plant invasions, but McFadyen (2003)
estimated that only 55% of biological control agents
established contribute to the suppression and control
of their target weeds. In addition, more than half of
the successful biological control programs can link
their success to a single agent (Denoth et al., 2002).
Introduction of ineffective biological control agents
increases the probability of direct risks to non-target
University of Idaho, Department of Plant, Soils, and Entomological
Sciences, Moscow, ID 83844-2339, USA.
2
CABI EuropeSwitzerland, 1 Rue des Grillions, Delmont 2800, Switzerland.
Corresponding author: K.P. Puliafico <[email protected]>.
CAB International 2008
plant species, interference with established agents and
indirect effects on ecosystem functions without the
benefits gained from reduction of weed dominance or
abundance. To minimize the cost of host-specificity
screening and risks associated with establishment and
proliferation of ineffective agents, improved methods
for predicting the impact of candidate agent species before their release are recommended (Sheppard, 2003;
McClay and Balciunas, 2005). Pre-release efficacy testing of candidates provides quantitative data to improve
the likelihood of selecting the agents most capable of
reducing weed abundance.
The most commonly used technique for pre-release
efficacy assessment is to expose candidate agents to
target plants in garden or greenhouse experiments in
their native range or under quarantine conditions in
their introduced range. The advantage of these methods is that changes in plant performance can be directly
attributed to agents at experimentally controlled densities. However, screening several candidate species
429
XII International Symposium on Biological Control of Weeds
simultaneously may be particularly difficult because
potential agents are often scarce initially or may require development of novel rearing protocols. Results
of field experiments conducted in the area of origin
may also be confounded by the use of weed genotypes
different from those in the invaded range, interference
by other herbivores unlikely to be utilized in biological
control and competitive interactions with plant species
that are not present in the invaded range.
Another technique that has recently been advocated
is the use of simulated herbivory for pre-screening insect efficacy (Raghu and Dhileepan, 2005; Schooler
et al., 2006). The advantage of simulated herbivory is
the ease of applying precise experimental treatments
(Hjltn, 2004) without the need for insect quarantine.
Although this approach has been shown to effectively
mimic insect feeding impact in some cases, plant responses to mechanical damage do not always match
the magnitude and quality of responses to herbivore
damage (Lehtil and Boalt, 2004). In addition, it can be
difficult to artificially simulate the diversity of insect
damage because of the unique spatial and temporal distribution of damage caused by each feeding mode.
The use of surrogate herbivores provides an alternative approach to pre-release efficacy testing that
combines the strengths of actual insect herbivory and
the convenience of simulated herbivory. Surrogate
herbivores can be used early in the development of a
biological control program for testing the sensitivity of
the target plant to damage inflicted by different feeding niches, which could then guide the search for new
insect agents. The protocols for subsequent pre-release
efficacy testing of candidate agents can be established
and the results compared with generalist herbivores of
the same niche. The main advantage of this approach
is that it can be conducted with resources abundant
in the invaded range of the weed without the expense
of quarantine facilities. Surrogate herbivores can be
indigenous or cosmopolitan insects that occur in the
introduced range of the weed. These insects are often
seasonally abundant and easily collected from the habitats invaded by the target weed or from agricultural
habitats. They may already utilize the target plant or
be generalist pests of confamilial plant species. Several
agricultural pest species have well-established massrearing protocols for laboratory colonies or are available for purchase.
We used surrogate herbivores to assess the potential
effects of candidate biological control agents on hoary
cress, Lepidium draba L. (Brassicaceae). An extensive
survey of L. draba insects by Cripps et al. (2006) indicated that several above-ground feeding guilds are
represented in North America by oligophagous native
and polyphagous pest species. To test the impact of
these species and to determine the most effective feeding guild for biological control, we investigated the
response of hoary cress to four different types of oligophagous herbivore damage at different densities.
Materials and methods
L. draba ssp. draba (=Cardaria draba (L.) Desv.) is a
Eurasian perennial mustard introduced to North America in the 19th century (Mulligan and Findlay, 1974).
L. draba is currently considered a noxious weed in 15
western states and three Canadian provinces (Rice,
2007). L. draba occurs in a wide range of habitats, including cultivated land, rangeland, wilderness, pastures,
roadsides and waste areas, but thrives particularly well
in disturbed, riparian or irrigated areas (Mulligan and
Findlay, 1974). It reproduces sexually and vegetatively
through rhizomes. Seeds usually germinate in early autumn and produce overwintering rosettes, which bolt in
spring and flower from April to June in the northwestern USA.
Surrogate insect herbivores were selected based on
the feeding modes of two specialist insects currently
under consideration for biological control of L. draba:
the stem-mining weevil, Ceutorhynchus merkli Korotyaev (Coleoptera: Curculionidae), and the stem and
root-crown feeding flea beetle, Psylliodes wrasei Leonardi & Arnold (Coleoptera: Chrysomelidae) (Cripps
et al., 2006). The native stem-miner, Ceutorhynchus
americanus Buchanan, was the only insect found mining in the shoots of L. draba in North America and was
used as a surrogate for C. merkli. Adult C. americanus
feed on foliage and oviposit in the stems of L. draba
in laboratory no-choice tests (K.P. Puliafico, unpublished data). The host affinity of this native weevil is
unknown, but adults have been collected from Brassica
and Lepidium species and it has only been successfully
reared from Lepidium virginicum (Buchanan, 1937).
The crucifer flea beetle, Phyllotreta cruciferae
(Goeze) (Coleoptera: Chrysomelidae), was selected as
a surrogate to mimic adult feeding of P. wrasei. Crucifer flea beetles are introduced Brassicaceae pests,
which cause crop damage primarily from feeding on
seedling plants but usually have only a minor effect on
established plants (Feeny et al., 1970).
Two additional insects were utilized as surrogates
to mimic feeding modes for which there are currently
no biological control candidate species: a lepidopteron defoliator and a piercing/sucking true bug. The
diamondback moth, Plutella xylostella (L.) (Lepidoptera: Plutellidae), is considered native to the Mediterranean region but has a cosmopolitan distribution and
occurs wherever crucifer crops are grown (Talekar and
Shelton, 1993). This defoliating moth is oligophagous
within the Brassicaceae (Talekar and Shelton, 1993)
and is abundant on L. draba in North America (Cripps
et al., 2006). The tarnished plant bug, Lygus hesperus Knight (Heteroptera: Miridae), is native to North
America and was one of the most abundant polyphagous species found on invasive L. draba (Cripps et al.,
2006). Lygus species lacerate plant tissue, inject salivary fluids that digest the tissue extra-orally and then
ingest the liquefied tissue (Butts and Lamb, 1990).
430
The use of surrogate herbivores
This study was conducted at the climate-controlled
Manis Entomological Greenhouse, University of Idaho,
Moscow, ID. Environmental conditions were maintained with a 15L/9D photoperiod at 24 2C (day)/18
1C (night) in the greenhouse throughout all phases
of the experiment except for the seedling vernalization
described below.
L. draba seeds were collected from a population
near Moscow, ID (4644N, 11658W) in August 2003.
Seeds were sown into Cornell Peat-Lite A artificial
potting medium (Hartmann et al., 1990) in the greenhouse. After formation of the first true leaves, seedlings
were transplanted into plug trays using the same potting medium. Seedlings were grown for 6 weeks in
the greenhouse and were then transferred to a 3 1C
cold room with 12L/12D photoperiod for a minimum
of 60 days to initiate bolting and flower production.
During vernalization seedlings were fertilized once a
week with 0.33 ml of Miracle-Gro per litre of water
(15-30-15 NPK, Scotts Miracle-Gro, Marysville, OH).
After cold treatment, plants were transplanted into 3-l
plastic pots in artificial potting media (1:2:1 peat moss,
vermiculite and perlite mix; augmented with 5% sand,
pH stabilizers and trace elements) and fertilized with
2.8 g Osmocote slow release fertilizer per litre soil
(14:14:14 NPK, Scotts Miracle-Gro, Marysville, OH).
Plants were then grown in the greenhouse for 6 to 8
weeks before they were used for experimentation.
C. americanus was collected from L. draba near
Vale, Oregon (4405N, 11718W) on 24 April 2004.
Females were tested individually for oviposition on cut
L. draba stems before the start of the experiment (Harmon and McCaffrey, 1997). P. cruciferae adults were
collected on 29 April 2004 from Sinapis alba seedlings
southeast of Genesee, ID (4633N, 11655W). Adult
and nymph L. hesperus were collected from a mixed
field of alfalfa and pasture grasses at the University of
Idaho Sheep Research Farm, Moscow, ID (4644N,
11658W) 2 days before the start of the experiment
on 12 August 2004. P. xylostella eggs were obtained
from Benzon Research Inc. (Carlisle, PA). A laboratory
colony was reared on artificial diet in 473-ml Styrofoam cups with plastic lids following the protocols of
Shelton et al. (1991). Approximately 300 adults of both
sexes were caged together, and females were allowed
to oviposit for 24 to 48 h on cabbage-juice-coated aluminum foil. Foil was then cut into 1 cm2 pieces; eggs
were counted and combined into treatment densities,
which then were placed at the base of the plants.
Stem number and individual stem lengths were
measured for all plants before the start of experiments.
Plants were assigned to herbivore treatments in a complete randomized block design with position on greenhouse bench as the blocking factor. Plants were caged
with 80-cm-long mesh sleeve cages supported with internal wire frames with the bottom of each sleeve held
in place with a metal hose clamp. Herbivores were released according to treatments (see below). All plants
were caged for 40 to 44 days and then destructively
harvested in blocks. Plants were clipped at the soil
surface, the numbers of shoots counted and individual
shoot lengths recorded to the nearest 1 cm. Roots were
carefully washed to remove soil. Root and shoot biomasses were recorded to the nearest 0.1 g after drying
for a minimum of 24 h at 80C.
The effects of density-dependent herbivory on plant
biomass, shoot number and height were determined in
two experiments using identical protocols. Insect densities were selected along an exponentially increasing
scale and chosen to encompass normal field densities
from survey data (Cripps et al., 2006). Experiment 1
started on 29 April 2004 with three insect species with
ten replicates for each herbivore and density treatment.
Treatments were comprised of either P. cruciferae
(0, 10, 20, 40 and 80 unsexed adults), P. xylostella (0,
75, 150, 300 and 900 eggs on foil) or C. americanus
(0, 1, 2, 4 and 8 adult females). Male C. americanus
(0, 1, 1, 2 and 4 individuals, respectively) were added
to the females to ensure continual fertilization of eggs.
Experiment 2 started on 12 August 2004 and tested the
impact of L. hesperus adults (0, 10, 20, 40 and 80) with
five replicates and nymphs (0, 20, 40, 80 and 160) with
seven replicates.
Data were analyzed using general linear models for
analysis of covariance with herbivore treatments as the
fixed factor and position on the greenhouse bench as the
random blocking factor. Pre-treatment shoot length was
log10-transformed and used as a co-variable for biomass and post-treatment shoot lengths. Pre-treatment
shoot number was used as a co-variable for post-treatment
shoot number. Upon a significant herbivore effect,
means were compared using a Tukey Honestly Significantly Different test, and per-capita herbivore effects
on shoot length were examined using regression analysis. All analyses were conducted using Minitab v15
(Minitab Inc., 2006).
Results
P. xylostella was the only herbivore that significantly
decreased L. draba above-ground (F4,35 = 40.09, P <
0.001) and below-ground (F4,35 = 47.17, P < 0.001)
biomass accumulation (Fig. 1). Damage to plants was
extensive at all treatment levels of diamondback moth
and resulted in total defoliation of plants at the highest
egg density. Caterpillars at the highest density starved
to death before pupation, allowing compensatory regrowth of plants before harvest.
Defoliation by P. xylostella (F4,35 = 6.54, P < 0.001),
shot-hole feeding by P. cruciferae adults (F4,35 = 3.42,
P = 0.018) and stem-mining by C. americanus larvae
(F4,35 = 5.24, P = 0.002) significantly decreased maximum shoot elongation (Fig. 2). Sap feeding by L. hesperus had no impact on shoot length by either adults (F4,20
= 0.51, P = 0.729) or nymphs (F4,28 = 0.29, P = 0.884).
C. americanus had the highest per-capita effect on
431
XII International Symposium on Biological Control of Weeds
shoot production of L. draba, which is an important
determinant of the rate of spread for clones.
Despite the lack of direct impacts on vegetative
propagules, above-ground feeding may contribute to
the cumulative stress on plants and consequently reduce L. drabas ability to tolerate attack by agents in
other niches. Defoliators are not currently being considered for the biological control of L. draba, but our
data suggest that they can have significant impacts on
plant biomass. Shot-holes produced by P. cruciferae
adults had minimal impact on the performance of bolting plants; we therefore expect that adult feeding by
P. wrasei will not greatly impair bolting plants either.
Adult flea beetle feeding on seedlings and spring rosettes may affect hoary cress population density within
patches and sexual reproduction success.
Above-ground
biomass (g)
17
15
13
11
9
7
0
100
200
300
900
Below-ground
biomass (g)
20
15
10
65
5
0
60
0
100
200
300
55
900
50
P. xylostella egg density
40
Effect of different Plutella xylostella egg densities on mean above- and below-ground plant
biomass (95% CI) of Lepidium draba after 40
days of larval feeding.
35
30
Maximum shoot length (cm)
Figure 1.
45
maximum shoot length (-1.25 0.46 cm/female,
P = 0.01, r2 = 0.113), followed by P. cruciferae (-0.151
0.038 cm/adult, P < 0.001, r2 = 0.234) and P. xylostella
(-0.017 0.004 cm/egg, P < 0.001, r2 = 0.217) (Fig. 2).
None of the herbivore treatments reduced the number
of vegetative shoots produced per plant.
C. americanus laid eggs in more than 90% of the
shoots regardless of female density; however, plant defence responses caused high egg and larval mortality
within the stems. Callus tissues were commonly found
growing around oviposition holes, eggs and larval
mines.
10
20
30
40
C. americanus adult density
65
60
55
50
45
40
35
30
80
P. cruciferae adult density
65
Discussion
60
The principal objective of L. draba biological control
is to reduce the plants vegetative growth capability,
thereby limiting stand dominance and rate of spread of
hoary cress. We found that above-ground plant architecture was significantly affected by chewing insects
in the defoliating (P. xylostella), shot-hole feeding (P.
cruciferae) and stem-mining niches (C. americanus).
Our results suggest that the introduction of endophagous stem-miners may impact plant performance even
at low densities and should be prioritized as candidates
for the biological control of L. draba. However, none
of the insect species investigated decreased vegetative
50
55
45
40
35
30
100
200
300
900
P. xylostella egg density
Figure 2.
432
Effect of different densities of three surrogate
herbivores on maximum shoot length (95% CI)
of Lepidium draba after 40 days of feeding.
The use of surrogate herbivores
Faunistic surveys are increasingly being conducted
on invasive plants in both their native and introduced
ranges (Hinz and Schwarzlnder, 2004), providing information on potential surrogate insects. The biogeographic comparison of herbivore faunas from both
ranges of L. draba by Cripps et al. (2006), for instance, provided the opportunity to explore the impact
of feeding niches potentially important for biological
control of this plant. Another example of the potential
uses of surrogate herbivory comes from the search for
biological control agents for tropical soda apple, Solanum viarum Dunal (Solanaceae), which resulted in the
release of the defoliating leaf beetle, Gratiana boliviana Spaeth (Coleoptera: Chrysomelidae) in the USA
in 2003 (Cuda et al., 2004). Ten major crop pests that
occupy five distinct feeding guilds have been identified
on tropical soda apple in the USA (Cuda et al., 2004),
and additional feeding guilds are known from the insect pests of potato (S. tuberosum; Radcliffe, 1982) and
tomato (S. lycopersicum; Norris and Kogan, 2005).
These pest species could be used as surrogates along
with G. boliviana to investigate potential interactions between herbivores on S. viarum before the introduction of any new biological control agents. Our
proposed protocol does not intend to introduce an extra
step in the agent selection process but could provide a
useful alternative technique to identify and prioritize
candidate biological control agents based on empirical
efficacy data for agents in different feeding niches.
The greatest limitation for the use of surrogate insects to pre-screen candidate biological control agent
efficacy is the availability of specialist feeding niches
within the invaded range. Although we identified several above-ground feeding niches with available surrogates, we did not find any appropriate below-ground
endophagous feeders or shoot gall formers, both niches
for which candidate species have been identified (Fumanal et al., 2004; Cripps et al., 2006). The stem-miner
C. americanus had high rates of oviposition, but the
observed egg and larval mortality may indicate that L.
draba is not an ideal host plant. Observed effects of C.
americanus may underestimate the potential impact of
better adapted specialist stem-miners such as C. merkli
and P. wrasei, currently being studied at CABI Europe,
Switzerland (Cripps et al., 2006). Surrogate herbivores
can be used in numerous invasive plant systems beyond
relatives of agronomic species, but the best known generalist insect herbivores are crop pests. Several weeds
that are unrelated to crops, including L. draba (Cripps
et al., 2006), are recognized as reservoirs of important
economic pests; however, suitable surrogates may not
be available for all target weed species and potential
biological control niches.
Organizing foreign exploration can take a long time
due to logistical, financial, political or safety concerns
and may require establishment of new collaborations
in areas without a tradition of biological weed control.
Furthermore, once candidate insects are found, it takes
additional time before mass-rearing techniques are developed to produce sufficient numbers for host-testing
and pre-release efficacy testing. Testing the efficacy
of surrogate herbivores can provide the opportunity
to investigate protocols for screening agents before
candidates are identified while simultaneously removing ineffective feeding niches from consideration. The
use of surrogate herbivores is therefore not intended to
replace efficacy testing but could be used at an early
stage of a biological control program to assess the impact of specific feeding niches on the target weed in the
invaded range to direct the search for effective candidate agents in the area of origin.
Acknowledgements
We thank M. Cripps and J. McKenney for discussions
on experimental design, K. Schotzko, S. Gersdorf and
M. Cole (all University of Idaho) for technical assistance. We would also like to thank P. Hatcher (University of Reading), M.C. Bon (USDA-ARS) and K.
Marske (University of Auckland) for critical review of
previous drafts of this manuscript. Funding was provided by USDA NRI grant agreement IDA00108-CG
to MS, Idaho State Department of Agriculture through
its cost share program and USDI Bureau of Indian
Affairs.
References
Buchanan, L.L. (1937) A new species of Ceutorhynchus from
North America (Coleoptera: Curculionidae). Bulletin of
the Brooklyn Entomological Society 32, 205207.
Butts, R.A. and Lamb, R.J. (1990) Comparison of oilseed
brassica crops with high or low-levels of glucosinolates
and alfalfa as hosts for 3 species of Lygus (Hemiptera,
Heteroptera, Miridae). Journal of Economic Entomology
83, 22582262.
Cripps, M.G., Hinz, H.L., McKenney, J.L., Harmon, B.L., Merickel, F.W. and Schwarzlaender, M. (2006) Comparative
survey of the phytophagous arthropod faunas associated
with Lepidium draba in Europe and the western United
States, and the potential for biological weed control. Biocontrol Science and Technology 16, 10071030.
Cuda, J.P., Coile, N.C., Gandolfo, D., Medal, J.C. and Mullahey, J.J. (2004). Tropical soda apple. In: Coombs, E.M.,
Clark, J.K., Piper, G.L. and Cofrancesco, A.F.J. (eds) Biological Control of Invasive Plants in the United States. Oregon State University Press, Corvallis, OR, pp. 395401.
Denoth, M., Frid, L. and Myers, J.H. (2002) Multiple agents
in biological control: improving the odds? Biological
Control 24, 2030.
Feeny, P., Paauwe, K.L. and Demong, N.J. (1970) Flea beetles and mustard oils: host plant specificity of Phyllotreta
cruciferae and P. striolata adults (Coleoptera: Chrysomelidae). Annals of the Entomological Society of America
63, 832841.
Fumanal, B., Martin, J.F., Sobhian, R., Blanchet, A. and
Bon, M.C. (2004) Host range of Ceutorhynchus assimilis
433
XII International Symposium on Biological Control of Weeds
(Coleoptera : Curculionidae), a candidate for biological
control of Lepidium draba (Brassicaceae) in the USA.
Biological Control, 30, 598607.
Harmon, B.L. and McCaffrey, J.P. (1997) Laboratory bioassay to assess Brassica spp. germplasm for resistance to
the cabbage seedpod weevil (Coleoptera: Curculionidae).
Journal of Economic Entomology 90, 13921399.
Hartmann, H.T., Kester, D.E. and Davies, F.T. (1990) Plant
Propagation Principles and Practices, 5th ed. PrenticeHall, Englewood Cliffs, NJ.
Hinz, H.L. and Schwarzlaender, M. (2004) Comparing invasive plants from their native and exotic range: What
can we learn for biological control? Weed Technology 18,
15331541.
Hjltn, J. (2004). Simulating herbivory: problems and possibilities. In: Weisser, W.W. and Siemann, E. (eds) Insects
and Ecosystem Function, vol. 173. Springer, New York,
pp. 243255.
Lehtil, K. and Boalt, E. (2004). The use and usefulness of artificial herbivory in plantherbivore studies. In: Weisser,
W.W. and Siemann, E. (eds) Insects and Ecosystem Function, vol. 173. Springer, New York, pp. 257275.
McClay, A.S. and Balciunas, J.K. (2005) The role of pre-
release efficacy assessment in selecting classical biological control agents for weeds applying the Anna Karenina
principle. Biological Control 35, 197207.
McFadyen, R.E. (2003). Does ecology help in the selection
of biocontrol agents? In: Spafford Jacob, H. and Briese,
D.T. (eds) Improving the Selection, Testing and Evaluation of Weed Biological Control Agents, vol. 7. CRC for
Australian Weed Management, Glen Osmond, Australia,
pp. 59.
Mulligan, H.A. and Findlay, J.N. (1974) The biology of Canadian weeds. 3. Cardaria draba, C. chalepensis and C.
pubescens. Canadian Journal of Plant Science 54, 149
160.
Norris, R.F. and Kogan, M. (2005) Ecology of interactions
between weeds and arthropods. Annual Review of Entomology 50, 479503.
Radcliffe, E.B. (1982) Insect pests of potato. Annual Review
of Entomology 27, 173204.
Raghu, S. and Dhileepan, K. (2005) The value of simulating
herbivory in selecting effective weed biological control
agents. Biological Control 34, 265273.
Rice, P.M. (2007). INVADERS Database System. Division of
Biological Sciences, University of Montana. Available at:
http://invader.dbs.umt.edu (accessed February 2007).
Schooler, S., Baron, Z. and Julien, M. (2006) Effect of simulated and actual herbivory on alligator weed, Alternanthera philoxeroides, growth and reproduction. Biological
Control 36, 7479.
Shelton, A.M., Cooley, R.J., Kroening, M.K., Wilsey, W.T.
& Eigenbrode, S.D. (1991) Comparative analysis of two
rearing procedures for diamondback moth (Lepidoptera,
Plutellidae). Journal of Entomological Science 26, 1726.
Sheppard, A.W. (2003). Prioritising agents based on predicted
efficacy: Beyond the lottery approach. In: Spafford Jacob,
H. and Briese, D.T. (eds) Improving the Selection, Testing
and Evaluation of Weed Biological Control Agents, vol.
7. CRC for Australian Weed Management, Glen Osmond,
Australia, pp. 1121.
Talekar, N.S. and Shelton, A.M. (1993) Biology, ecology, and
management of the diamondback moth. Annual Review of
Entomology 38, 275301.
434
The evolutionary history of an
invasive species: alligator weed,
Alternanthera philoxeroides
A.J. Sosa1, E. Greizerstein2, M.V. Cardo1, M.C. Telesnicki1 and M.H. Julien3
Summary
The eco-evolutionary mechanisms of biological invasions are still not thoroughly understood. Alligator weed, Alternanthera philoxeroides (Martius) Gisebach (Amaranthaceae), is a plant native to South
America and a weed in Australia and other countries. To better understand its success as an invader,
we assessed the morphological and cytogenetic variability of 12 Argentine populations and the cytogenetic variability of seven Australian populations. We found differences in leaf shape (width-tolength ratio) and stem architecture in the Argentine populations, in reproduction (sexual with regular
meiosis in two Argentine populations vs completely asexual with irregular meiosis and low pollen
viability in all other populations) and ploidy level (tetraploids with sexual reproduction and seed
production vs hexaploids with or without sexual reproduction). We propose a hypothesis about the
mechanism that drove alligator weed to form highly invasive hybrid populations with vegetative reproduction from diploid ancestors, and we consider the implications for plantherbivore interactions
and biological control of this weed.
Keywords: Alternanthera philoxeroides, biological invasions, plantherbivore
interactions, hybrids, polyploids.
Introduction
The identification and characterization of the native
range and the centre of origin of a weed are crucial in
a biological control program. Alligator weed, Alternanthera philoxeroides (Martius) Gisebach (Amaranthaceae), is a target of biological control in Australia.
Its native range is southern South America (Argentina,
Paraguay, Uruguay and Brazil). In Argentina, the genus Alternanthera Forsskal includes 27 species, four of
them endemic (Pedersen, 1999), indicating that this is
probably its natural area of distribution and perhaps its
centre of origin. Alligator weed is represented by two
known morphological forms in Argentina, A. philoxeroides f. philoxeroides (Mart.) Griseb. and A. philoxeroides f. angustifolia Sssenguth, and a third, intermediate
USDA-ARS South American Biological Control Laboratory, Bolivar
1559 (B1686EFA), Hurlingham, Buenos Aires, Argentina.
2
Universidad de Buenos Aires, Facultad de Ciencias Exactas y Naturales, Gentica y Evolucin, Pabelln II, Departamento de Ecologa,
(1428) Ciudad Autnoma de Buenos Aires, Argentina.
3
CSIRO Entomology European Laboratory Campus International de
Baillarguet. 34980 Montferrier sur Lez, France.
Corresponding author: A.J. Sosa <[email protected]>.
CAB International 2008
1
form, all of which were recently associated to a complex of hybrids (Sosa et al., 2004).
A. philoxeroides reproduces both sexually and asexually. However, production of viable seeds seems to be
restricted to its native range. Understanding why some
populations of alligator weed are fertile and others sterile may be important in understanding why this plant is
invasive and in the development of management strategies. Factors, both intrinsic, e.g. gametogenesis, and
extrinsic, e.g. pollination processes, affect the ability
of plants to produce seeds. Seeds and seedlings of both
known forms of alligator weed were recorded in the
field and in germination trials in the lab (Sosa et al.,
2004). Nevertheless, the requirements for successful
sexual reproduction, the characterization of hybrids
and their role in the invasiveness of the species remain
uncertain.
All alligator weed in Argentina propagates vegetatively, and only in particular situations does it also
propagate by seed. The reliance on vegetative reproduction could indicate the presence of hybrids in the native range, as in other Amaranthaceae in South America
(Greizerstein and Poggio, 1992). Hybrids can be fertile
or sterile, depending on the differences between the
parental genomes. They can also develop an enhanced
435
XII International Symposium on Biological Control of Weeds
vegetative reproduction capacity. The strongest evidence for the presence of hybrids can be observed in
cytogenetic studies, such as the characterization of meiosis (Mallet, 2005).
For biological control of alligator weed, the flea
beetle, Agasicles hygrophila Selman and Vogt (Chrysomelidae: Alticinae), the stem borer, Arcola malloi
(Pastrana) (Lepdidoptera: Pyralidae), and the thrips,
Amynotrips andersoni ONeil (Thysanoptera: Phaelothripidae), were released in the USA to control
the weed. The flea beetle and the moth were subsequently introduced in Australia (Julien, 1981; Julien
and Griffiths, 1999). In spite of their success, there are
limitations, mainly concerning terrestrial growth of the
weed and the lack of biological control in cooler areas
of Australia and the USA (Cofrancesco, 1988; Julien
and Bourne, 1988; Julien et al., 1995; Julien and Stanley, 1999).
There are a number of reasons why the flea beetle,
currently the most important of the biological control
agents, may be restricted in space and habitat. Its life
cycle is strongly associated with aquatic conditions,
including high humidity. For successful pupation, it
requires hollow stems with thin walls (normally found
in aquatic growth) rather than thick-walled stems
with small cavities (normally terrestrial growth; Vogt,
1973). The flea beetle is also sensitive to low temperatures (Stewart et al., 1996). However, no possible relationship between its performance and plant genetic
differences within alligator weed has been considered
thus far.
Genetic studies were initiated to understand the
morphological variation found in A. philoxeroides in
its native range and to characterize alligator weed populations in Australia. In addition, a preliminary study
was conducted for demographic parameters of the flea
beetle under laboratory conditions to evaluate the role
of plant ploidy level in the insects performance.
and Tukeys honestly significant difference (HSD) test
was used to separate the means.
Cytogenetic studies
The cytogenetic variation of alligator weed was
studied to identify hybrid forms in both the native and
adventive ranges and to characterize their reproductive status (sexual or asexual). Young flowers were
collected into vials with a 6:3:1 solution of 96% ethyl
alcohol, chloroform and acetic acid, from 12 sites in
Argentina (Fig. 1) and nine in Australia (Fig. 2). The
process of meiosis in flower-bud cells was studied. In
addition, mitosis was studied in at least 30 cells taken
from fine root-tips. Viability of pollen from anthers was
estimated using Alexander stain, a stain for chromatin.
Unstained pollen grains indicate absence of chromatin
and non-viability.
A. hygrophila performance in relation to
ploidy levels
Adults of the flea beetle were collected from the
field at Hurlingham, near Buenos Aires, and raised in a
chamber at 25C and 12-h light. Ten first-instar larvae
were placed in a plastic container (8 cm diameter, 5 cm
height) with moistened tissue paper and fed on plants
from different alligator weed populations: Santa F
(hexaploid), Predelta (hexaploid), Hurlingham (hexaploid), Cauelas (hexaploid) and Tandil (tetraploid).
Each treatment was replicated ten times. To estimate
performance of the flea beetle on plants from the different localities, survivorship of larvae and immature
developmental time were measured. The results were
analysed with multivariate ANOVA (MANOVA), and
Tukeys HSD test was used to compare means.
Results
Morphological studies
Methods and materials
Morphological studies
Cultures from seedlings and stems from the following different localities in Argentina were cultivated in
the greenhouse for 6 months under controlled conditions: 1 Rt. 11, 22 km SW Reconquista, Santa F
Province, 291651.3S, 594912.8W (from now
on called Santa F); 2 Rt. 380, to Lules, Tucumn
Province, 265227.5S, 651825.1W (Tucumn); 3
Hurlingham, Buenos Aires Province 343514.2S,
583824.0W (Hurlingham) and 4 Rt.30, 23 km
from Tandil, Buenos Aires Province, 371135.5S,
590329.7W (Tandil). Twenty-one morphological parameters were measured and analysed using a principal
component analysis (PC ord 4). The variables that
explained the greater portion of the variance were analysed using a one-way analysis of variance (ANOVA),
The first three axes of the principal component
analysis explained 75% of the variance. Populations
were ordered according to the variables associated to
these axes: shoot diameter, width-to-length ratio of
the first leaf and apical angle of the first leaf (Fig. 3).
The populations that were more representative of each
form (Table 1) were analysed in terms of these three
variables. The Cauelas population closely resembled
Tandil population in shoot diameter, but leaves were
larger and the width-to-length ratio was intermediate
between Hurlingham and Santa F. The Predelta population, on the contrary, closely resembled Hurlingham
population, although with bigger leaves and shoots.
A. philoxeroides f. angustifolia differs from A.
philoxeroides f. philoxeroides in the diameter of the
stems and in the shape of the leaves. In the laboratory,
A. philoxeroides f. philoxeroides from Tandil (APP
in Table 1) had significantly thinner stems (smaller
436
The evolutionary history of an invasive species: alligator weed, Alternanthera philoxeroides
Figure 1.
Figure 2.
The locations of the 12 Argentine populations of Alternanthera philoxeroides that were
studied. Numbers refer to populations described in Table 2.
The locations of the nine Australian populations of Alternanthera philoxeroides that were
studied. Letters refer to populations described in Table 2.
437
XII International Symposium on Biological Control of Weeds
Figure 3.
Ordination obtained with principal component analysis of four populations of Alternanthera philoxeroides;
Santa F (SF), Hurlingham (H), Tandil (Tandil) and Tucumn (TUCU), using 21 morphological variables.
internode diameter), ovate leaf tips (greater leaf apical
angle) and shorter leaves (higher values in leaf widthto-length ratio). The angustifolia form from Hurlingham (APA in Table 1) significantly differed in its more
acute leaf apical angle, longer leaves (lower leaf widthto-length ratio) and broader stems (greater internode
diameter). Plants from Santa F (intermediate form in
Table 1) resembled the angustifolia form due to their
broad stems, whereas they had similar leaf width-tolength ratio and leaf apical angle to the philoxeroides
form (Table 1).
Cytogenetic studies
Alligator weed has small and numerous chromosomes that are difficult to distinguish and count. The
results revealed that alligator weed populations in Argentina are composed of a complex of hybrids. Chromosome number differed among populations within
the native range (Table 2). Populations from Tandil had
66 chromosomes compared to other populations with
higher numbers (approximately 100). In comparison,
samples from Santa F, Cauelas, Tucumn and Australia (Table 2) had aberrant meiosis in which several
univalent chromosomes were not aligned in the equatoTable 1.
rial plate. The resulting cells had different chromosome
numbers, approximately 100. Material from Australian
populations could not be analysed for chromosome
number because root-tips were in an advanced state of
cell division at the time of the study.
It is interesting to note that Alternanthera aquatica
Chod., a species closely related to alligator weed, also
had 66 chromosomes. This is the first report of a chromosome number for this plant species.
Pollen staining confirmed low viability of pollen
grains, suggesting hybrid forms of A. philoxeroides.
Pollen viability from Tandil and Predelta was higher
than in Santa F, Cauelas, Hurlingham or Australia
(Table 2).
A. hygrophila performance in relation
to ploidy levels
Preliminary results show that host genetic variability affects the performance of A. hygrophila (MANOVA: Wilks lambda = 0.63, P = 0.008). Differences
were found in immature survivorship (Fig. 4, ANOVA:
F4, 44 = 4.07, P = 0.006), which was higher in plants
from Hurlingham (0.77 0.03) than in those from
either Santa F (0.44 0.07) or Predelta (0.49 0.08),
Morphometric data for laboratory-grown Alternanthera philoxeroides from three locations in Argentina. APP,
A. philoxeroides f. philoxeroides; APA, A. philoxeroides f. angustifolia; AP?, an intermediate form. Means
and standard errors are shown. Means within a column followed by different letters are significantly different
(P < 0.05; ANOVA, Tukey post hoc multiple comparisons).
Form
Collecting site
APP
AP?
APA
Tandil, n = 10
Santa F, n = 10
Hurlingham, n = 10
Internode diameter (mm)
1.95 1.11a
2.84 0.77 b
3.16 0.48b
Leaf length to width ratio
0.40 0.07a
0.42 0.12a
0.24 0.05b
438
Leaf apical angle ()
137.33 43.16a
105.56 31.00a
68.83 14.45b
The evolutionary history of an invasive species: alligator weed, Alternanthera philoxeroides
Table 2.
esults of cytogenetic studies on 12 Argentine populations of Alternanthera A. philoxeroides. The dash () indiR
cates that data is not yet available. The numbers in the first column relate to locations in Figs. 1 and 2.
Population
Number of
individuals
Argentina populations
Reconquista,
5
Santa F (1)
Tandil, Buenos
6
Aires (2)
Tandil, Buenos
4
Aires (3)
La Paz, Entre
5
Ros (4)
Cauelas,
5
Buenos
Aires (5)
Predelta, Entre
4
Ros (6)
Cazn, Buenos
4
Aires (7)
Hurlingham (8)
5
Chaco (9)
5
Paranacito (10)
5
La Plata (11)
5
Tucumn (12)
5
Australian populations
Hunter Valley,
3
NSW (A)
Dandenong,
2
Victoria (B)
Kaotara,
4
NSW (C)
Maitland East,
3
NSW (D)
Wallsend,
4
NSW (E)
Oakville,
5
NSW (F)
Richmond,
3
NSW (G)
Form
Chromosome
number 2n
Pollen
stainability (%)
Intermediatea
50
philoxeroides
Approximately
100
66
philoxeroides
66
95
angustifolia
95
61
angustifolia
ca. 100
angustifolia
ca. 100
94
*intermediate
65
angustifolia
A. aquatica?
66
0
15
A. aquatica?
*intermediate
angustifolia
66
100
angustifolia
angustifolia
angustifolia
angustifolia
angustifolia
angustifolia
angustifolia
Observations on
pollen grains
Different size
(aneuploid)
Normal morphology
and size
Normal morphology
and size
Different size
(aneuploid)
Normal morphology
and size
Different size
(aneuploid)
Different size
(aneuploid)
Fruits
No
Yes
Yes
No
Yes
No
No
Yes?
No
No
Yes
No
Different size
(aneuploid)
Different size
(aneuploid)
Different size
(aneuploid)
Different size
(aneuploidy)
Intermediate: A form that appears to be intermediate between the two forms angustifolia and philoxeroides (see Table 1).
but showed no difference from plants from Tandil (0.57
0.06) or Cauelas (0.57 0.06).
However, developmental time (in days) of different
hybrids were not different (Santa F 22.2 0.8; Predelta 21.1 1.1; Hurlingham 21.2 0.6; Tandil 23.1
0.9; Cauelas 20.4 0.5; ANOVA: F4, 44 = 1.85,
P = 0.136).
Discussion
Irregularities in meiosis division are associated with
hybrid organisms (Mallet, 2005). Irregularities observed through cytogenetic analysis, and subsequent
correlation with pollen staining strongly suggests that
the entity A. philoxeroides is a complex of hybrids. Additional chromosomes (through hybridization) could be
beneficial for these plants (Levin, 2002), particularly as
they do not depend on sexual reproduction. New polyploids may possess novel physiological, ecological or
phenological characteristics that allow them to colonize new niches, and they may be wholly or partially
reproductively isolated from their diploid progenitors
(Ramsey and Schemske, 1998).
Based on our results and knowing that in the Gomphreninae tribe the basic chromosome number is x =
1617 (Okada et al., 1985), we propose a hypothetical
model showing the evolution of A. philoxeroides in its
native area (Fig. 5). Diploid ancestors gave origin to
439
XII International Symposium on Biological Control of Weeds
Figure 4.
Figure 5.
Survivorship of larvae of Alternanthera hygrophila on five populations
of A. philoxeroides from Argentina; Santa F (SF), Predelta (PRE),
Hurlingham (HU), Tandil (TA) and Cauelas (CA). Vertical bars denote
0.95 confidence intervals. Different letters indicate significant differences (ANOVA P < 0.05).
Hypothetical hybridization model of Alternanthera philoxeroides in Argentina.
440
The evolutionary history of an invasive species: alligator weed, Alternanthera philoxeroides
auto- and allotetraploids, the latter being represented
by the Tandil population. Due to a lack of reduction in
meiosis I, these subsequently produced two different
types of hexaploids, one with aberrant meiosis (Santa
FCauelasHurlingham) and another with normal
meiosis and sexual reproduction (Predelta). However,
it is still uncertain which parents gave rise to the alligator weed hybrids. To assess this issue, the genomic in
situ hybridization technique will be used in the future to
determine the origins of the different chromosomes in
each population. This technique is used with the combination of molecular tools to study the evolutionary
aspects and the role of polyploids in other biological
invasions (Fehrer et al., 2007). For the alligator weed,
they should provide a more complete picture of its evolutionary history.
As the only cytotype of A. philoxeroides found in
Australia is a hexaploid, we suggest that its invasive
capability could be attributed to traits from one of the
native range hexaploid (hybrid) populations and that
these traits were gained before invasion. As previously
documented for other weeds (Lee, 2002), alligator weed
shows positive effects of hybridization on invasibility,
such as faster growth, greater size and increased aggression (Okada et al., 1985; Alonso and Okada, 1996).
In addition, alligator weed plants growing in aquatic
and terrestrial environments are morphologically different, and this might be due to genotype environment interactions.
The alligator weed flea beetle that was released in
the USA and Australia was collected in Buenos Aires
Province in Argentina and in Montevideo, Uruguay
(Coulson, 1977). The insects used in the present study
were collected in Hurlingham. As it is highly probable
that these populations have the same origin, the use of
insects from Hurlingham for plantherbivore interaction
studies may validate comparisons between the outcome
of our experiments and the behaviour of A. hygrophila
in the adventive range. In our experiments, higher survivorship on Hurlingham plants could be explained by
differences in the genotypes or by plant maternal effects.
This warrants further study, and a current experiment is
testing for differences in fertility and fecundity.
The presence of different cytotypes of alligator
weed should be considered when explaining the lack
of success of the flea beetle in controlling the weed in
cool areas, particularly in the USA, where the existence
of distinct alligator weed biotypes has been confirmed
(Kay and Haller, 1982).
Acknowledgements
Many thanks to Lidia Poggio for letting us use the
facilities at the Laboratorio de Citogentica. We also
appreciate comments and suggestions on the original
manuscript by reviewers. We thank the Australian Government for supporting this project through the Defeating the Weed Menace program.
References
Alonso, S.I. and Okada, K.A. (1996) Capacidad de propagacin de Alternanthera philoxeroides en suelos agrcolas.
Ecologa Austral 6, 916.
Cofrancesco, A.F. Jr. (1988) Alligator weed survey of ten
southern states. Miscellaneous Paper A-8883. US Army
Corps of Engineers Waterways Experimental Station,
Vicksburg, MS, 69 pp.
Coulson, J.R. (1977) Biological control of alligatorweed,
19591972. A review and evaluation. Technical Bulletin
No. 1547. Agricultural Research Service, US Department
of Agriculture, Washington, DC, 98 pp.
Fehrer, J., Krahulcov, A., Krahulec, F., Chrtek, J. Jr., Rosenbaumov, R. and Brutigam. S. (2007) Evolutionary aspects in Hieracium subgenus Pilosella. In: Grossniklaus,
U., Hrandl, E., Sharbel, T. and van Dijk, P. (eds) Apomixis:
Evolution, Mechanisms and Perspectives. Regnum Vegetabile 147, Koeltz, Knigstein, Germany, pp. 359390.
Greizerstein, E.J. and Poggio, L. (1992) Estudios citogenticos de seis hbridos interespecficos de Amaranthus
(Amaranthaceae). Darwiniana 31, 159165.
Julien, M.H. (1981) Control of aquatic Alternanthera philoxeroides in Australia: another success for Agasicles hygrophila. In: DelFosse, E.S. (ed.) Proceedings of the 5th
International Symposium on Biological Control of Weeds.
CSIRO, Melbourne, Australia, pp. 583588.
Julien, M.H. and Bourne, A.S. (1988) Alligator weed is spreading in Australia. Plant Protection Quarterly 3, 9196.
Julien, M.H. and Griffiths, M.W. (1999) Biological Control
of Weeds. A World Catalogue of Agents and Their Target Weeds, 4th ed. CAB International, Wallingford, UK,
223 pp.
Julien, M.H., Skarratt, B. and Maywald, G.F. (1995) Potential
geographical distribution of alligator weed and its biological control by Agasicles hygrophila. Journal of Aquatic
Plant Management 33, 5560.
Julien, M.H. and Stanley, J.N. (1999) The management of
alligator weed, a challenge for the new millennium. In:
Ensbey, R., Blackmore, P. and Simpson, A. (eds) 10th
Biennial Noxious Weeds Conference. NSW Agriculture,
Australia, pp. 213.
Kay, S.H. and Haller, W.T. (1982) Evidence for the existence
of distinct alligator weed biotypes. Journal of Aquatic
Plant Management 20, 3741.
Lee, C.E. (2002) Evolutionary genetics of invasive species.
Trends in Ecology and Evolution 17, 386391.
Levin, D.A. (2002) The Role of Chromosomal Change in
Plant Evolution. Oxford University Press, Oxford, UK.
Mallet, J. (2005) Hybridization as an invasion of the genome.
Trends in Ecology and Evolution 20, 229237.
Okada, K.A., Alonso, S.I. and Rodriguez, R.H. (1985) Un
citotipo hexaploide de Alternanthera philoxeroides como
nueva maleza en el partido de Balcarce, provincia de
Buenos Aires. Revista de Investigaciones Agropecuarias
INTA 20, 3753.
Pedersen, T.M. (1999) Amaranthaceae. In: Zuloaga, F.O. and
Morrone, O. (eds) Catlogo de las Plantas Vasculares de
la Argentina II. Monographs in Systematic Botany from
the Missouri Botanical Garden, Missouri Botanical Garden, St. Louis, MO, pp. 1231.
Ramsey, J. and Schemske, D.W. (1998) Pathways, mechanisms and rates of polyploidy formation in flowering
441
XII International Symposium on Biological Control of Weeds
plants. Annual Review of Ecology and Systematics 29,
467501.
Sosa, A.J., Julien, M.H. and Cordo, H.A. (2004) New research on alligator weed (Alternanthera philoxeroides) in
its South American native range. In: Cullen, J.M., Briese,
D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott,
J.K. (eds) Proceedings of the XI International Symposium
on Biological Control of Weeds. CSIRO Entomology,
Canberra, Australia, pp. 180185.
Stewart, C.A., Emberson, R.M. and Syrett, P. (1996) Temperature effects on the alligator weed flea-beetle, Agasicles
hygrophila (Coleoptera: Chrysomelidae): implications for
biological control in New Zealand. In: Moran, V.C. and
Hoffmann, J.H. (eds) Proceedings of the IX International
Symposium on Biological Control of Weeds. Stellenbosch,
South Africa, pp. 393398.
Vogt, G.B. (1973) Exploration for natural enemies of alligator
weed and related plants in South America, Appendix B.
In: Gangstad, E.O., Scott, R.A. and Cason, R.G (eds) Biological Control of Alligatorweed. Technical Report 3. US
Army Engineer Waterways Experiment Station, Aquatic
Plant Control Program, Vicksburg, MS, pp. 166.
442
Landscape genetics and climatic
associations of flea beetle lineages
and implications for biological control
of tansy ragwort
M. Szcs1, C.L. Anderson2 and M. Schwarzlnder3
Summary
While Longitarsus jacobaeae Waterhouse (Coleoptera: Chrysomelidae) has shown the best results for
biocontrol of tansy ragwort, Senecio jacobaea L. (Asteraceae), the precise effects of this flea beetle
on tansy ragwort distribution and abundance are confused by the introduction in North America of
two distinct strains, of Italian and Swiss origin. Beetles of the two biotypes differ in their phenology,
while hybrids of the two strains show life-cycle characteristics different from either parental strain.
However, it is not known which biotype(s) currently provides control of tansy ragwort infestations
or whether both biotypes or hybrids of the two have been established in the USA. Moreover, mixed
populations may exhibit lower efficiency in controlling the target weed relative to pure strains, and
specificity for the invasive host may be compromised. In this study, molecular markers were employed to distinguish between biotypes of L. jacobaeae. Analysis of mitochondrial sequence of specimens from Switzerland, Oregon and California revealed higher than expected sequence variation,
within and between strains. Despite the high levels of polymorphism, cladistic analysis did not show
distinct separation of strains into well-defined clades. These results indicate that Swiss L. jacobaeae
beetles may have established in North America in contrast with the general assumptions. As the revealed mitochondrial DNA (mtDNA) polymorphisms themselves will not identify populations, we
are now employing nuclear markers to characterize North American populations, their ancestry and
their association with certain climatic regions. The goal is to match the most effective genotypes of
L. jacobaeae with new tansy ragwort infestations east of the Cascade mountain range.
Keywords: COI, sequence variation, hybrid strain, insect biotypes.
Introduction
Tansy ragwort is one of the fastest spreading invasive
plants in the western USA since its introduction in 1922
(McEvoy, 1984). It is particularly prevalent in Washington, California and Oregon, has spread recently into
Montana and Idaho and is listed as a noxious weed in
eight states. West of the Cascade Mountain range, tansy
University of Idaho, Department of Plant, Soil, and Entomological Sciences, Moscow, ID 83844-2339, USA.
2
University of Idaho, Department of Fish and Wildlife Resources, Moscow, ID 83844-1136, USA.
3
University of Idaho, Department of Plant, Soil, and Entomological Sciences, Moscow, ID 83844-2339, USA.
Corresponding author: M. Szcs <[email protected].
edu>.
CAB International 2008
1
ragwort infestations have been most effectively controlled with biological control agents (McEvoy et al.,
1991). Tansy ragwort biomass was reduced by 93%
in western Oregon (Coombs et al., 1996) and by 99%
in sites in northern California (Hawkes and Johnson,
1978) after the introduction of three biological control
agents, but the success is primarily attributed to Longitarsus jacobaeae Waterhouse (McEvoy et al., 1991).
Although L. jacobaeae is the most effective biocontrol
agent, quantifying the effects of this flea beetle on tansy
ragwort distribution and abundance is difficult due to
the introduction of two distinct biotypes, or strains, of
Italian and Swiss origin in North America (Frick, 1971;
Frick and Johnson, 1973).
Beetles of the two biotypes differ in their phenology
and environmental requirements, and hybrids of the two
biotypes show phenologies different from that of either
443
XII International Symposium on Biological Control of Weeds
parental strain (Frick, 1971; Frick and Johnson, 1973).
While the establishment of the Italian biotype of L. jacobaeae is widely assumed in western coastal areas,
the fate of the originally released Swiss beetles in Del
Norte County, California (Frick, 1970) is not known.
Thus, it is not clear which biotype(s) are currently providing effective control of coastal infestations of tansy
ragwort or whether hybrids of these biotypes have been
established in the USA. Beetles derived from these
North American coastal populations have failed to establish in areas east of the Cascade Mountain range,
which are characterized by a continental climate with
cold winters (Turner and McEvoy, 1995). Mixed populations might be less effective in controlling tansy ragwort infestations, and hybrids may differ in their host
specificity, placing native confamilial plants at risk
(Hoffman et al., 2002). Consequently, it is a matter of
great importance to determine the origins and composition of extant populations in North America.
Comparison of the genetic composition of extant
L. jacobaeae populations in North America with European source populations will not only indicate the
origins of North American populations but also reveal
patterns of hybridization and climatic associations of
L. jacobaeae lineages. In this study, we investigated
two introduced North American populations and one
Swiss population of L. jacobaeae to assess whether
mtDNA sequences can reveal sufficient genetic variation and differentiation between strains that may allow
development of molecular markers that can distinguish
biotypes of the flea beetle.
Methods and materials
DNA extraction
DNA was extracted from 50 individuals using a
DNeasy Tissue Kit (Qiagen Inc., Valencia, CA), following the manufacturers protocol. Twenty-two individuals were collected from Mettembert, Switzerland
(4724N, 720E), 14 specimens were collected both
from Salem, OR (4493N, 12299W) and from Crescent City, CA (4144N, 12408W).
DNA amplification
Polymerase chain reaction (PCR) was performed
with general-purpose insect-derived primers that are
known to amplify a fragment of the coleopteran mtDNA
genome (Szalanski and Owens, 2003):
C1-J-2797: 5-CCTCGACGTTATTCAGATTACC-3
C2-N-3400: 5-TCAATATCATTGATGACCAAT-3
These primers amplify the 3 end of the cytochrome
oxidase I gene, the transfer RNA for leucine, and the
5 end of cytochrome oxidase II gene. Each reaction
was carried out in 20 ml reaction volume, containing 1
Colorless GoTaq buffer pH 8.5, 1.5 mM MgCl2, 1 U of
GoTaq DNA polymerase (Promega), 10 pmol of each
primer, 0.2 mM deoxyribonucleotide triphosphate and
approximately 25 ng DNA. Thermal cycling conditions
were 3 min at 94C, followed by 40 cycles of 30 s at
94C, 20 s at 51C, 50 s at 72C, with a final step of
3 min at 72C.
DNA sequencing and analysis
PCR products were prepared for sequencing using
ExoSAPit (GE Healthcare Corp., Piscataway, NJ), following the manufacturers protocol. Cleaned samples
were sequenced with Big Dye version 3.1 following
the manufacturers protocol and the reactions run on
an Applied Biosystems 3130xl automated sequencer.
Sequences were edited and aligned with Sequencher
4.1 (Gene Codes Corp., Ann Arbor, MI). Parsimony
analysis was carried out using PAUP 4.0b10 for Macintosh (Swofford, 2002). For the tree shown in Fig. 1,
Apthona cyparissiae Koch was used as an outgroup to
root the tree (accession number gi88810025), incorporating 541 bp of sequence data. Thirty-two sites were
parsimony-informative, including 50 haplotypes from
North American and Swiss populations. Search was
full heuristic, with branch swapping [tree bisection-reconnection (TBR)] and gaps treated as missing data.
To gauge the reliability of the observed topology, we
did a further round of bootstrap analysis using a different outgroup, Diabrotica barberi Smith and Lawrence
(accession number gi11275649), and a reduced number
of taxa. For the bootstrap analysis, we used 27 haplotype samples representing all the major branches of the
tree shown in Fig. 1, incorporating 588 bp of sequence
data, of which 13 sites were parsimony-informative.
Search was full heuristic with branch swapping, gaps
were treated as missing data and 100 replicates were
performed.
Results
The L. jacobaeae sequences revealed an unexpectedly
high level of polymorphism. Ninety-eight polymorphic
sites were detected within the analysed 541 bp region, of
which 32 were parsimony-informative. The sequences
were obtained from 50 individuals from three populations (two North American and one Swiss). By way of
comparison, sequencing the same region from 22 individuals of the southern corn rootworm, Diabrotica
undecimpunctata, revealed two haplotypes, differentiated by a single nucleotide polymorphism (Szalanski and Owens, 2003). High levels of polymorphism,
notwithstanding, parsimony analysis of the unique
haplotypes did not show clustering of strains into welldefined clades. Results of this analysis are shown in
Fig. 1. To test the reliability of the topology we obtained
in our parsimony analysis, we carried out bootstrap
analysis (Hillis and Bull, 1993). Bootstrap analysis
indicated a 50% majority-rule consensus tree in which
444
Landscape genetics and climatic associations of flea beetle lineages
Figure 1.
Rooted parsimony tree constructed in PAUP using heuristic search with branch swapping (TBR) and gaps treated as missing data. Branch lengths as shown. Longitarsus
jacobaeae originating from Switzerland (SW122), Oregon (OR114) and California
(CA114).
all North American and Swiss samples cluster in a simple polytomy (Fig. 2).
Discussion
Our results show that there is substantial sequence
variation in the mitochondrial genome of L. jacobaeae,
both within and between strains, suggesting that further
investigation of other regions will yield genetic markers indicative of strain type. Both Italian- and Swissstrain beetles were originally released in California, but
only the establishment of the Italian biotype was confirmed, and it is now widely accepted that all beetles
distributed along the west coast are of Italian origin
(Turner and McEvoy, 1995). According to this assumption, the CA and OR populations sampled for this study
445
XII International Symposium on Biological Control of Weeds
effective genotypes of L. jacobaeae with new tansy ragwort infestations east of the Cascade Mountain range.
Acknowledgements
We thank Urs Schaffner (CABI Bioscience, Switzerland), Eric Coombs (Oregon Department of Agriculture) and Baldo Villegas (California Department of
Food and Agriculture) for providing us with specimens.
This project is funded by the Palouse Cooperative
Weed Management Area (CWMA), the USDA Forest
Service Clearwater National Forest and the Potlatch
Corporation.
References
Figure 2.
Bootstrap analysis conducted with full heuristic search with branch swapping, gaps treated
as missing data and 100 replicates performed.
Branch lengths as shown. Longitarsus jacobaeae
originating from Switzerland (SW122), Oregon (OR114) and California (CA114).
were derived from the Italian strain. Our results are
surprising in light of this general assumption. Because
of the strikingly high level of observed polymorphism,
a trend would be expected for the clustering of haplotypes if the sampled beetles represent two distinct
biotypes, such as the Swiss and Italian. One possible
explanation for the lack of segregation of strains may
be that Swiss beetles from the original releases have
established in California, and either significant hybridization or introgression occurred between the Italian
and Swiss beetles. This seems a plausible scenario, as
L. jacobaeae used for this analysis was collected in Del
Norte County in California where Swiss beetles were
initially released and Californian populations later provided the source populations for the redistribution of
flea beetles along the west coast.
However, definitive conclusions cannot be drawn
from this sampling because it lacks representative beetles of known Italian origin. Therefore, we are making
efforts to obtain beetles from Italy. In addition, we will
be employing nuclear markers to assess the ancestry of
North American populations and their association with
certain climatic regions. The goal is to match the most
Coombs, E., Radtke, H., Isaacson, D. L. and Snyder, S. P.
(1996) Economic and regional benefits from the biological control of tansy ragwort Senecio jacobaea in Oregon,
USA. In: Moran, V. C. and Hoffmann, J. H. (eds) Proceedings of the IX International Symposium on Biological
Control of Weeds. University of Cape Town, Stellenbosch,
South Africa, pp. 489494.
Frick, K. E. (1970) Ragwort flea beetle established for biological control of tansy ragwort in Northern California.
California Agriculture April 24, 1213.
Frick, K. E. (1971) Longitarsus jacobaeae (Coleoptera:
Chrysomelidae), a flea beetle for the biological control of
tansy ragwort. II. Life history of a Swiss biotype. Annals
of the Entomological Society of America 64, 834840.
Frick, K. E. and Johnson, G. R. (1973) Longitarsus jacobaeae
(Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy ragwort. 4. Life history and adult aestivation of an Italian biotype. Annals of the Entomological
Society of America 66, 358366.
Hawkes, R. B. and Johnson, G. R. (1978) Longitarsus jacobaeae aids moth in the biological control of tansy ragwort.
In: Freeman, T.E. (ed.) Proceedings of the IV International
Symposium on Biological Control of Weeds.. University
of Florida, Gainesville, FL, pp. 193196.
Hillis, D. M. and Bull, J. J. (1993). An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology 42, 182192.
Hoffmann, J. H., Impson, F. A. C. and Volchansky, C. R.
(2002) Biological control of cactus weeds: implications
of hybridization between control agent biotypes. Journal
of Applied Ecology 39, 900908.
McEvoy, P. (1984) Depression in ragwort (Senecio jacobaea)
abundance following introduction of Tyria jacobaeae and
Longitarsus jacobaeae on the central coast of Oregon. In:
Delfosse, E.S. (ed.) Proceedings of the VI International
Symposium on Biological Control of Weeds. Agriculture
Canada, Ottawa, Canada, pp. 5764.
McEvoy, P., Cox, C. and Coombs, E. (1991) Successful
biological control of ragwort, Senecio jacobaeae, by introduced insects in Oregon. Ecological Applications 1,
430442.
Swofford, D. L. (2002) PAUP*. Phylogenetic Analysis Using
Parsimony (*and Other Methods), Version 4. Sinauer Associates, Sunderland, MA.
446
Landscape genetics and climatic associations of flea beetle lineages
Szalanski, A.L. and Owens, C.B. (2003) Genetic variation
of the southern corn rootworm (Coleoptera: Chrysomelidae). Florida Entomologist 86, 329333.
Turner, C. E. and McEvoy, P. B. (1995) Tansy Ragwort Senecio jacobaea (Asteraceae). In: Nechols, J.R., Andres, L.A.,
447
Beardsley, J.W., Goeden, R.D. and Jackson, C.J. (eds) Biological Control in the U. S. Western Region: Accomplishments and Benefits of Regional Research Project W-84
(19641989). University of California, Division of Agriculture and Natural Resources, Berkeley, CA, pp. 264269.
XII International Symposium on Biological Control of Weeds
Genetic characterization of the whitetop collar
gall weevil, Ceutorhynchus assimilis, enhances its
potential as biological control agent
M.C. Bon,1 B. Fumanal,1 J.F. Martin2 and J. Gaskin3
1
USDAARSEBCL, 34980 Montferrier le Lez, France
2
CBGP, 34980 Montferrier le Lez, France
3
USDAARS, NPRL, Sidney, MT, USA
In the field of biological control, it is becoming clear that genotypes of invasive weeds vary in their
susceptibility to natural enemies and that genotypes of actual or candidate biological control agents
vary in their ability to control different genotypes of target weeds, influencing the level of total control
achieved. Few studies attempt to both pinpoint the area of provenance of the invasive weed in the introduced range and search for matching genetic relationships between the target weed and the herbivore
before selection of biological control agent. The case of hoary cress, Lepidium draba (Brassicaceae,
native to Eurasia), which is one of the most invasive noxious weeds in North American rangelands and
croplands, and one of its natural enemies, a collar gall weevil, Ceutorhynchus assimilis (Coleoptera:
Curculionidae), explicitly illustrates the benefit of using evolutionary knowledge to refine efficacy and
safety of biological control. Within the geographic distribution of the phytophagous weevil, genetic
analysis uncovered several morphocryptic genetic lineages including one race specific to L. draba with
regard to the larval development and restricted to southern France and northern Italy. Crossing experiments were carried out to assess the level of reproductive isolation of these lineages. Concomitantly, a
phylogeography study of the weed in its native range gave evidence of a cluster of haplotypes originating from the same region as the L. draba specific race found in the weevil.
Pinpointing the origin of North American invasive
Vincetoxicum spp. using phylogeographical markers
M.C. Bon,1 R. Sforza,1 W. Jones,1 C. Hurard,1 L.R. Milbrath2 and S. Darbyshire3
2
1
USDAARSEBCL, 34980 Montferrier sur Lez, France
USDAARSPlant, Soil and Nutrition Lab, Ithaca, NY 14853, USA
3
Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada
Three European species of swallow-worts belonging to the Apocynaceae family are established in
North America. Vincetoxicum nigrum (L.) Moench (black swallow-wort) and Vincetoxicum rossicum
(Kleo.) Barb. (pale swallow-wort) are both highly invasive in natural areas, abandoned pastures and
rural sites. Vincetoxicum hirundinaria Medik. (white swallow-wort) occurs sparsely in the Northeast
as a horticultural escape. As current control measures for the swallow-worts are unable to alleviate
their weedy impact and because of the numerous natural enemies associated with Vincetoxicum sp. in
Europe, classical biological control of swallow-worts in North America is being considered. Ascertaining the insect fauna of Vincetoxicum species in Eastern Europe and western Russia is confounded by
problems in target plant taxonomy at both species and genus levels. Tracing the origins of invasive
weeds and knowing the levels of genetic variation relative to the native range seems to be increasingly
important for conducting rigorous specificity tests in the time frame of a biological control programme.
Currently, nothing is known about the genetic relationships between native and introduced populations
of these target weeds. More importantly, with the complexity of the genus, the present taxonomic
identity of individuals is questionable. In collaboration with national research agencies, plant material
of these species is being collected from populations in native and introduced ranges. Phylogeography
is being explored using chloroplast DNA sequences in combination with ploidy determination and first
data presented in this paper.
CAB International 2008
448
Abstracts: Theme 6 Evolutionary Processes
Population genetics of invasive North American diffuse
and spotted knapweed (Centaurea diffusa and C. stoebe)
R.A. Hufbauer,1 R.A. Marrs1 and R. Sforza2
1
Colorado State University, Department of Bioagricultural Science and Pest Management,
Fort Collins, CO 80523-1177, USA
2
USDAARS, European Biological Control Laboratory, Campus Intl de Baillarguet,
CS90013, Montferrier-sur-Lez, 34988 St. Gely du Fesc, France
Knowing the possible origins of invasive weeds, whether multiple introductions have occurred, and
levels of genetic variation relative to the native range is vital to conducting rigorous tests of several hypotheses that underlie classical biological control. We explore the population genetics of two Eurasian
species that are invasive in North America, Centaurea diffusa and Centaurea stoebe, using variable
chloroplast DNA (cpDNA) sequences and microsatellite loci. C. diffusa has lower haplotype diversity
and cpDNA allelic richness in the introduced range relative to the native range, suggesting that the
introduction imposed a bottleneck in population size. However, variation at microsatellite loci does
not differ, and the data suggest a minimum of two introductions of C. diffusa. Three of the haplotypes
of C. stoebe found in North America match haplotypes in species other than C. stoebe from the native
range, suggesting the possibility of cryptic invasions. Additionally, C. diffusa and C. stoebe share several cpDNA haplotypes, including their most common haplotype, and they share most microsatellite
alleles. This suggests ongoing hybridization between the species or incomplete segregation of alleles.
These data can guide further exploration for the origins of these species and point out locations within
the introduced range with unique and diverse genetic makeup.
Morphological and genetic methods to differentiate
and track strains of Phoma clematidina on
Clematis in New Zealand
H.M. Harman, N.W. Waipara, H. Kitchen, R.B. Beever,
B. Massey, S. Parkes and P. Wilkie
Landcare Research Ltd, Private Bag 92170, Auckland, New Zealand
A highly pathogenic strain of the leaf pathogen, Phoma clematidina (Thm.) Boerema, has been deliberately introduced to New Zealand from North America for biocontrol of old mans beard, Clematis vitalba L. However, the disease levels of this biocontrol agent have been inconsistent, and it
is sometimes present as a symptomless endophyte. Local strains of P. clematidina that are mildly or
non-pathogenic on C. vitalba were present in New Zealand on C. vitalba and native Clematis species
before initiation of the biological control programme. An understanding of how the introduced virulent
biocontrol strain is interacting with the avirulent endemic strains, and how this interaction is impacting on the pathogenicity and disease expression on C. vitalba is critical to fully evaluate biocontrol
efficacy of the disease. We are using morphological and genetic methods to differentiate local strains of
P. clematidina from the exotic biocontrol strain. These are also being used to determine the distribution, pathogenicity and host specificity of the different pathogenic strains present in New Zealand.
Marked P. clematidina strains are being developed to track the fungus in the field to further understand the epidemiology of P. clematidina leaf disease on Clematis.
449
XII International Symposium on Biological Control of Weeds
Polyploidy, life cycle, herbivory and invasion success:
work on Centaurea maculosa
H. Mller-Schrer,1 H. Bowman Gillianne,1 U. Treier,1 C. Bollig,1
U. Schaffner2 and T. Steinger1
1
Universit de Fribourg, Unit dEcologie & Evolution, Departement de Biologie,
1700 Fribourg, Switzerland
2
CABI Bioscience Centre, 2800 Delmont, Switzerland
The knapweed, Centaurea maculosa, has been introduced from Europe (EU) into North America (NA)
during the late 19th century, where it has become a prominent rangeland weed. Flow cytometry studies
of populations sampled in EU and NA revealed diploid (2x) and tetraploid (4x), as well as few mixed
populations in EU but, so far, only 4x populations in NA. Field observations suggest that 2x populations are predominantly monocarpic and 4x populations polycarpic. Age structure using herb chronology will be presented for various populations. In the greenhouse, we are growing plants of 77 native
EU populations, both 2x and 4x, and of 23 invasive 4x NA populations, conducting performance tests
with specialist and generalist herbivores and analyzing defence traits. We specifically explore the link
between ploidal level, life history traits, phenotypic plasticity and reproductive strategy to investigate
trade-offs with defence traits. To test if a polycarpic habit has been negatively selected by specialist
herbivores in the native but positively in the introduced range, where specialist herbivores are absent,
we started a 2x vs 4x competition experiments in the presence and absence of herbivores. The results
will be integrated with information from niche modelling and community invasibility experiments.
Use of morphometrics and multivariate analysis for
classification of Diorhabda ecotypes from the old world
J. Sanabria,1 J.L. Tracy,2 T.O. Robbins2 and C.J. DeLoach2
1
Texas A&M University-Blackland Research Center, 720 E. Blackland Road, Temple, TX 76502, USA
2
USDAARS, 808 E. Blackland Road, Temple, TX 76502, USA
Six years of data have shown high potential of Diorhabda elongata as an effective biological control of
Tamarix spp. in some regions of the USA. There is evidence that five ecotypes may represent different
sibling species. Consequently, taxonomic studies of the saltcedar beetle are critical in Tamarix control
programs. In addition, there is disagreement among taxonomists about the existence and number of
D. elongata sibling species. Five genitalic ecotypes based on morphology of the genitalia are reported:
elongata, carinata, sublineata, carinulata and meridionalis. These ecotypes may be suitable for control
of Tamarix in differing biogeographic areas of the western USA. We developed a classification system
of Diorhabda ecotypes based upon measurements of both genitalic and external structures using a
combination of factor and cluster analyses. The first factor associated with 59% of the variability is
explained by external body parts; the other four factors are associated with genitalic measurements
and together explained 34.24% of the data variability. The cluster analysis was able to reproduce a
good separation of the 85 specimens into five ecotypes. A dendrogram constructed from the analysis
shows the highest affinity between the carinata and sublineata morphotypes and the highest difference
between elongata and the rest of ecotypes.
450
Abstracts: Theme 6 Evolutionary Processes
Why are there no species-specific natural enemies
for giant hogweed?
M.K. Seier1 and M.J.W. Cock2
CABI UK Centre, Silwood Park, Buckhurst Road, Ascot, Berkshire SL5 7TA, UK
2
CABI Switzerland Centre, Rue des Grillons 1, 2800 Delmont, Switzerland
Based on surveys for and laboratory studies of the insect herbivores and fungal pathogens recorded
from giant hogweed, Heracleum mantegazzianum Sommier & Levier (Apiaceae) in the Caucasus as its
native range, we assess the potential for classical biological control of giant hogweed in Europe. Surveys revealed a guild of natural enemies, arthropods and pathogens, associated with the target plant and
other Heracleum spp. in the western Caucasus Mountains. However, none of the evaluated insects and
pathogens was considered to be suitably host-specific for introduction into Europe. A hypothesis is proposed to explain the absence of monospecific natural enemies of giant hogweed in the Caucasus, based
on the dynamic and interactive evolution of populations of closely related and hybridizing species of
Heracleum spp. in this mountain range over successive glaciation events during the Pleistocene.
Specificity and plant host phenology: the case of
Gephyraulus raphanistri (Diptera: Cecidomyiidae)
J. Vitou,1 J.K. Scott2 and A.W. Sheppard3
CSIRO-EL, Campus de Baillarguet, 34980 Montferrier sur Lez, France
CSIRO Entomology, Private Bag 5, PO Wembley, Western Australia 6913, Australia
3
CSIRO Entomology, GPO Box 1700, Canberra, Australian Capital Territory 2601, Australia
1
Wild radish (Raphanus raphanistrum L.) is one of the most important weeds of crops in southern Australia. The potential for classical biological control of this weed was investigated, and recent confirmed
host records show that the flower gall midge, Gephyraulus raphanistri, is restricted to R. raphanistrum
throughout Europe. G. raphanistri has never been confirmed from Canola in Europe, where 3 million
hectares are grown each year. Field host specificity testing G. raphanistri by manipulating host plant
phenology of actual and potential hosts in the genera Raphanus and Brassica revealed that no host
plant preference was observed. All tested species, Raphanus raphanistrum raphanistrum (wild radish),
Raphanus raphanistrum landra (coastal wild radish), Raphanus sativus (radish) and Brassica napus
(an oilseed rape cultivar) were resynchronized for initial flowering to the natural R. raphanistrum
landra plants hosting a natural population of G. raphanistri. The high field host specificity observed in
this gall midge in Europe is driven by synchrony of oviposition and flower availability. When phenologically resynchronized, Canola was an equally acceptable host in the field for oviposition and larval
development. In Australia, the new environment might generate new phenological conditions and thus
significantly increase the risk associated with this midge as a biological control agent.
451
XII International Symposium on Biological Control of Weeds
Comparative invasion histories of Australians
invading South Africa
J.R.U. Wilson,1 D.M. Richardson,1 A.J. Lowe,2,3 T.A.J. Hedderson,4 J.H. Hoffmann,5
A.W. Sheppard,6 A.B.R. Witt7 and L.C. Foxcroft8
Stellenbosch University, Department of Botany & Zoology, Centre for Invasion Biology,
Stellenbosch, 7602, South Africa
2
University of Adelaide, School of Earth and Environmental Science, Adelaide
3
Adelaide Botanic Gardens, Plant Biodiversity Centre, Adelaide, Australia
4
University of Cape Town, Botany Department, Cape Town, South Africa
5
University of Cape Town, Zoology Department, Cape Town, South Afric
6
CSIRO Entomology, Canberra, Australia
7
ARC-Plant Protection Research Institute, Pretoria, South Africa
8
South African National Parks, Savanna Ecological Research Unit, Kruger National Park,
Skukuza, South Africa
An increasing number of studies are exploring the phylogeography of invasive alien species and how
this may impact the success rate of biological control programs. We are in the first year of a comparative study looking at plants native to Australia and invasive in South Africa (and vice versa), focussing
on acacias in particular. We believe that, by comparing species, we can gain general insights that are of
value both to direct management and to our broader understanding of invasion biology. We also contend that the collection and curation of plant material samples suitable for genetic analysis should form
part of any biological control survey. We are keen to hear comments, share experiences and establish
collaborations.
452
Theme 7:
Opportunities and Constraints for the
Biological Control of Weeds
in Europe
Session Chair: Paul Hatcher
453
This page intentionally left blank
Keynote Presenter
Opportunities and constraints for the
biological control of weeds in Europe
M. Vurro1 and H.C. Evans2
Summary
Although there has been increasing interest in Europe over the last few decades in trying to harness
the potential of biological control as a management tool for weeds, securing funding for projects
continues to be problematic whilst the successes have been limited. The incentives to use alternative
technology are based on a number of interrelated factors, not least the dramatic rise in popularity of
organic products (linked to increasing environmental awareness) and a powerful anti-GMO lobby.
Combine this with new legislation to remove some chemical herbicides from the market or to restrict
their usage, as well as soaring development costs for new products, and the opportunities for biological control have never been better. Here, we analyse the reasons for the limited uptake, and the challenges or constraints facing weed biocontrol from two different approaches: classical and inundative.
For classical biological control against invasive alien weeds, Europe has lagged behind other continents (e.g. Australasia, North America) to the point that there have been no introductions thus far,
whilst those that are in the pipeline still need to clear considerable legislative hurdles. These issues
are highlighted for past and ongoing projects. For inundative biological control against indigenous
or naturalized weedsin this case, the use of products based on plant pathogens (bioherbicides)
the constraints are largely technological and commercial rather than bureaucratic: e.g. stability and
efficacy; costs of production and registration; and limited markets. Opportunities to support the study
and use of mycoherbicides, and strategies to overcome these constraintsimproving production, formulation and application systems; genetic enhancement; synergistic mixtures of agents/metabolites
are discussed.
Keywords: bioherbicides, classical biological control, environmental weeds.
Introduction
In a review of the biological control of weeds and its
prospect in Europe, Schroeder and Mller-Schrer
(1995) stated optimistically that: Although biological weed control so far [has] received little attention
in Europe, more recent developments indicate that this
may change in the near future. These developments included the commercialization of, and increasing potential for, mycoherbicide use, particularly in agricultural
systems in North America (Charudattan, 1991; Smith,
1991). In the intervening decade or so, has this optimism been realized? Certainly, the odds should have
moved favourably in this direction, not least because of
the current awareness of environmental issues in EuInstitute of Sciences of Food Production, CNR, via Amendola 122/O,
70125 Bari, Italy.
2
CABI E-UK Centre, Silwood Park, Ascot, Berkshire, SL5 7TA, UK.
Corresponding author: M. Vurro <[email protected]>.
CAB International 2008
1
rope by the public and politicians alike, leading to: a
dramatic rise in the popularity of organic products; a
powerful anti-GM lobby, negating the use of herbicideresistant crops; and new legislation to remove chemical
herbicides from the market, or to restrict their usage.
Here, this question is addressed, together with an
assessment of the opportunities and constraints for biological control of weeds in Europe.
Classical
Crispy concepts, soggy concerns: Classical (or inoculative) biological control of weeds has had a long history
and has been one of the main strategies for the management of invasive alien plants worldwide, apart from
Europe, especially using coevolved arthropod natural
enemies (McFadyen, 1998; Julien and Griffiths, 1998). In
contrast, the use of coevolved pathogens is still relatively new although there have been notable successes (Evans, 2002). The high success rate with arthropod agents
has also been extremely cost effective, with an impres-
455
XII International Symposium on Biological Control of Weeds
sive safety record considering that over 600 agents have
been released thus far (Marohasy, 1996). However, as
McClayand Balciunas (2005) point out, it is a highstakes game to achieve successful control of an invasive
weed without collateral damage to non-target plants or
other, more cryptic negative environmental impacts.
Paradoxically, if all of these issues were addressed,
no classical biological control projects would ever get
off the ground because of the prohibitive costs involved
in undertaking such an in-depth ecological impact assessment. Indeed, as with any biological system, predictive models are just thatinformed guesswork. There
must be transparent risk assessment and on the question of risk to non-target plants the centrifugal phylogenetic protocol originally proposed by Wapshere (1974)
has proven to be extremely robust (Marohassy, 1996;
Evans, 2000). However, as has been stressed recently,
classical biological control could be viewed as the introduction of yet another alien species: an issue
of growing public concern (Miller and Aplet, 2005).
Although these misgivings were aimed specifically at
the situation in the United States, where legal issues
concerning hidden regulations have been identified,
decision-makers in Europe would have taken note. In a
blame culture, perhaps this will color decisions about
future project proposals, not to say gaining permission
to release classical agents in Europe.
As highlighted by Schroeder and Mller-Schrer
(1995), although Europe has been a source of classical
biological control agents for use in other continents for
over 30 years, only two biological control programmes
had been implemented within Europe (Sheppard et al.,
2006). One of these programmes was targeted at invasive bracken (Pteridium aquilinum (L.) Kuhn) in the
UK but it foundered on legal, political and environmental issues (Lawton, 1988). A potential biological
control agent, a leaf-feeding lepidopteran from South
Africa, expired in quarantine after lengthy wranglings
over additional testing and release protocols. One
somewhat bizarre recommendation involved caged
releases on bracken infestations in the Isle of Man:
The suggested cage size was such that it was deemed
both physically and economically impractical (CIBC,
1988, unpublished). Based on weed biological control
success stories elsewhere, principally in Australia and
South Africa, as well as the changing public opinion
concerning biodiversity and conservation, Schroeder
and Mller-Schrer (1995) concluded that Europe was
ready for classical biological control. Amongst the
weeds they identified as potential targets were giant
hogweed, Japanese knotweed, and Himalayan balsam.
Earlier, Evans and Ellison (1990) also considered these
same invasive neophytes to be ideal candidates for this
management strategy in the UK because of their increasing environmental or amenity importance. Since
this time, biological control programmes have been
initiated against these three weeds and, therefore, they
make ideal examples to illustrate the opportunities and
the constraints for the classical biological control of
weeds in Europe.
Opportunities and constraints: examples from projects: Giant hogweed. Heracleum mantegazzianum
Sommier and Levier is both an environmental and a
health-threat in western Europe and an EU-funded,
multidisciplinary project on its ecology and management was initiated in 2002. The outputs of this threeyear programme have recently been published (Pysek et
al., 2007). Unfortunately the biological control component failed to positively identify any potential biological
control agents following surveys for natural enemies in
the plants centre of diversity, the Caucasus Mountains.
A damaging leaf-spot pathogen, Phloeospora heraclei
(Lib.) Petr. showed exceptional promise but further
evaluation was suspended when it was found to attack
several related crop species, including parsnip, in labbased screening trials. Nevertheless, questions remain
unanswered concerning its true (field) host range because this pathogen has never been reported as a disease
problem in cultivated parsnip, either in the UK or mainland Europe. Neither has it been found on H. mantegazzianum in its invasive range, despite records of its occurrence on indigenous Heracleum spp. (Seier and Evans,
2007). It would appear, therefore, that pathotypes or
formae speciales specific to H. mantegazzianum occur.
However, because of scientific and legislative uncertainties, as well as negative public perception or wariness about pathogens (pathophobia), it was decided that
this would not be a good model system to launch the
classical biological control concept in Europe although
it was also concluded that: This does not negate the
potential for weed biological control in Europe (Cock
and Seier, 2007). Nonetheless, how the decision makers
and stakeholders in the EU have viewed this failure to
deliver remains unclear but it is tempting to speculate
that it will only provide additional ammunition for critics of classical biological control.
Japanese knotweed. This project, funded by a
consortium of UK environmental agencies and local
stakeholders, is still ongoing and has now reached the
critical phase of submitting the research data on two
Japanese natural enemies of Fallopia japonica (Houtt.)
R. Decr. for permission to release in the UK, following a four-year assessment. One of these is a hemibiotrophic fungus (Mycosphaerella sp.), ubiquitous and
damaging in Japan and specific to the target weed (Kurose et al., 2006). The legislative hurdles still to clear
are discussed comprehensively elsewhere in this Proceedings (see Ehlers, R.), as well as in the literature
(Sheppard et al., 2006), and at first sight appear to be
daunting. However, as well as a test case, this project
could serve as the flagship to launch classical biological control into European waters. It is now very much
a high-profile weed in the UK, receiving considerable
publicity because of its diverse impactsin natural
ecosystems, as a riparian invader; in business/industrial
situations, where it dramatically increases the costs of
456
Opportunities and constraints for the biological control of weeds in Europe
construction work; and last but not least, at the urban
or grassroots level, contaminating gardens and devaluing property. It has also been in the public domain
since London was chosen as the venue for the 2012
Olympics, as the Lea Valley olympic site is infested
with both Japanese knotweed and giant hogweed. Initial estimates have put the costs of clearing knotweed
from an area smaller than an Olympic swimming pool
at around 50,000 or US$100,000 (The Guardian, 13
Sept. 2005). More recent reports on treating the whole
site to remove F. japonica declared that it would add
nearly 70 million to the already burgeoning and highly controversial budget (Sunday Times, March 2007).
Such is its infamy that F. japonica has now entered the
national psyche and vernacular, with one newspaper
choosing to describe the success of a certain supermarket chain as . . . spreading through the UK like
knotweed (The Guardian, 19 April 2007). In fact, an
official study on the costs involved in the conventional
control of this weed in the UK has arrived at the worryingly precise figure of 1.56 billion (DEFRA, 2003).
Himalayan balsam. This is the new project on the
block, receiving seed money from several UK agencies
to undertake a preliminary survey in parts of its native
range in the lower Himalayan region of Pakistan. The
results have proven to be extremely encouraging with
several fungal and insect natural enemies showing good
potential. However, there is still considerable onus on
the scientists to raise the stakes with the donors in order
to achieve the funding necessary to implement a viable
biological control programme against this increasingly
problematic and invasive riparian weed in the UK.
Inundative
Stability of research: Besides the scientific constraints that limit weed biological control in Europe,
the main problems are the low number of stable or established groups working on inundative weed biocontrol strategies in Europe and the low number of projects
that are or have been internationally funded on this topic. The European Framework Programmes, which represent the main source of cooperative research funds,
have supported only a limited number of projects
specifically dealing with biological control of weeds
(http://cordis.europa.eu/en/home.html). In particular,
within the 5th Framework only one project was funded.
This has been mirrored in the recently completed 6th
Framework, where only one project devoted to enhancing and exploiting biocontrol agents, including weed
pathogens, was funded. Within the COST programme
(European Co-operation in the Field of Scientific and
Technical Research), one of the longest-running instruments supporting cooperation among scientists and
researchers across Europe, few projects received funding, including Biological control of weeds in Europe
(COST 816, 1994-1999) and Parasitic plant management in sustainable agriculture (COST 849, 2001-
2006), which included a biological control component
(http://www.cost.esf.org/).
Favorable and unfavourable legislation: Thanks to
the enforcement of the pertinent EU directives (e.g.
91/414, 2002/2076), many dangerous pesticides have
been banned in the EU countries in recent years or are
scheduled to be banned within 2007. Since 2001, the
use of 127 active ingredients has already been prohibited, and this list should lengthen by 14 more by the
end of 2007. Among the 191 compounds recognized as
herbicidal active ingredients, 71 of them were already
excluded by the so-called Annex 1, which is the list
of the permitted herbicides, and many others are under
evaluation for their prospective exclusion (data gathered from http://fitorev.imagelinenetwork.com). This
should give a renewed impetus to research on mycoherbicides and could revitalize public interest in weed
biological control.
However, the Commission Directive 2001/36, which
amended the Council Directive 91/414/EEC specifically for biopesticides, is very restrictive with regard
to the procedures of risk assessment, registration and
use of microbial plant-protection products. This is a
further and potent reason why no microbial products
are currently included as bioherbicides in the register kept by the Directorate of the Consumer Health
Protection.
Choice of suitable targets and efficacious agents: The
choice of an appropriate target weed is of utmost importance for the success of any biocontrol programme.
Weeds that cannot be controlled by traditional methods
or by chemicals due to resistance and/or environmental
factors should be the preferred targets. The lack of alternative control strategies should increase the acceptability of a biological control agent, even if the price is higher due to a more sophisticated method of application.
For example, parasitic plants such as Orobanche spp.
could represent ideal targets. These weeds represent an
unsolved and increasing problem in many countries of
the Central European and Mediterranean regions due to
their complex life cycle and lack of available control
methods, especially in light of the ban on methyl bromide and other dangerous soil fumigants. In addition,
perennial weeds, such as Cirsium arvense (L.) Scop. or
Sonchus arvensis L., which occur throughout Europe
and whose control appears particularly problematic
because of their vegetative spread by subterranean organs, could be appropriate targets at the European level.
Even plants spreading in anthropic environments or
causing health problems (e.g. the allergenic plant Ambrosia artemisiifolia L.) could be perfect targets for biological control.
The search for suitable agents has not been pursued
to the extent that it has been in other parts of the world,
due mainly to funding restrictions (see above). The
Mediterranean basin or the Caucasian region, for example, has more often represented sources of agents for
classical weed biocontrol programmes overseas rather
457
XII International Symposium on Biological Control of Weeds
than has been considered a source of potential mycoherbicides for European programmes.
Improving production: Mycoherbicides can only
compete successfully with chemical herbicides if the
products are as effective as the chemicals they seek to
replace and if they are not significantly more expensive. Apart from the efficacy of the strain of the microorganism used, this is mainly dependent on how it
is produced and formulated. The production technology used must ensure the highest possible yield of
live propagules and the formulation must be such that
an application of the propagules to the soil or to the
plant is as easy or nearly as easy as the application of
a chemical pesticide. For the development of biological control agents it is usually stated that the formulation must improve or at least assist the effectiveness of
the microorganism and must ensure a shelf life of the
product of at least one year, preferably two or more
years. Most probably the main element is cost-effectiveness in order to make the new product competitive
with current technologies. With modern inventory control, shelf life is less important, and if a product is truly
efficacious but requires special application techniques
or equipment, farmers will make the necessary investments just as they have for other specialized machinery
now regularly used in modern agriculture.
Suitable culture media for the production of fungal
propagules should be selected using an appropriate fermentation technology, followed by: evaluation of the
most suitable growth conditions; selection of the best
technology to separate the propagules from the fermentation product; evaluation of the most suitable methods
and conditions for the formulation of the propagules
produced; and assessment of the shelf life of the formulated products. The use of biomasses obtained as
residues of industrial food processing or agricultural
practices could reduce the cost of production. Molasses, exhausted olive cakes, residues from breweries
and waste from tinned fruit industries are a few examples of such by-products that still contain nutrients
and could be exploited to grow microbes. Often their
disposal presents both an economic and environmental
problem. Therefore, they could be ideal and inexpensive media for the production of many microbial biocontrol agents.
Application systems: One of the main problems in using microbial biocontrol agents to control weeds is to
find suitable methods of application that allow the uniform distribution of the agent at the desired site, without
waste or excessive consumption of the mycoherbicide
that would increase the cost of treatment and the risks
of non-target effects. The uniform and precise application of microbial particles close to the target weed and
to the crop to be protected can increase the success of a
biological control treatment, and the use of systems or
technologies which are usually available in agriculture
could influence the acceptability of biocontrol agents by
farmers and enlarge the market. For example, the use
of drip irrigation systems for the application of suspensions containing conidia of potential mycoherbicides
has recently been suggested (Boari et al., 2007). An
advantage of using propagules of soil-borne pathogens
that normally infect at or below the soil surface is that
the propagules may be more protected from environmental factors such as wind and UV radiation, which
can negatively affect conidial viability and uniformity
of distribution. Applying fungal inoculum by drip irrigation does not require growers to invest in new equipment for application since this strategy is already quite
widely used in agriculture to supply water, nutrients or
chemicals, especially in vegetable crops where perennial and parasitic weeds often represent difficult problems. A further advantage could be the limiting of the
applied doses to the crop root zone and not to the whole
field, and therefore a reduction in the cost of treatment.
Several potential mycoherbicides could be applied
at the soil level (Charudattan, 2001), as could microbial
antagonists (Whipps and Lumsden, 2001) and biopesticides (Copping, 1998), during transplanting or through
soil-drenching or root-dip, although these techniques
of application can be expensive. As the fungal community already in the soil can affect the persistence of
microbial treatments, longer watering intervals involving multiple treatments with lower concentrations of
spores could be considered. This would result in a better distribution of the microbes in the soil in terms of
volume of protected soil and amount of inoculum and
reduce the risk of clogging the dripper.
Leaf-applied mycoherbicides could take advantage of sophisticated technologies, such as the use of
advanced optics and computer assistance to sense if a
weed is present. In this way, a precise amount of mycoherbicide could be applied only to the weeds and not
wasted on bare ground. Such systems could be used
where weeds occur intermittently, optimizing the consumption of spray suspension, and thus reducing the
treatment costs.
Potential for genetic enhancement of pathogen biological control agents: Several transformation-based
techniques allow reproducible genetic modifications
in fungi. It should be possible to knock out genes in a
biological control agent, as well as to transfer specific
genes into it, and then determine effects on pathogenicity/virulence. Gressel et al. (2007) have recently
inserted into some promising biological control agents
genes considered soft, such as those encoding carbohydrases, auxin and oxalate, or hard such as those encoding phytotoxins.
Physiological enhancement of biological control activity: Different approaches are being used to increase
the efficacy of biological control agents without using
genetic or transgenic manipulation. The transgenically
enhanced hypervirulence of a biological control agent
has the advantage of constitutiveness; it is already present, and there is no need for additives. Conversely, if
the same effect can be achieved physiologically with
458
Opportunities and constraints for the biological control of weeds in Europe
an additive, then there is the advantage that the organism is no different from the wild type after the additive
has gone. For example, an organism could be engineered to overproduce oxalate (Gressel, 2002) in order to overcome calcium-dependent weed defences, or
the biological control organism could be provided with
exogenous oxalate to achieve similar hypervirulence
(Gressel et al., 2002); yet, the organism lacks hypervirulence when the oxalate is not present.
Another possibility is the use of natural mutants,
such as those that are able to overproduce and excrete
amino acids that are inhibitory to the target plant, resulting in enhanced virulence and improved efficacy of
the bioherbicide (Thompson et al., 2007).
Environmental impact: One of the main constraints to the release of microbes in the environment
for weed biological control is the lack of knowledge
about issues such as the fate of the strains after their
release into the environment, their stability and the
risk of dispersal. The environmental impact of a variety of biological control agents can be assessed
by tracking their movement, assaying non-target
\effects and any changes in host range (especially after genetic or physiological modifications), together
with determining long-term environmental persistence.
The introduction of biological control agents into soil
may pose a risk of unforeseen or detrimental activities on the soil microbial population. The EU Directive
2001/36 states clearly that side effects on non-target soil
microorganisms should be addressed, but there are no
validated methods available. Until recently, techniques
for monitoring direct effects on microorganisms have
been restricted to in vitro culture-based methods that
ignored 90% or more of the microbial population that
could not grow on culture media in the laboratory. The
development of DNA-fingerprinting techniques makes
it possible to compare the genomes of all strains and to
use molecular markers to recognise strains of biological
control agents after their release into the environment.
The study of the microbial community composition can
be based on the direct extraction of DNA from soils or
other complex matrices. Practically, techniques such
as terminal restriction fragment length polymorphism
analysis (T-RFLP), ribosomal intergenic spacer analysis (RISA) and AFLP are relatively rapid, economically
feasible, and within the technological capabilities of
most microbiological research laboratories. The genetic
diversity within species can also be determined by using DNA molecular analyses such as sequencing of nuclear ribosomal DNA, or the beta-tubulin, calmodulin,
or elongation factor genes. In addition, real-time PCR
techniques can be set up for the qualitative/quantitative
detection of DNA from biological control agents (Anderson and Cairney, 2004).
Integration between diverse biological control
agents, bioactive fungal metabolites, herbicides, or
other chemicals: An important factor that can reduce
the efficacy of a potential mycoherbicide is the ability
of the target weed to resist invasion and colonization
by the biological control agent. Several attempts have
been made to combine mycoherbicides with bioactive
metabolites, in order to enhance agent efficacy. Such
combinations can suppress or weaken plant defence
mechanisms by blocking the synthesis of secondary
plant metabolites or breaking down physical barriers to
pathogen attack, resulting in increased biological control (Duke et al., 2007, and references therein cited).
The effect of herbicides on plant disease is an important but generally overlooked aspect of integrated
weed management. Nevertheless, understanding herbicide/plant pathogen interactions can be critical in
designing effective and efficient integrated weed management programmes.
Synthetic herbicides have the potential to influence
plant disease by several mechanisms. It is not unusual
for low rates of herbicides to stimulate in vitro pathogen growth or sporulation (Wyss et al., 2004; Yandoc
et al., 2006). On the other hand, herbicides such as
glyphosate can also be very effective at lowering plant
resistance to pathogens and acting as a synergist for
microbial weed biological control products (Duke et
al, 2007). In such strategies, dose rates are likely to be
highly important to both the direct and indirect effects
of herbicides on plant disease.
Enhanced bioherbicidal efficacy of Exserohilum
monoceras (Drechs.) K. J. Leonard and E. G. Suggs on
Echinochloa crus-galli (L.) Beauv., a weed in paddy
rice (Oryza sativa L.), was obtained when the fungus
was applied with -aminolevulinic acid, a precursor
of tetrapyrolescompounds which are involved in
the bleaching and killing of plant tissue (Hirase et al.,
2006). A system to integrate low doses of glyphosate
with a foliar desiccant (ammonium sulphate) and the
biological control agent Alternaria destruens (Smolder) to control Cuscuta pentagona Engelm., has also
been reported (Cook et al., 2005).
The efficacy of a weak biological control agent,
Colletotrichum coccodes (Wallr.) S. Hughes, on velvetleaf (Abutilon theophrasti Medik.) was improved
by applying calcium chelators that repressed host-plant
defences by reducing callose formation (Gressel et al.,
2002). Also phytotoxic metabolites produced by plant
pathogens can weaken defence mechanisms of plants,
rendering them more susceptible to pathogen attack.
Thus, the application of toxins jointly with the pathogens could strongly enhance their bioherbicidal properties (Vurro, 2007).
A mixture of three host-specific pathogens: Alternaria cassiae (specific to sicklepod), Phomopsis
amaranthicola Rosskopf, Charudattan, Shabana &
Benny (specific to pigweeds), and Colletotrichum
dematium f. sp. crotalariae (specific to showy crotalaria), proved to be very efficacious in the simultaneous control of the three weeds (Chandramohan
and Charudattan, 2003). Good control of seven
weedy grass species has also been obtained using a
459
XII International Symposium on Biological Control of Weeds
suspension containing conidia of three host-specific
pathogens: Drechslera gigantea (Heald and Wolf)
Ito, Exserohilum rostratum (Drechsler) Leonard and
Suggs, and Exserohilum longirostratum Subram.
(Chandramohan et al., 2002).
Discussion
Perhaps we can borrow the use of the Anna Karenina
Principle from McClay and Balciunas (2005), who
first applied it to biological control [the list of borrowers, of course, goes back to Tolstoy (1877)] in order
to compare the constraints and opportunities for classical vs inundative biological control. For the inundative
approach, these are essentially similar for every weed
target in every country, region or continent: happy
families are all alike, and the issues involved have been
addressed here. However, for the classical approach,
especially against invasive environmental weeds, the
factors involved are extremely, and often uniquely,
complex and therefore must be dealt with on a caseby-case basis: every unhappy family is unhappy in its
own way.
In the European context, many of these issues have
already been highlighted and reviewed comprehensively by Sheppard et al. (2005). Suffice it to say that
no biological control agents have been released thus
far, and the few that are in the pipeline face an uphill
struggle and uncertain future for acceptance, despite the
fact that: Classical biological control remains the only
tool available for permanent ecological and economic
management of invasive alien species (Sheppard
et al., 2005). This approach has even received the seal
of approval from the Convention on Biological Diversity (CBD), the European and Mediterranean Plant Protection Organisation (EPPO) and the European Strategy on Invasive Alien Species (ESIAS). However, is
this the kiss of death, as the bureaucratic red tape kicks
in? As previously mentioned, there are now so many
more environmental concerns to address, compared to
earlier times, that the costs of implementing all of them
would put any biological control project out of the financial reach of traditional donors. Certainly, we have
moved on from the hunter-gatherer, quick-release,
lets-try-this-one approach to the position where it is
essential to abide by the CBD and to undertake scientifically driven risk assessments. These are in place but
still subject to the whims and interpretations of individual countries and international organisations, as well
as the critics of biological control. In the present climate, it would still take only one mistake, or unexpected
non-target issue, to seriously undermine the solid scientific foundations on which classical biological control is based.
In the case of the CBD, this has recently created additional barriers hampering free exchange of germplasm
between countries. For example, permission to release
an Argentinian strain of the rust, Puccinia spegazzinii
DeToni for use against the highly invasive mile-a-minute
weed (Mikania micrantha Kunth) in China has been
blocked, seemingly permanently, by Misiones Department, which has a separate CBD policy from that of
Argentina. Thus, biological control scientists are now
expected to negotiate delicate political issues in order
to implement classical projects. In addition, there are
constant attacks on or criticisms of the safety-testing
procedures employed to screen classical weed biological control agents, despite an impeccable track record
(McFadyen, 1998), as well as new concerns about the
indirect impacts of even host-specific agents on nontarget species (Pearson and Callaway, 2003; Louda et al.,
2003). These concerns provide further ammunition for
biological control sceptics to shoot down any proposals, which are based solely on the classical biological control approach, before they have gotten off the
ground. Moreover, this relates only to relatively inoffensive insect agents! What hope is there for pathogens? It could be argued that if all the environmental
concerns and risks involved in undertaking a motorized
shopping trip were analysed as critically, supermarkets
would go out of business (even the ones spreading like
knotweed)!
As flagged by Thomas et al. (2004), such concerns
serve to highlight . . . the need for proper ecological
and socioeconomic evaluation of pests before control
to determine probable costs and benefits. The financial stakes are raised yet again, as well as time frames,
before biological control can even be considered as
a management option, putting such proposals on a
different donor level. The giant hogweed project, for
example, was large by EU standards because of its
multidisciplinary approach and composition. Even so,
fundamental questions relating to the safety of several potential biological control agents still remained
unanswered because of insufficient time and funding.
Thus, we are left with the possibility that future classical weed biological control projects for Europe need
to be multidisciplinary and last a minimum of 510
years to achieve all the goals now expected, once the
target posts have been moved. Such multi-million euro
proposals must thus be the norm if proposals are
to be approved and project aims realized. It is no
coincidence, perhaps, that the multinational, multiorganisational, multidisciplinary biological control
project against migratory locusts did achieve its objectives, but only after a massive injection of funds
from a consortium of international donors for over
more than a decade.
This leads finally to the perennial question: biological controlrisky or necessary? (Thomas and Willis,
1998). Critical decisions need to be taken in Europe regarding the long-term management of invasive weeds,
especially those with serious environmental impacts,
but perhaps it is better not to fiddle around too much
while the aliens continue their destruction of fragile
ecosystems.
460
Opportunities and constraints for the biological control of weeds in Europe
References
Anderson, I.C. and Cairney, J.W. (2004) Diversity and ecology of soil fungal communities: increased understanding
through the application of molecular techniques. Environmental Microbiology 6, 769779.
Boari, A., Zuccari, D. and Vurro, M. (2007) Microbigation:
delivery of biological control agents through drip irriga
tionsystems. Irrigation Science, http://dx.doi.org/10.1007/
s00271-007-0076-x.
Chandramohan, S. and Charudattan, R. (2003) A multiplepathogen system for bioherbicidal control of several
weeds. Biocontrol Science and Technology 13, 199205.
Chandramohan, S., Charudattan, R., Sonoda, R.M. and
Singh, M. (2002) Field evaluation of a fungal mixture
for the control of seven weedy grasses. Weed Science 50,
204213.
Charudattan, R. (1991) The mycoherbicide approach with
plant pathogens. In: TeBeest, D.O. (ed.) Microbial Control
of Weeds. Chapman and Hall, New York, USA, pp. 2457.
Charudattan, R. (2001) Biological control of weeds by means
of plant pathogens: significance for integrated weed management in modern agro-ecology. BioControl 46, 229
260.
Cock, M.J.W. and Seier, M.K. (2007) The scope for biological
control of giant hogweed, Heracleum mantegazzianum.
In: Pysek, P., Cock, M.J.W., Nentwig, W., and Ravn, H.P.
(eds) Ecology and Management of Giant Hogweed. CABI
Publishing, Wallingford, UK, pp. 255271.
Cook, J., Charudattan, R., Rosskopf, E., Zimmerman, T.,
MacDonald, G. and Stall, W. (2005) Integrated control
of dodder (Cuscuta pentagona) using glyphosate, ammonium sulphate, and the biological control agent Alternaria
destruens. In: Proceeding of the 40th Annual Caribbean
Food Crops Society Meeting. USWI, St. John. [Abstract.]
Copping, L.G. (1998) The BioPesticide Manual. BCPC,
Berkshire, UK. 333 p.
DEFRA (2003) Review of Non-Native Species Policy: Report
of the Working Group. DEFRA Publication, London, UK.
90 p.
Duke, S.O., Wedge, D.E., Cerdeira, A.L. and Matallo, M.B.
(2007) Interactions of synthetic herbicides with plant disease and microbial herbicides. In: Vurro, M. and Gressel, J. (eds) Novel Biotechnologies for Biocontrol Agent
Enhancement and Management. Springer, Dordrecht, The
Netherlands, pp. 277296.
Evans, H.C. (2000) Evaluating plant pathogens for biological
control of weeds: an alternative view of pest risk assessment. Australasian Plant Pathology 29, 114.
Evans, H.C. (2002) Biological control of weeds. In: Kempken, D. (ed.) The Mycota XI: AgriculturaI Applications.
Springer-Verlag, Berlin, Germany, pp.135152.
Evans, H.C. and Ellison, C.A. (1990) Classical biological
control of weeds with micro-organisms: past, present,
prospects. Aspects of Applied Biology 24, 3949.
Gressel, J. (2002) Molecular Biology of Weed Control. Taylor
& Francis, London, UK. 520 p.
Gressel, J., Michaeli, D., Kampel, V., Amsellem, Z. and Warshawsky, A. (2002) Ultralow calcium requirements of
fungi facilitate use of calcium regulating agents to suppress host calcium-dependent defences, synergizing infection by a mycoherbicide. Journal of Agriculture and
Food Chemistry 50, 63536360.
Gressel, J., Meir, S., Herschkovitz, Y., Al-Ahmad, H.,
Greenspoon, I., Babalola, O. and Amsellem Z. (2007)
Approaches to and successes in developing transgenically enhanced mycoherbicides. In: Vurro, M. and Gressel,
J. (eds) Novel Biotechnologies for Biocontrol Agent
Enhancement and Management. Springer, Dordrecht,
The Netherlands, pp. 297305.
Hirase, K., Nishida, M. and Shinmi, T. (2006) Effect of aminolevulinic acid on the herbicidal efficacy of foliarapplied MTB-951, a mycoherbicide to control Echinochloa crus-galli L. Weed Biology and Management 6,
4449.
Julien, M.H. and Griffiths, M.W. (1998) Biological control
of weeds: a world catalogue of agents and their target
weeds. CABI, Wallingford, UK. p 223.
Kurose, D., Renals, T., Shaw, R., Furuya, N., Takagi, M.
and Evans, H. (2006) Fallopia japonica, an increasinglyintractable weed problem in the UK: can fungi
help cut through the Gordian knot? Mycologist 20,
126129.
Lawton, J.H. (1988) Biological control of bracken in Britain:
constraints and opportunities. Philosophical Transactions
of the Royal Society B, 318, 335350.
Louda, S.M., Pemberton, R.W., Johnson, M.T. and Follet,
P.A. (2003) Nontarget effects the Achilles heel of
biological control? Annual Review of Entomology 48,
365396.
Marohasy, J. (1996) Host shifts in biological weed control:
real problems, semantic difficulties or poor science? International Journal of Pest Management 42, 7175.
McClay, A.S. and Balciunas, J.K. (2005) The role of prerelease efficacy assessment in selecting biocontrol agents
for weeds applying the Anna Karenina principle. Biological Control 35, 197207.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
Miller, M.L. and Aplet, G.H. (2005) Applying legal sunshine
to the hidden regulations of biological control. Biological
Control 35, 358365.
Pearson, D.E. and Callaway, R.M. (2003) Indirect effects of
host specific biological control agents. Trends in Ecology
and Evolution 18, 456461.
Pysek, P., Cock, M.J.W., Nentwig, W. and Ravn, H.P. (2007)
Ecology and Management of Giant Hogweed. CABI Publishing, Wallingford, UK. 324 p.
Schroeder D. and Mller-Schrer, H. (1995) Biological control of weeds and its prospectives in Europe. Medical Facultade Landbouw, Universdat Gent 60, 117123.
Seier, M.K. and Evans, H.C. (2007) Fungal pathogens associated with Heracleum mantegazzianum in its native and
invaded distribution range. In: Pysek, P., Cock, M.J.W.,
Nentwig, W., and Ravn, H.P. (eds) Ecology and Management of Giant Hogweed. CABI Publishing, Wallingford,
UK, pp. 189208.
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006)
Top 20 environmental weeds for classical biological
control in Europe: a review of opportunities, regulations and other barriers to adoption. Weed Research 46,
93117.
Smith, R.J. (1991) Integration of biological control agents
with chemical pesticides. In: TeBeest, D.O. (ed.) Microbial Control of Weeds. Chapman and Hall, New York,
USA, pp.189208.
461
XII International Symposium on Biological Control of Weeds
Thomas, M.B. and Willis, A.J. (1998) Biocontrol risky but
necessary? Trends in Ecology and Evolution 13, 325329.
Thomas, M.B., Casula, P. and Wilby, A. (2004) Biological
control and indirect effects. Trends in Ecology and Evolution 19, 61.
Thompson, B.M., Kirkpatrick, M.M., Sands, D.C. and Pilgeram, A.L. (2007) Genetically enhancing the efficacy of
plant pathogens for control of weeds. In: Vurro, M. and
Gressel, J. Novel Biotechnologies for Biocontrol Agent
Enhancement and Management. Springer, Dordrecht, The
Netherlands, pp. 267275.
Tolstoy, L. (1877) Anna Karenina. Ruskii Vestnik, Moscow,
Russia.
Vurro, M. (2007) Benefits and risks of using fungal toxins
in biological control strategies. In: Vurro, M. and Gressel, J. (eds) Novel Biotechnologies for Biocontrol Agent
Enhancement and Management. Springer, Dordrecht, The
Netherlands. pp. 5374.
Wapshere, A. J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology, 77, 201211.
Whipps, J.M. and Lumsden, R.D. (2001) Commercial use of
fungi as plant disease biological control agents: status and
prospects. In: Butt, T.M., Jackson, C. and Magan, N. (eds)
Fungi as Biocontrol Agents: Progress, Problems and Potential. CABI Publishing, Wallingford, UK, pp. 922.
Wyss, G., Rosskopf, E. N., Charudattan, R. and Littell, R.
(2004) Effects of selected pesticides and adjuvants on
germination and vegetative growth of Phomopsis amaranthicola, a biocontrol agent for Amaranthus spp. Weed
Research 44, 114.
Yandoc, C. B., Rosskopf, E. N., Pitelli, R. L. and Charudattan, R. (2006) Effects of selected pesticides on conidial
germination and mycelial growth of Dactylaria higginsii,
a potential bioherbicide for purple nutsedge (Cyperus rotundus). Weed Technology 20, 255260.
462
Could Fallopia japonica be the first target
for classical weed biocontrol in Europe?
D.H. Djeddour,1 R.H. Shaw,1 H.C. Evans,1 R.A. Tanner,1 D. Kurose,2
N. Takahashi2 and M. Seier1
Summary
Japanese knotweed, Fallopia japonica (Houtt.) Ronse Decr., (Polygonaceae), is a serious environmental and economic weed in its adventive range of Europe and much of North America. Such is the
scale of the problem in the UK that a pioneering biocontrol programme began in 2003 which would
possibly make it the first target of a full classical biological control of weeds programme in Europe.
This paper summarizes the current status of the plant, reviews the literature associated with its natural
enemies and reports the progress with the programme for UK sponsors, as well as referring to North
American interests. We conclude that, should appropriate permissions be made available, the prospects for biological control of this high profile weed, using arthropod and fungal agents, are good.
Keywords: Japanese knotweed, classical biological control, Fallopia japonica.
Introduction
In December 2003, the European Strategy on Invasive
Alien Species (ESIAS) came into being (Genovesi and
Shine, 2004), supporting the Convention on Biological Diversity (CBD), calling for a regional approach
to the invasive alien species problem, and highlighting
the need for cost/benefit analyses of long-term control
measures. Any country intending to control those invasive alien species that threaten ecosystems, habitats or
species are encouraged by the Convention to consider
classical biological control for environmental weeds.
There have been over a thousand releases of biological control agents against weeds worldwide. Despite
European countries being the source for 381 releases
of classical biological control agents for alien plants
around the world (Julien and Griffiths, 1998), no full
classical weed biocontrol programme has yet been
carried out for the benefit of an EU country. The reasons for this are manifold and a source of frustration
for many weed biocontrol experts working in Europe,
particularly in light of the long list of potential targets
(Shaw, 2003; Sheppard et al., 2006). Fallopia japonica
CABI E-UK Centre, Bakeham Lane, Egham, Surrey TW20 9TY.
Kyushu University, Faculty of Agriculture, Fukuoka, 812-8581, Japan.
Corresponding author: D.H. <[email protected]>.
CAB International 2008
1
2
(Houtt.) Ronse Decr., Japanese knotweed (Polygonaceae), is one such target whose profile is so high that
many of the usual hurdles have been easier to overcome than for previous potential targets.
Nomenclature
Japanese knotweed was independently classified as
Reynoutria japonica by Houttuyn in 1777 and as Polygonum cuspidatum by Siebold in 1846. It was not
until the early part of the 20th century that these were
discovered to be the same plant (Bailey, 1990). Generally referred to as Polygonum cuspidatum by Japanese and American authors, recent evidence vindicates
Meissners 1856 classification as Fallopia japonica
var. japonica (Bailey, 1990). The closely related giant knotweed, Fallopia sachalinensis (F. Schmidt ex
Maxim.) Ronse Decraene can hybridise with F. japonica to form Fallopia x bohemica (Chrtek and Chrtkov)
J. Bailey, first described in 1983, and is rapidly proving
to be more difficult to manage than either of its parents
(Bmov et al., 2001; Mandk et al., 2004).
Common names include Japanese/Mexican bamboo,
pea-shooter plant, Sally/donkey/gypsy/wild rhubarb,
Hancocks curse, Japanese fleece-flower and horse
buckwheat. The Cornish name, Ladir Tir, is a rare example of the democratic addition to the lexicon since
the Cornwall Knotweed Forum, voted for this translation of their preferred English descriptor, land thief.
Itadori, the Japanese name for the plant, translates as
463
XII International Symposium on Biological Control of Weeds
take away pain, presumably reflecting its medicinal
properties. The subject of this paper will be referred to
as Japanese knotweed.
Reproduction
In its native range, the plant is functionally dioecious but in its introduced range it has spread solely
by vegetative means from a very small number of
initial introductions. Consequently, much of the invasive knotweed in the world may be clonal, as is the case
in the UK (Hollingsworth and Bailey, 2000). However,
recent research in the USA has shown that wild F. japonica can produce large quantities of viable seed and
seedlings have been found in the field (Forman and
Kesseli, 2003).
Morphology
Detailed descriptions of Japanese knotweeds morphology are available (Beerling et al., 1994; Lousely
and Kent, 1981). It is a vigorous, herbaceous perennial,
with annual, glabrous, tubular stems which ascend from
an often extensive rhizome system, to reach heights of
over 3 m in 3 months (Beerling et al., 1994).
Spread
The history of alien Polygonum and Reynoutria
species in the UK has been well reported (Bailey and
Conolly, 2000; Bailey, 2005; Conolly, 1977). The most
likely date of introduction of Japanese knotweed to
Europe is 1849, received at the nursery of Philipp von
Siebold in the Netherlands. This was also the first year
that the japonica variety was made available to the public as a much-prized ornamental. In the UK, the plant
had become naturalized by the late 1880s, having been
first recorded in the wild in Maesteg, South Wales, in
1886 (Conolly, 1977). Its status as a weed was soon
recognized, and today it is one of only two terrestrial
plants which are illegal to cause to grow in the wild
under the UK 1981 Wildlife and Countryside Act, as
well as being classed as a controlled waste, meaning
that a licence is required for its disposal.
Damage
The costs of Japanese knotweed can be considered
as both economic and environmental. To control Japanese knotweed on a national scale in the UK would
cost an estimated 1.56 billion, as noted by a review
team reporting to the UK Department of Environment,
Food and Rural Affairs in its recent non-native species
policy review (Defra UK, 2003). An accepted estimate
of control costs is 10,000 per hectare for a three-year
spraying regime with two sprays per year, although this
is probably an underestimate if revegetation costs are
taken into account. With fragments as small as 0.6g
capable of generating new plants, the presence of Japanese knotweed can add around 10% to the costs of a
development project, especially if soil is considered
contaminated and subject to removal fees. A worst-case
scenario could see a 1m2 patch costing up to 46,000 to
eradicate (M. Wade, 2006, personal communication).
Its reputation as a concrete-cracking super-weed is
justified; seven designs of reinforced channel revetment
blocks were specifically tested against penetration and
displacement by Japanese knotweed (Beerling, 1991),
and all seven failed. In East London, work has begun
to clear four hectares of knotweed infesting the 2012
Olympics site, an activity which is gleefully reported
by the press to have added 65 million to the expanding
development budget.
Though harder to quantify, the impact the weed has
on ecosystem function and biodiversity are considerable. Its early emergence and great height combine to
shade out other vegetation and prohibit regeneration of
other species (Sukopp and Sukopp, 1988). Dead knotweed stems can persist for two to three years producing large quantities of debris and slowly decomposing
litter, which also leads to low floristic diversity (Child
and Wade, 2000). Observations on knotweed in the UK
revealed that invertebrate species richness was lower
on F. japonica than on sympatrically occurring native
plant species (Beerling and Dawah, 1993). More recent
work in Switzerland, Germany and France, comparing
the diversity of plants and invertebrates in invaded and
non-invaded habitats, showed a reduced diversity on
both taxa, as well as a halving of invertebrate biomass
under knotweed (E. Gerber, unpublished data). Impacts
on fish and other vertebrates further up the food chain
are likely and knotweed-invaded sites appear to be
less suitable habitats for foraging frogs (Maerz et al.,
2005).
Dense knotweed stands can also exacerbate flooding, damage riverbank protection works and impede
flow, whilst dead stems can cause blockages downstream when swept away. Knotweeds influence on riparian systems is particularly pertinent in the light of
the EU Water Framework Directive, which demands
that member nations waterways achieve good status
by 2015.
Current control measures
The effectiveness of control and eradication interventions has recently been thoroughly reviewed by
Kabat et al. (2006), who included 65 articles in their
meta-analyses. Six control interventions were considered, none of which could eradicate Japanese knotweed or its hybrid in the short term. Cutting treatments
alone were not found to result in significant decreases
of knotweed abundance. However, they found that statistically significant reductions in abundance could be
achieved through limited application of: a) glyphosate,
464
Could Fallopia japonica be the first target for classical weed biocontrol in Europe?
imazapyr, or imazapyr plus glyphosate, b) cutting followed by filling stems with glyphosate, or c) cutting
followed by spraying with glyphosate. These authors
were unable to conclude any clear long-term efficacy.
As a general rule of thumb, based on discussions
with numerous experts in the UK and the United States,
a late-season application of glyphosate, when the plant
is at maximum height, is the most cost-effective control
measure.
Literature and field observations
Methods and materials
The phytophagous arthropods and fungi recorded
from Japanese knotweed were collated from the printed
and electronic literature in both the English and Japanese language and, for arthropods, their feeding habits were categorized. Surveys were carried out across
the growing season of the plant each year from 2003
to 2006. Early surveys covered the complete range of
the plant from Northern Honshu to Southern Kyushu
islands, whilst later collections focussed on Kyushu Island. The focus on Kyushu Island was a consequence
of molecular studies carried out at Leicester University,
that showed the closest match to the UK clone was to
be found in the Nagasaki Prefecture of Kyushu. At each
site, knotweed plants were first visually assessed for
natural enemies, their activity and/or damage inflicted,
before using a beating tray to collect any natural enemies that had been missed. A subset of the plants had
their stems and rhizomes split open to reveal any endophagous species. Simultaneous assessments were
carried out on other members of the Polygonaceae
family growing in the vicinity to provide data on field
host range.
Results
The literature review of natural enemies revealed
186 arthropod species and over 30 fungal plant pathogens to be associated with F. japonica in Japan. This
is in stark contrast with the situation in the UK where
only 14 arthropods and no fungal plant pathogens have
been recorded on the plant (Figure 1). In Japan, leaf
feeders and sap suckers together made up over 87% of
the arthropod species recorded (Figure 2). The dearth
of rhizome feeders was notable and this surprising observation was supported by subsequent field surveys,
which revealed this large resource to be almost solely
exploited by the polyphagous hepialid moth Endoclyta
excrescens (Butler).
Surveys revealed that knotweed was subject to
significantand in many cases, severenatural enemy
damage. In undisturbed areas, this led to it being outcompeted by the many large forbs characteristic of the
Japanese flora. Observations on sympatric Polygonaceae
revealed that a handful of these natural enemies had a
very narrow host range. It should be noted that at sites
where the natural enemy cycles had been disrupted by
cutting, Japanese knotweed revealed its potential as a
dominant species.
80
Number of species
70
60
50
UK
40
Japan
30
20
Pathogens
Orthoptera
Lepidoptera
Hymenoptera
Hemiptera
Coleoptera
Diptera
10
Taxon
Figure 1.
A comparison of phytophagous natural enemies recorded on Fallopia japonica
in Japan and the UK by order (fungal pathogens are presented collectively).
465
XII International Symposium on Biological Control of Weeds
Figure 2.
Arthropod natural enemies found on Fallopia japonica in Japan, grouped by
feeding habit.
Potential biological control agents
The status of the more interesting potential biocontrol
agents is presented below including a summary of any
host-range testing that has been undertaken. Despite
the sound arguments for shorter and more tailored hostrange test plant lists (Briese, 2005), the approach taken
here was the more traditional one including apparently
irrelevant species. This decision was taken because any
consideration of the use of a classical biological control
agent in the UK would be novel and the authorities
are likely to be most interested in economic and crop
plants. The list includes 74 species from 23 families,
consisting of 33 plants native to the UK, 15 introduced
species, three native to Europe, 13 ornamentals and ten
economically important crop species.
Arthropods
Ostrinia ovalipennis Ohno (Lepidoptera: Crambidae) is a recently identified (Ohno et al., 2003; Ohno,
2003) close relative of O. latipennis (Warren), a wellknown and widely distributed knotweed borer feeding
on other species in the field in Japan. Ostrinia ovalipennis appears to be univoltine and restricted to two
distinct populations; one from Hokkaido Island and the
other from highland areas in the Nagano Prefecture of
central Japan (Ohno et al., 2006). It has only been recorded from Japanese and giant knotweed. Identification and likely rearing difficulties meant that this potential agent was not prioritized for the UK but remains
of interest for North America where giant knotweed is
more of an issue.
Macchiatella itadori (Shinji) (Hemiptera: Aphididae) is a very common aphid which causes severe
damage to both F. japonica and F. sachalinensis from
June to September, often in association with leafspot
and various ant species that tend it. Unfortunately, its
primary winter hosts are recorded as Rhamnus japonica
Maxim and R. purshiana De Candolle and, as such, it
was dismissed for the UK due to the likely attack of native Rhamnus spp. It is also unlikely to be of interest to
North America, unless the latter host has been recorded
erroneously.
Ametastegia polygoni Takeuchi (Hymenoptera:
Tenthredinidae). This stem-mining sawfly has been
collected from both Japanese and giant knotweeds in
the field, although the Japanese literature only reports
Japanese knotweed as a host plant. Attempts to establish a culture for host-range testing have failed so far
but this sawfly has not been rejected.
Gallerucida bifasciata (Motchulsky) syn. Gallerucida nigromaculata (Baly) (Coleoptera: Chrysomelidae). Originally two species, these have recently been
synonymized to G. bifasciata. Some doubt remains,
however, since our collections revealed considerable
differences in both morphology and behaviour in populations from different regions of Japan. A more southerly
population was used for preliminary larval no-choice
testing and when presented with both cut plant and
live plant material, larvae fed significantly within the
Polygonaceae family. Subsequent field observations
in Kyushu revealed larvae feeding on Rumex acetosa
L. leaves and consequently, this beetle was not prioritised. A more northerly population has been collected
and is undergoing testing in the United States.
Lixus impressiventris Roelofs (Coleoptera: Curculionidae) is a very common stem-boring weevil
which has only ever been collected from Japanese and
giant knotweeds in the field even when suitably sized
stems of Rumex spp. were present, at a site where almost every internode section of the knotweed stems
contained a Lixus larva. Nonetheless, adult no-choice
and choice tests showed a physiological host range that
included many members of the Polygonaceae. Despite
a Japanese paper reporting the weevil as a pest of a
466
Could Fallopia japonica be the first target for classical weed biocontrol in Europe?
very minor crop, Polygonum tinctoria Lour. (Sekiguchi
and Wakiya, 1988), every attempt to rear the weevil on
this plant has failed. No-choice oviposition and development tests showed that one native UK plant, Polygonum hydropiper (L.), was able to support development
of the weevil, albeit producing significantly smaller
adults in the process. The possibility of adults feeding
on non-targets and the risk of development on a native
plant species have meant that this weevil is no longer
a prioritized agent for the UK. Further studies, perhaps
in the native range, may well prove this weevil to be
highly specific.
Aphalara itadori Shinji (Hemiptera: Psyllidae) is
found from southern Kyushu to as far north as Nagano
Prefecture on Honshu Island and was observed feeding on Japanese knotweed from sea level to 2150 m
a.s.l. Adults were collected from late April to midAugust and although widespread, were rarely present
in high numbers. One unidentified eulophid parasitoid
has been reared out from a late nymph. Adults lay eggs
on the leaves or under the papery sheath surrounding
the petiole and once hatched, the nymphs pass through
five instars feeding on the phloem of the plant. In the
laboratory, at 22oC, the mean development time was
32.9 days (0.8= SE, n=21) and reproductive females
laid a mean of 637 eggs each (121= SE, n=11). Impact studies are ongoing but early signs indicate that
the presence of feeding nymphs restricts plant height
and increases leaf production.
Host-range tests have focussed on multiple-choice
oviposition studies since the nymphs are not very mobile and adult feeding was hard to observe and quantify. Host-absent multiple-choice tests were used to
test the validity of host-present tests and no significant
difference was found when very closely related plants
were used. Over a 20-month period, the location and
fate of just under 125,000 eggs have been recorded
during tests on 83 test plant species. So far, only 700
eggs (0.6%) have been laid on non-knotweed or knotweed hybrid hosts and not one of these has developed
through to adult. Although more replication is required
on some non-target species, these results are extremely encouraging. The question of what happens when
above-ground knotweed dies off at the first frost remains. Adult Aphalara are presumed to shelter in the
bark of trees such as Cryptomeria spp. (N. Takahashi,
2007, personal communication). This is currently being investigated.
Pathogens
Puccinia polygoni-amphibii var. tovariae Arthur
(Basidiomycota: Pucciniaceae). Several strains of
this rust have been found in the field on F. japonica in
Honshu and Kyushu Islands, from sea level to 1550
m a.s.l. Collected all year round, either as cinnamon
brown-coloured uredinia or as the over-wintering,
brownish-black, telial stage, it was also recorded on
F. sachalinensis, F. japonica var. compacta and the
somewhat hairy-leaved F. japonica var. uzenensis. The
uredinial spore stage was tested against more than 40
non-target plants. It showed a great deal of potential by
infecting all stages of the target plant, causing severe defoliation in the lab. Unfortunately, infective symptoms
and subsequent sporulation to produce viable spores
were consistently recorded on the native Rumex longifolius DC, and Fallopia baldschuanica (Regel) Holub,
an ornamental. Furthermore, its life cycle could not
be resolved in quarantine since telial dormancy could
not be broken under these artificial conditions. These
facts, coupled with reports in the Japanese literature of
closely related Puccinia varieties being heteroecious,
with Geranium spp. as alternative hosts, meant that this
damaging rust was no longer prioritized for the UK.
Aecidium polygoni-cuspidati Dietel (Basidiomy
cota: Incertae sedis). This conspicuous rust was found
on F. japonica var. compacta and F. japonica var.
japonica in the field and has been recorded from F.
sachalinensis in the Japanese literature. It is found from
late April to September at altitudes of up to 1270 m,
but occurring more frequently in the lower, warmer areas of both Honshu and Kyushu islands, commonly in
humid riparian and woodland habitats. Failure to infect
knotweed plants with aeciospores in the lab reinforced
suspicions that this rust may in fact be heteroecious and
synonymous with Puccinia phragmitis (Schum.) Krn.,
using Phragmites communis Trin. as its alternative host.
This was confirmed in the Japanese literature (Harada,
1978) where it was identified that the rust had many specialized biologic forms or strains, one of which infected
F. japonica and F. sachalinensis in its aecial stage. This
agent has therefore been dismissed.
Mycosphaerella polygoni-cuspidati Hara (Ascomycota: Mycosphaerellaceae). This hemibiotrophic
pathogen which cycles through the sexual ascospores
and causes a highly damaging and ubiquitous leafspot is found on Kyushu, Honshu and Shikoku islands,
from sea level to altitudes over 900 m. Displaying a
high degree of polymorphism, the lesions may appear
as large, dark-tan coalescing forms, or as circular or
irregular, chestnut-brown lesions, or as more discrete
spotting. This, and considerable variation shown in
cultured isolates, suggests that there is a range of morphotypes and/or pathotypes. The leafspot appears to be
restricted to F. japonica var. japonica in the field, and
coincides with knotweed stem emergence in late April
through to its senescence in October/November. After
considerable investigation, this slow-growing ascomycete has now been confirmed as the causal agent,
following Kochs postulates. Host-range testing has
been carried out, using ascospore inoculum for closely
related species and mycelial inoculum for the entire
host-test list since the availability of the former is limited. To date, over 50 plant species have been tested,
and results confirm the extremely narrow host range
shown in the field. Genetic characterization of the
467
XII International Symposium on Biological Control of Weeds
various strains is being carried out to ascertain their
relatedness.
Discussion
The short answer to the question posed by the title
of this paper is no, this will not be the first target
for classical biocontrol of weeds in Europe. This is
not because an eventual agent release is unlikely but
rather because it would not actually be novel for Europe. Closer examination of a biological control study
against creeping thistle (Circium arvens) in the UK in
1969 (Baker et al., 1972) revealed that, although the
initial releases of hundreds of adult beetles (Haltica
carduorum Guerin) from France were made into field
cages, these cages were removed later in the study. The
eventual results were similar to those encountered in
Canada, with no successful survival over winter (Peschken et al., 1970).
Despite this, the completion of a full, official classical biological control programme for a weed in Europe
is effectively a new concept and would be expected to
face various challenges from the outset (Shaw, 2003;
Sheppard et al., 2006). A team at the University of Coimbra in Portugal is currently studying the safety and
efficiency of the gall wasp, Trichilogaster acaciaelongifoliae Froggatt, against Acacia longifolia (Andr.)
Willd. in quarantine. This agent was successfully released in South Africa (Julien and Griffiths, 1998). This
is part of a larger project, but it could be that this excellent agent will be the first classical agent released in
Europe against a weed.
Regulatory challenges are likely to be the most difficult to overcome especially when it comes to fungal
agents, although proposed arthropod releases for weeds
have not been welcomed as much as those for insect
pests. At this stage, the psyllid Aphalara itadori and
the leafspot Mycosphaerella sp. seem likely to pass the
host-range testing process, but whether the prospect of
an actual release into the environment becomes a reality
is likely to depend on individuals within the appropriate
UK government and EU department(s) taking a pragmatic approach to often inappropriate or absent legislation.
If the eventual goal of release is achieved, then this programme will indeed have laid the groundwork, helped
establish the rules and opened the door to classical biological control of weeds in Europe (Kurose et al., 2006).
Acknowledgements
Much of the work outlined in this paper would not have
been possible without the help of the Japanese knotweed team at Kyushu University, in particular Professor
Masami Takagi, as well as technical support in the UK
from Sarah Bryner, Valerie Coudrain and Lynn Hill. We
would like to thank Defra, the UK Environment Agency,
Network Rail, The Welsh and South West Regional
Development Agencies, British Waterways, Cornwall
County Council and the USDA Forest Service for their
funding, and the Royal Entomological Society and the
European Weed Research Society for the travel grants
that were used to attend this symposium.
References
Bailey, J. (2005) The history of Japanese knotweed. Ecos
British Association of Nature Conservationists 26, 5562.
Bailey, J.P. (1990) Breeding behaviour and seed production
in alien giant knotweed in the British Isles; Biology and
control of invasive plants. In: Richards, Moorehead &
Laing Ltd. (eds) Biology and Control of Invasive Plants.
BES Industrial Ecology Group Symposium, pp.110120.
Bailey, J.P. and Conolly, A.P. (2000) Prize-winners to
pariahsa history of Japanese knotweed s.l. (Polygonaceae) in the British Isles. Watsonia 23, 93110.
Baker, C.R.B., Blackman, R.L. and Claridge, M.F. (1972)
Studies on Haltica carduorum Guerin (Coleoptera: Chrysomelidae) an alien beetle released in Britain as acon
tribution to the biological control of creeping thistle,
Cirsium arvense (L.) Scop. Journal of Applied Ecology
9, 819830.
Bmov, K., Mandk, B. and Pyek, P. (2001) Experimental
control of Reynoutria congeners: a comparative study of a
hybrid and its parents. In: Brundu, G., Brock, J., Camarda,
I., Child L. and Wade, M. (eds) Plant Invasion: Species
Ecology and Ecosystem Management. Backuys, Leiden,
Netherlands, pp. 283290.
Beerling, D. (1991) The testing of cellular concrete revetment blocks resistant to growths of Reynoutria japonica
Houtt. (Japanese knotweed). Water Research (Oxford) 25,
495498.
Beerling, D.J. and Dawah, H.A. (1993) Abundance and diversity of invertebrates associated with Fallopia japonica
(Houtt. Ronse Decraene) and Impatiens glandulifera
(Royle): two alien plant species in the British Isles. Entomologist 112, 127139.
Beerling, D.J., Bailey, J.P. and Conolly, A.P. (1994) Fallopia
japonica (Houtt.); Ronse Decraene (Reynoutria japonica
Houtt.; Polygonum cuspidatum Sieb. & Zucc.), Journal of
Ecology 82, 959979.
Briese, D.T. (2005) Translating host-specificity test results
into the real world: the need to harmonize the yin and
yang of current testing procedures. Biological Control 35,
208214.
Child, L. and Wade, M. (2000) The Japanese Knotweed Manual: the Management and Control of an Invasive Alien
Weed. Packard Publishing Limited, Chichester, UK 123p.
Conolly, A.P. (1977) The distribution and history in the British Isles of some alien species of Polygonum and Reynoutria. Watsonia 11, 291311.
Defra, UK (2003) Review of Non-native Species Policy
Report of the Working Group www.defra.gov.uk/wildlifecountryside/resprog/findings/non-native/report.pdf
Forman, J. and Kesseli, R.V. (2003) Sexual reproduction in
the invasive species Fallopia japonica (Polygonaceae).
American Journal of Botany 90, 586592.
Genovesi, P. and Shine, C. (2004) European Strategy on Invasive Alien Species. Nature and Environment, n. 137,
t-pv (2003)7. Council of Europe Publishing, Strasbourg.
468
Could Fallopia japonica be the first target for classical weed biocontrol in Europe?
Harada, Y. (1978) New hosts and biologic specialization in
the aecial state of Puccinia phragmitis in Japan. Transactions of the Mycological Society of Japan 19, 433438.
Hollingsworth, M.L. and Bailey, J.P. (2000) Evidence for
massive clonal growth in the invasive weed Fallopia
japonica (Japanese Knotweed). Botanical Journal of the
Linnean Society 133, 463472.
Julien, M.H. and Griffiths, M.H. (1998) Biological Control
of Weeds. A World Catalogue of Agents and their Target
Weeds. Fourth edition. CABI Publishing, Wallingford,
UK. 223 p.
Kabat, T.J., Stewart, G.B. and Pullin, A.S. (2006) Are Japanese knotweed (Fallopia japonica) control and eradication interventions effective? In: Centre for Evidence
Based Conservation Internal Report, Systematic Review
no. 21. Birmingham, UK.
Kurose, D., Renals, T., Shaw, R., Furuya, N., Takagi, M.
and Evans, H. (2006) Fallopia japonica, an increasingly
intractable weed problem in the UK: Can fungi help cut
through this Gordian knot? Mycologist 20, 126129.
Lousely, J.E. and Kent, D.H. (1981) Docks and Knotweeds of
the British Isles. BSBI, London, UK. 205p.
Maerz, J.C., Blossey, B. and Nuzzo, V. (2005) Green frogs show
reduced foraging success in habitats invaded by Japanese
knotweed. Biodiversity and Conservation 14, 29012911.
Mandk, B., Pyek, P. and Bmov, K. (2004) History of the
invasion and distribution of Reynoutria taxa in the Czech
Republic: a hybrid spreading faster than its parents. Preslia 76, 1564.
Ohno, S. (2003) A new knotweed-boring species of the genus
Ostrinia Hbner (Lepidoptera: Crambidae) from Japan.
Entomological Science 6, 7783.
Ohno, S., Hoshizaki, S., Tatsuki, S. and Ishikawa, Y. (2003)
New records of Ostrinia ovalipennis (Lepidoptera: Crambidae) from Hokkaido, and morphometric analyses for
species identification and geographic variation. Applied
Entomology and Zoology 38, 529535.
Ohno, S., Ishikawa, Y., Tatsuki, S. and Hoshizaki, S. (2006)
Variation in mitochondrial COII gene sequences among
two species of Japanese knotweed-boring moths, Ostrinia
latipennis and O. ovalipennis (Lepidoptera: Crambidae).
Bulletin of Entomological Research 96, 243249.
Peschken, D., Friesen, H.A., Tonks, N.V. and Banham,
F.L. (1970) Releases of Altica carduorum (Chrysomelidae: Coleoptera) against the weed Canada thistle (Cirsium arvense) in Canada. Canadian Entomologist 102,
264271.
Sekiguchi, T. and Wakiya, H. (1988) Ecology and chemical
control of a Lixus impressiventris Roelofs infesting Polygonum tinctorium Lour. Tokushima Agricultural Experimental Station Reports 25, 5257.
Shaw, R.H. (2003) Biological control of invasive weeds in the
UK: opportunities and challenges. In: Child, L., Brock,
J.H., Brundu, G., Prach, K., Pyek, K., Wade, P.M. and
Williamson, M. (eds) Plant Invasions: Ecological Threats
and Management Solutions. Backhuys, Leiden Publishers, pp.337354.
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006) Top 20
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 125.
Sukopp, H. and Sukopp, U. (1988) Reynoutria japonica
Houtt. in Japan and in Europe. Veroffentlichungen Geobotanishes Institut, Zurich 98, 354372.
469
Biological control of Rumex species in
Europe: opportunities and constraints
P.E. Hatcher,1 L.O. Brandsaeter,2 G. Davies,3 A. Lscher,4
H.L. Hinz,5 R. Eschen5 and U. Schaffner5
Summary
The increasing problems caused by dock infestations (especially Rumex obtusifolius L., R. crispus L.,
and R. longifolius DC.) to organic agriculture in Great Britain, Norway and Switzerland are discussed.
Inadequate, costly, or time-consuming non-chemical control options for Rumex are among the major
barriers for farmers converting to organic production. Potential biological control agents for Rumex
in Europe are discussed. We conclude that the chrysomelid beetle Gastrophysa viridula Degeer and
the rust fungus Uromyces rumicis (Schum.) Wint. remain the most promising of the researched indigenous species and that G. viridula can be combined with other non-chemical control methods. However, there is a need for biological control agents that target dock roots; we suggest that Pyropteron
chrysidiformis (Esper), one of several sesiid moth species present in Europe which attack dock roots,
has good potential for Rumex spp. biological control and merits further study within Europe.
Keywords: dock, organic farming, Rumex crispus, Rumex longifolius, Rumex obtusifolius.
Introduction
Docks, especially Rumex obtusifolius L. and R. crispus
L., have been recognized as problem weeds in conventional agriculture for centuries (Foster, 1989; Zaller,
2004). These species grow rapidly, are resilient to cutting (being able to quickly regrow from their root stock,
and replenish carbohydrates used in regrowing within
two to three weeks), are long-lived and are able to produce up to 80,000 seeds per plant per year (Cavers and
Harper, 1964). These seeds form a long-lasting soil seed
bank with seeds surviving for possibly up to 80 years
(Cavers and Harper, 1964). More recently, docks have
been recognized as a serious problem for organic agriculture and an important limiting factor in the conversion from conventional to organic farming is thought to
be the worry of many farmers over their ability to control docks without chemical herbicides. In this paper,
School of Biological Sciences, The University of Reading, Reading,
UK.
2
Bioforsk, Norway.
3
HDRA, Ryton Organic Gardens, Coventry, UK.
4
Agroscope Reckenholz-Tnikon, Research Station ART, Zurich,
Switzerland.
5
CABI Europe-Delmont, Switzerland.
Corresponding author: P.E. Hatcher <[email protected]>.
CAB International 2008
1
we examine this problem as well as recent and ongoing research into it in three European countriesGreat
Britain, Switzerland and Norway. We discuss possible
biological control methods (none of which are currently used in Europe) and our recommendations for the
way ahead in Europe. Of course, this paper only gives a
snapshot of the situation in three countries and there is
much work also taking place in other European countriesfor example, Germany (Zaller, 2004), the Czech
Republic (Martinkov and Honk, 2004) and Austria
(Hann and Kromp, 2003). Also, as we do not intend to
review the voluminous work on non-chemical control
of Rumex spp., readers are referred to Foster (1989),
Hatcher and Melander (2003), Zaller (2004) and Bond
et al. (2006) for this.
The problems
Great Britain
In January 2005, a total of 690,269 ha of agricultural land in the UK was registered as organic or in
conversion to organic; just over 4% of all agricultural
land (FiBL, 2006). Of this, 92% was fully organic and
the retail market for organic products in the UK was
worth GBP 1.213 billion in 2006. The problem caused
by docks to UK organic producers became very apparent during the course of a three-year UK Government
470
Biological control of Rumex species in Europe: opportunities and constraints
(DEFRA)-funded research programme into the management of weeds in organic production systems, carried out by the Henry Doubleday Research Association
(HDRA) (Turner et al., 2004, 2007). Of those farmers
surveyed, 92% listed Rumex spp. as one of their main
problem weed species. Farmers had a clear understanding of what encouraged docks in their systems (e.g.
poaching, inappropriate or untimely cultivation) and
many thought that their dock problems were historical
and had persisted due to poor weed management prior
to organic conversion. The general approach to control
was to prevent dock seeding and to reduce their vigour
by harvesting crops before docks seeded, keeping margins clean, planning control periods into rotations and
the use of grazing stock. Direct action included various
integrated topping and grazing strategies, with many
farmers using topping machinery. Sheep were used to
intensively graze young seedlings, and goats to strip
mature plants. Manual removal was also used at times,
and periods of summer fallow were used to cultivate
the land, cutting roots below 10 cm and exposing the
roots on the surface to desiccate (to prevent regrowth).
Raking off these roots and burning them was effective.
Switzerland
Switzerland was one of the pioneers in organic
farming, and there were already 5001000 such farms
in the 1960s (Niggli, 2005). From the 1940s there was
a steady increase in conversion to organic farming and
since the 1990s this conversion has increased rapidly: in
1990 there were 803 organic farms totalling 10,000 ha;
by 2005 there were 6462 of 112,000 ha comprising
roughly 10% of farms and cultivated land in the country
(Niggli, 2005). These farms are typically small, with an
average size of 14 ha in 1998. Farmers in Switzerland
have identified R. obtusifolius and other dock species
(e.g. R. crispus, R. alpinus L.) as a major limitation to
plant production on existing organic farms and a serious obstacle to conversion to this type of production
(LBL Bericht, 2001). Organic farmers are typically
prepared to put considerable time into weed control
with some devoting over 1000 man-hours per year to
dock control alone (Grossrieder and Keary, 2004) but
this amount of effort is not feasible for all farms and is
obviously limited by economics. Large-scale physical
control using machines was not suitable as it causes
soil disturbance and this promotes dock seedling establishment. Variation in cultural control methods, such
as cutting height and frequency, grass species sown,
and added nutrients have all been found to have only
limited effects on dock populations. A review of these
grassland experiments (Lscher et al., 2001) demonstrated that all these management options did not significantly reduce the competitive ability of established
R. obtusifolius plants. Dock seedlings, however, have
a much weaker competitive ability than most sown
grasses. Consequently, all measures that increase grass
sward density and prevent gaps are successful against
the new establishment of dock plants, and root competi
tion was much more important than shoot competition
[this has also been shown for R. longifolius in Norway
(Haugland, 1993)]. Such cultural control methods must
be part of a holistic management strategy if the weed is
to be controlled (LBL Bericht, 2001; Niggli, 2005) and
should focus on preventing the establishment of new
dock plants (Dierauer et al., 2007). Due to the restrictions on control methods in organic farming, biological control is a logical tool to be integrated into such
a strategy.
Norway
Organic farming started in Norway in the 1930s, but
there were few such farms until the 1970s. A national
organic certification procedure was adopted in 1986,
with 19 farms being certified originally, and since then
the number of organic farms has increased steadily. In
1996 there were 946 farms of 7900 ha total (0.8% total agricultural land) (Johnsen and Mohr, 2000), while
by 2006 there were 2500 farms of 38,798 ha (3.8% of
agricultural land) (Debio, 2006). At 13 ha, the average
organic farm size is slightly larger than that of conventional farms, and almost all are run as family farms.
The current national plan of action aims for 15%
of agricultural land to be organically farmed by 2015
(www.regjeringen.no).
Over 80% of organic agricultural land in Norway is
under grassland, meadows or green manure, 15% under
cereals, with little organic horticulture (Debio, 2006).
Thus, grassland weeds are a major problem. A report
from Sweden (Andersson, 2005) states that many farmers feel powerless to control their Rumex problem, and
some organically motivated farmers are prevented
from converting to organic production because of this.
This applies also to Norway. Along with R. crispus and
R. obtusifolius, R. longifolius DC is also present in Norway. It is the most widespread weedy Rumex species
(Fykse, 1986) and is one of the most troublesome dicot species in Norwegian grasslands (Haugland, 1993).
R. longifolius can grow up to 1250 m above sea level;
it develops much faster in the spring and forms twice as
many shoots from root fragments but regrows slower
after defoliation than R. crispus and R. obtusifolius
(Fykse, 1986). A major Norwegian study has started to
investigate the natural enemies and control options for
R. longifolius and other Rumex spp. in Norwegian organic agriculture (Brandster and Haugland, 2007).
Biological control
Insects
Grossrieder and Keary (2004) have reviewed recent
studies on potential insect biological control agents for
Rumex spp. Much of the research reported here was
471
XII International Symposium on Biological Control of Weeds
carried out on European species of potential for Rumex
and Emex biocontrol in Australia, and has concentrated
on agents with a southern European, or Mediterranean
distribution, to match that of Australia. It is unlikely
that these species will be suitable for biological control
in the three central and northern European countries
considered here, ruling out species such as Lixus cribricollis Boheman (Col., Curculionidae) and Synansphecia doryliformis (Ochsenheimer) (Lep., Sesiidae).
Nor is it likely that introducing non-indigenous biological control agents will be feasible in Europe within
the near future, due to the current regulatory climate.
There is only one species in this category that might be
worth investigating at present: Gastrophysa atrocyanea
Motschulsky (Col., Chrysomelidae) from Japan, which
is probably the most promising insect for classical biological control of R. obtusifolius.
Within Europe, we need to concentrate on those insect species that can cope with the current management
regimes used against docks (e.g. cutting, ploughing,
removal before flowering) and this eliminates further
species. For example, several species of Apion weevils are found on Rumex. These mainly bore into the
flowering stem as larvae, and adults emerge in July to
August, after flowering. However, no organic farmer is
likely to leave a flowering Rumex in their fields if they
can help it, although this could be considered in uncultivated areas. Biocontrol agents should also be easy to
rear or culture for potential inundative releases.
Brachycaudus rumexicolens (Patch) (Hom., Aphididae) is recommended for further study by Grossrieder and Keary (2004). First discovered in the US in
the early 20th century, this species is confined mainly
to the Polygonaceae (Scott and Yeoh, 1998), especially
Rumex and Emex, although it can also attack Lupinus
albus L. and Triticum aestivum L. and it may also be
a virus vector. Nevertheless, the aphid was considered
sufficiently safe to be used in a programme aimed at the
biological control of Emex australis Steinh. in Australia
(Scott and Yeoh, 1998). This sap-feeding insect caused
widespread death and stunting of Emex, reducing individual achene weight by 41% (Scott and Shivas, 1998),
and has a high intrinsic rate of increase [rm alatae =
0.32, apterae = 0.43 at 24oC (Scott and Yeoh, 1999)]
even for an aphid. Scott and Yeoh (1999) carried out
temperature studies and bionomic modelling on B. rumexicolens and showed that it should be able to survive
in northern Europe. It has already been recorded from
most of Europe, including the UK and Norway (e.g.
Ossiannilsson, 1962).
Hyperia rumicis L., a leaf-feeding weevil (Col.,
Curculionidae), and Pegomya nigritarsis (Zetterstedt)
(Dipt., Anthomyiidae), a leaf miner, can both occasionally cause extensive damage to docks (Grossrieder and
Keary, 2004) but seem to have few advantages over
the leaf beetle Gastrophysa viridula Degeer (Col.,
Chrysomelidae).
Gastrophysa viridula is the most-studied dock insect. It has up to four generations a year in Europe,
overwintering as an adult in the surface layers of the
soil, and passing through a generation in six weeks in
favourable conditions throughout the spring to autumn.
The species can show a large population increase during the year; with females able to lay over 1000 eggs
each. Outbreaks of this insect have been reported,
stripping Rumex plants of leaf material. Although Martinkov and Honk (2004) report that it will feed upon
nine other plant families, it can only complete its lifecycle on Rumex spp. and prefers R. obtusifolius to other
docks (Bentley and Whittaker, 1979). Dispersal of the
beetle is limited; it has been rarely observed to fly and
tends to occur in discrete patches. Martinkov and
Honk (2004) suggest that the beetle has become more
widespread in central Europe during the 20th century,
with a recent expansion since 1950 with the spread of
weedy docks in lowlands during the formation of large
farms. G. viridula can cause up to 50% reduction in dry
weight of R. obtusifolius during the first year of growth
(Hatcher et al., 1997), up to 80% shoot and 65% root
reduction of R. crispus and R. obtusifolius first-year
overwintering plants (Hatcher, 1996), and can cause up
to 70% reduction of dry mass and 65% reduction in
seed production in the first four years of R. obtusifolius
growth in the field (Hatcher, unpublished data).
Several species of clearwing moth may be suitable
biological control agents. While Synasphecia doryliformis has a Mediterranean distribution, the closely
related Pyropteron chrysidiformis Esper (Lep., Sesiidae) is native throughout western Europe and southern
England, but has not been recorded from Scandinavia
(Spatenka et al., 1999). As in the case of S. doryliformis, the species is univoltine and the larvae feed in the
roots of various Rumex spp. (Spatenka et al., 1999).
Synasphecia doryliformis, which was mass-released
into Australia in the early to mid-1990s as a biological control agent against docks (Fogliani and Strickland, 2000), reduced dock densities there by up to 90%
within five years of release (Faithful, 2000). Scott and
Sagliocco (1991a, b) considered P. chrysidiformis to be
as effective a biological control agent as S. doryliformis,
but attempts to adjust its life cycle to southern hemisphere conditions failed and therefore their work on
P. chrysidiformis was discontinued. Preliminary studies have been initiated at CABI Europe-Switzerland,
aiming to further study the biology of P. chrysidiformis
and to develop rearing protocols. Mass-rearing and release methods have been developed in Australia for S.
doryliformis (Fisher, 1992), using pieces of dock root
for larval rearing, and gluing eggs to swizzle sticks by
machine and inserting them directly into cut flowering dock stems. We believe that these methods could
easily be adapted to rearing P. chrysidiformis. In the
UK, P. chrysidiformis occurs on R. crispus only in a
couple of sea-cliff and shingle beach sites in Kent, SE
472
Biological control of Rumex species in Europe: opportunities and constraints
England, and is protected by law as an endangered species under Schedule 5 of the Wildlife and Countryside
Act (1981). Thus, working with this species in the UK
will be especially challenging, but the combination of
insect conservation and weed biological control, if successful, would be particularly rewarding.
Two other sesiids, P. minianiforme Freyer and
S. triannuliformis Freyer coexist on Rumex spp., especially R. crispus in SE Europe, with the latter species
extending into north and central Europe (Grossrieder
and Keary, 2004). The biological control potential of
both should also be considered.
Fungi
Three species of pathogenic fungi commonly infect
weedy Rumex spp. throughout Europe, and have potential for their biological control.
The rust Uromyces rumicis (Schum.) Wint. is the
most studied fungus on Rumex spp., and was considered
in the 1960s as a potential biological control agent for
R. crispus in the USA (Inman, 1970). However, work
was discontinued when it was impossible to confirm the
alternate hosts of the fungus [in Europe the fungus is
almost entirely spread through uredospores and teleutospores, but rarely forms spermogonia and aecidia on
Ranunculus ficaria L. as an alternate host (Schubiger
et al., 1985)]. U. rumicis can cause up to 35% reduction in dry weight of R. obtusifolius during the first year
of growth (Hatcher et al., 1997), up to 60% shoot and
52% root reduction of R. crispus first-year overwintering plants (Hatcher, 1996), and can cause up to 40% reduction of dry mass in the first four years of R. obtusifolius growth in the field (Hatcher, unpublished data.).
U. rumicis damage is not normally apparent in the field
until late in the year, after dock has flowered, and thus
it usually has little effect on seed production. It also
cannot infect young developing dock leaves and as it is
non-systemic, the plant is able to outgrow fungal damage (Hatcher et al., 1995). However, while U. rumicis
is not promising as a sole biological control agent for
Rumex spp., it combines well with G. viridula. The rust
infects the older leaves, causing the beetles to move to
the younger leaves; thus an additive amount of damage
is consistently produced by combined beetle and rust
attack (Hatcher, 1996; Hatcher et al., 1997; Hatcher
and Paul, 2001). Artificial inoculation with the rust
early in the year is possible, and in cool, moist climates
is likely to persist over much of the summer. It is easy
to produce large numbers of uredospores for artificial
inoculation from R. obtusifolius plants in the laboratory
or glasshouse (Hatcher et al., 1994; Hatcher, 1996).
The necrotrophic fungus Ramularia rubella (Bon.)
Nannf. also shows promise as a dock biological control
agent. Unlike U. rumicis, this fungus can be cultured on
agar and thus might be bulked up in the laboratory but
there has been insufficient work to ascertain whether
this is likely to be easy or not. It also occurs early, grows
throughout the year on dock, and can cause almost total
defoliationincluding younger leavesduring severe
outbreaks (Hatcher, personal observation). This fungus
is very common on R. longifolius in Norway. HberMeinicke et al. (1989) found that infection by this fungus reduced shoot weight of R. obtusifolius by 58%
after 11 weeks, and root weight by up to 48%.
Venturia rumicis (Desm.) Wint. also occurs on
R. crispus and R. obtusifolius. This hemibiotrophic
ascomycete is the least common of these three pathogens in the UK and usually causes damage in late summer and autumn, although low levels can be present
throughout the year. It is unclear whether this fungus
can be cultured in vitro and no work has yet been attempted on using it as a biological control agent.
The way forward
Along with meeting all the usual criteria for biological control agents, agents selected for dock biological
control in organic agriculture must also be able to cope
with the cultural control methods already being practiced by organic growers until and unless they can be
demonstrated to be superior to these. Thus, the interactions among and between insect and fungus species become important, and also the effects of cutting, grazing
and other control methods on the population dynamics
of the insects and pathogens need to be studied in situ,
under a range of conditions. For example, P. chrysidiformis oviposits on dry dock stalks, so any attempt to
clean dock-infested pastures by cutting the dry stalks
may seriously hamper the population buildup of this
moth.
Recent reviews of potential dock biological control
agents (Grossrieder and Keary, 2004; Zaller, 2004;
Bond et al., 2006) suggest that of the indigenous species, Gastrophysa viridula and Uromyces rumicis still
show the greatest promise for inundative biological
control at the moment, but that Ramularia rubella and
Brachycaudus rumexicolens need further investigation
to determine their potential.
One way forward is to engage organic growers in
enhancing or conserving the herbivores and pathogens
they already have on Rumex. This has started in the UK
with the HDRA-managed project mentioned above. A
questionnaire about G. viridula received 34 replies:
23 respondents had seen the beetle, and 18 said that it
occurred every year and noted that it caused significant damage to the plant, although this could be patchy.
Ideas for encouraging the beetle in field margins at the
start of the growing season by covering plants with
fleece or simple polytunnels, for example, were put forward. It is easy to rear the beetle in bulk, either in plastic boxes on detached dock leaves (the beetle pupates
under several layers of absorbent paper at the bottom
of the box) or on plants grown in pots and sleeved with
473
XII International Symposium on Biological Control of Weeds
perforated plastic bags. Both methods have been used
for many years at the University of Reading, UK, and
can produce many thousands of gravid female beetles
(thought to be the best stage for release). This has enabled populations to be introduced into new areas of
southern England, which have established after introduction (Hatcher, personal observation). Such methods
could easily be adopted by farmers and thus inundative
and conservation biological control could be practised
with this insect.
As mentioned above, the beetle can be combined
with other biocontrol agents, and a combination of G.
viridula and U. rumicis with early re-sowing of Lolium perenne L. can control the flush of emerging R.
obtusifolius seedlings after a pasture seed-bed is prepared (Keary and Hatcher, 2004). It is possible that
regular cutting of dock-infested grassland may inhibit G. viridula, for example if it occurs during a peak
egg-laying period. However, natural populations of
the beetle are rarely synchronized and the beetle has
persisted throughout a ten-year experiment at the University of Reading, UK, in a field which is mown at
least four times a year without regard for the beetle
(Hatcher, personal observation). It is also possible to
modify mowing regimes to accommodate the beetle.
In Austria, Hann and Kromp (2003) found that beetlefriendly mowing (mowing twice per year rather than
the three times used in conventional management, and
each mow timed to coincide with the period the beetle was in the soil as a pupa) had a positive effect on
G. viridula density and feeding damage, compared to a
conventional regime of three cuts per year. The beetles
were able to spread over the site, and unmown sites had
greater numbers of overwintering adults than the mown
ones. Thus, unmown refuges at the edge of fields could
be useful for the beetle.
However, as we noted above, Rumex spp. are resilient to defoliation and have fast regrowth rates due to
tap-root reserves that are rapidly replenished after regrowth of leaves. Hence, successful control strategies
should include organisms that target other parts of the
plant than the foliage, in particular the below-ground
storage organs. Further research on native European insects feeding on dock roots, such as the clearwing moth
P. chrysidiformis, may be fruitful and could provide,
alone or in combination with defoliating organisms,
effective biological control options. Additional control
organisms and combinations of effective control agents
also need further investigation to determine their suitability for mass-rearing and their potential to reduce
dock populations in European organic agriculture.
References
Andersson, P.-A. (2005) Skrppa ett vxande problem
i ekologisk odling. Delrsredovisning fr 2005. http://
fou.sje.se/fou/default.lasso.
Bentley, S. and Whittaker, J.B. (1979) Effects of grazing by a
chrysomelid beetle, Gastrophysa viridula, on competition
between Rumex obtusifolius and Rumex crispus. Journal
of Ecology 67, 7990.
Bond, W., Davies, G. and Turner, R.J. (2006) The biology
and non-chemical control of broad-leaved dock (Rumex
obtusifolius L.) and curled dock (R. crispus L.). Available
at: http://www.gardenorganic.org.uk/organicweeds (accessed 14 April 2007).
Brandsaeter, L.O. and Haugland, E. (2007) Kontroll av hymole (Rumex spp.) i kologiske og konvensjonelle dyrkingssystem. Bioforsk FOKUS 2(7), 5558.
Cavers, P.B. and Harper, J.L. (1964) Biological flora of the
British Isles, Rumex obtusifolius L. and R. crispus L.
Journal of Ecology 52, 737766.
Debio (2006) Organic plant production in Norway, statistics
2006. http://debio.no (accessed 14 April 2007).
Dierauer, H., Hermle, M., Lscher, A., Schaller, A. and Thalmann, H. (2007) Blackenregulierung: Vorbeugende Massnahmen ausschpfen. Merkblatt, Forschungsinstitut fr
biologischen Landbau (FiBL) und Arbeitsgemeinschaft
zur Frderung des Futterbaues (AGFF), Binkert Druck,
Laufenberg, pp. 116.
FiBL (2006) Organic farming in the United Kingdom. Available at: http://www.organic-europe.net/country_reports/
great_britain/default.asp (accessed 14 April 2007).
Faithful, I. (2000) Distribution of dock moth in Victoria. Under Control 12, 910.
Fisher, K. (1992) Clearwing moths are key to dock control.
Western Australia Journal of Agriculture 33, 152155.
Fogliani, R.G. and Strickland, G.R. (2000) Biological Control of Dock: Enhanced Distribution of the Dock Moth.
Research Report, Meat Research Corporation, Sydney,
Australia.
Foster, L. (1989) The biology and non-chemical control of
dock species Rumex obtusifolius and R. crispus. Biological Agriculture and Horticulture 6, 1125.
Fykse, H. (1986) Experiments with Rumex species. Growth
and regeneration. Scientific Reports of the Agricultural
University of Norway 65 (25), 11 p.
Grossrieder, M. and Keary, I.P. (2004) The potential for the
biological control of Rumex obtusifolius and Rumex crispus using insects in organic farming, with particular reference to Switzerland. Biocontrol News and Information
25/3, 65N79N.
Hann, P. and Kromp, B. (2003) Der Ampferkfer (Gastrophysa viridula Deg.) Ein Pflanzenfresser als Ntzling
in der biologischen Grnlandwirtschaft. Entomologica
Austrica 8, 1013.
Hatcher, P.E. (1996) The effect of insectfungus interactions
on the autumn growth and over-wintering of Rumex crispus and R. obtusifolius seedlings. Journal of Ecology 84,
101109.
Hatcher, P.E., Ayres, P.G. and Paul, N.D. (1995) The effect
of natural and simulated insect herbivory, and leaf age,
on the process of infection of Rumex crispus L. and
R. obtusifolius L. by Uromyces rumicis (Schum.) Wint.
New Phytologist 130, 239249.
Hatcher, P.E. and Melander, B. (2003) Combining physical,
cultural and biological methods: prospects for integrated
non-chemical weed management strategies. Weed Research 43, 303322.
474
Biological control of Rumex species in Europe: opportunities and constraints
Hatcher, P.E. and Paul, N.D. (2001) Plant pathogenherbivore
interactions and their effects on weeds. In: Jeger, M.J. and
Spence, N.J. (eds) Biotic Interactions in PlantPathogen Associations. CABI Publishing, Wallingford, UK,
pp. 193225.
Hatcher, P.E., Paul, N.D., Ayres, P.G. and Whittaker, J.B.
(1994) Interactions between Rumex spp., herbivores and a
rust fungus: Gastrophysa viridula grazing reduces subsequent infection by Uromyces rumicis. Functional Ecology
8, 265272.
Hatcher, P.E., Paul, N.D., Ayres, P.G. and Whittaker, J.B.
(1997) Added soil nitrogen does not allow Rumex obtusifolius to escape the effects of insectfungus interactions.
Journal of Applied Ecology 34, 88100.
Haugland, E. (1993) Competition between an established
grass sward and seedlings of Rumex longifolius DC. and
Taraxacum officinale (Web.) Marss. Norwegian Journal
of Agricultural Sciences 7, 409420.
Hber-Meinicke, G., Dfago, G. and Sedlar, L. (1989) Ramularia rubella (Bon.) Nannf. as a potential mycoherbicide
against Rumex weeds. Botanica Helvetica 99, 8189.
Inman, R.E. (1970) Observations on the biology of Rumex
rust Uromyces rumicis (Schum.) Wint. Botanical Gazette
131, 234241.
Johnsen, K.K. and Mohr, E. (2000) Organic agriculture in
Norway. Available at: http://www.organic-europe.net/
country_reports/norway/default.asp (accessed 13 April
2007).
Keary, I.P. and Hatcher, P.E. (2004) Combining competition
from Lolium perenne and an insectfungus combination
to control Rumex obtusifolius seedlings. Weed Research
44, 3341.
LBL Bericht (2001) Forschungsttigkeiten des Bundesamtes
fr Landwirtschaft fr den Biologischen Landbau. March
2001.
Lscher, A., Nsberger, J., Jeangros, B. and Niggli, U. (2001).
Jugendentwicklung und Konkurrenzverhalten von Rumex obtusifolius L.. In: Kurzfassungen der Referate und
Poster, 45. Jahrestagung Arbeitsgemeinschaft fr Grnland und Futterbau in der Gesellschaft fr Pflanzenbauwissenschaften, Wissenschaftlicher Fachverlag, Giessen,
pp. 4546.
Martinkov, Z. and Honk, A. (2004) Gastrophysa viridula
(Coleoptera: Chrysomelidae) and biocontrol of Rumex
a review. Plant, Soil and Environment 50, 19.
Niggli, U. (2005) Organic farming in Switzerland 2005.
Available at: http://www.organic-europe.net/country_
reports/switzerland/default.asp (accessed 13 April 2007).
Ossiannilsson, F. (1962) Hemipterfynd i Norge 1960. Norsk
Entomologisk Tidskrift 12, 5662.
Schubiger, F.X., Dfago, G., Sedlar, L. and Kern, H. (1985)
Host range of the haplontic phase of Uromyces rumicis. In:
Delfosse, E.S. (ed.) Proceedings of the VI th International
Symposium on the Biological Control of Weeds. Agriculture Canada, Ottawa, Ontario, Canada, pp. 653659.
Scott, J.K. and Sagliocco, J.-L. (1991a) Host specificity of
a root borer, Bembecia chrysidiformis [Lep.: Sesiidae], a
potential control agent for Rumex spp. [Polygonaceae] in
Australia. Entomophaga 36, 235244.
Scott, J.K. and Sagliocco, J.-L. (1991b) Chamaesphecia doryliformis [Lep.: Sesiidae], a second root borer for the
control of Rumex spp. [Polygonaceae] in Australia. Entomophaga 36, 245251.
Scott, J.K. and Shivas, R.G. (1998) Impact of insects and
fungi on doublegee (Emex australis) in the Western Australian wheatbelt. Australian Journal of Agricultural Research 49, 767773.
Scott, J.K. and Yeoh, P.B. (1998) Host range of Brachycaudus
rumexicolens (Patch), an aphid associated with the Polygonaceae. Biological Control 13, 135142.
Scott, J.K. and Yeoh, P.B. (1999) Bionomics and the predicted distribution of the aphid Brachycaudus rumexicolens
(Hemiptera: Aphididae). Bulletin of Entomological Research 89, 97106.
Spatenka, K., Gorbunov, O., Lastuvka, Z., Tosevski, I. and
Arita, Y. (1999) Handbook of Palearctic Macrolepidoptera: SesiidaeClearwing Moths. Gem Publishing, Wallingford, UK.
Turner, R.J., Bond, W. and Davies, G. (2004) Dock management: a review of science and farmer approaches. In:
Hopkins, A. (ed.) Organic Farming, Science and Practice
for Profitable Livestock and Cropping, BGS Occasional
Symposium 37, British Grassland Society, Reading, UK,
pp. 5356.
Turner, R.J., Davies, G., Moore, H., Grundy, A.C. and Mead,
A. (2007) Organic weed management: a review of the current UK farmer perspective. Crop Protection 26, 377382.
Zaller, J.G. (2004) Ecology and non-chemical control of
Rumex crispus and R. obtusifolius (Polygonaceae): a review. Weed Research 44, 414432.
475
Opportunities for classical biological
control of weeds in European
overseas territories
T. Le Bourgeois,1* V. Blanfort,2 S. Baret,3 C. Lavergne,3
Y. Soubeyran4 and J.Y. Meyer 5
Summary
European overseas territories are home to biodiversity and endemism of worldwide importance,
vastly superior to that of continental Europe as a whole. They are, however, much more threatened
by invasive species, including hundreds of alien invasive plant species having a huge impact on
natural and agricultural habitats. As in continental Europe, invasive plants have only recently been
recognized as a threat to the local environment and biodiversity. Mechanical and chemical control
programmesunderway for several decadeshave not been entirely successful for permanent, costeffective, environment-friendly management. Biological control of weeds has long been successfully
used in other neighbouring countries with similar climates, environmental conditions and invasions,
but has barely been implemented in European overseas territories. There have been very few attempts
to set up classical biological control programmes in these regionsa few of the species that have
been the focus of biological control are Lantana camara L., Rubus alceifolius Poir., Opuntia stricta
(Haw.) Haw., Acanthocereus tetragonus (L.) Britton & Rose, Ligustrum robustum (Roxb.) Blume,
Miconia calvescens DC., Ulex europaeus L., Prosopis juliflora (SW.) DC., and Leucaena leucocephala (Lam.) de Wit. Many invasive plants occurring in European overseas territories are also invasive
elsewhere and already targets of biological control programmes. Biological control agent specificity
requires particular attention due to the high level of endemism in such islands. This paper reviews
some of the most threatening species for which classical biological control could be achieved through
regional or international collaboration.
Keywords: tropical island, invasive plants, biological control agent.
Introduction
It is well known that invasive alien species are considered to be one of the greatest threats to biodiversity
after habitat degradation, particularly in island ecosystems. European overseas territories consist of seven
Cirad-UMR PVBMT, Ple de Protection des Plantes, Route ligne Paradis, 97410 Saint-Pierre, La Runion.
2
IAC-Cirad, Centre de recherche Nord, BP 6, 98825 Pouembout, Nou
velle Caldonie.
3
Conservatoire Botanique National de Mascarin, 2 rue Pre Georges,
97436 Colimaons Saint-Leu, La Runion.
4
UICN France, Cirad, Ple de Protection des Plantes, Route ligne Paradis, 97410 Saint-Pierre, La Runion.
5
Dlgation la recherche, Gouvernement de Polynsie franaise, B.P.
20981, 98712 Papeete, Tahiti, Polynsie franaise.
Corresponding author: T. Le Bourgeois <thomas.le_bourgeois@
cirad.fr>.
CAB International 2008
1
Ultra-Peripheral Regions (UPRs) that are an integral
part of the European Union and 21 Overseas Countries
and Territories (OCTs) that benefit from a system of
close association (Table 1). Hereafter these two groups
are jointly referred to as European Overseas Regions
and Territories (EORTs). These EORTs are home to
biodiversity of worldwide importance and vastly superior to that of continental Europe as a whole. Three
French UPRs and 13 OCTs are involved in four of
the 34 world biodiversity hotspots (ConservationInternational, 2006; Mittermeier et al., 2005). R.A.
Mittermeier, President of Conservation International,
stated that the most remarkable places on Earth are
also the most threatened, and it is in these territories
that the speed of species extinction is the fastest worldwide. These territories have also hosted many species
introductionsmainly plants, some of which have become invasive. For instance, over the last 300 years,
476
Opportunities for classical biological control of weeds in European overseas territories
Table 1.
European Overseas Regions and Territories selected according to their climate.a
European Overseas Regions and Territories
Country
European status
Climate
Azores
Canaries
Guadeloupe
French Guiana
Madeira
Martinique
Runion
Anguilla
Aruba
BAT (British Atlantic Territories)
Bermuda
BIOT (British Indian Ocean Territories)
British Antarctic
BVI (British Virgin Islands)
Cayman
Greenland
Mayotte
Montserrat
Nederland Antilles
New Caledonia
Pitcairn
French Polynesia
Saint Pierre et Miquelon
St Helena (+ Ascencion, Tristan da Cua)
TAAF (Terres Australes et Antarctiques Franaises)
Turks & Cacos
Wallis and Futuna
Portugal
Spain
France
France
Portugal
France
France
United Kingdom
Nederland
United Kingdom
United Kingdom
United Kingdom
United Kingdom
United Kingdom
United Kingdom
Danmark
France
United Kingdom
Netherlands
France
United Kingdom
France
France
United Kingdom
France
United Kingdom
France
UPR
UPR
UPR
UPR
UPR
UPR
UPR
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
OCT
warm temp./subtrop.
warm temp./subtrop.
tropical
tropical
warm temp./subtrop.
tropical
tropical/temperate
tropical
tropical
temperate
tropical
tropical
polar
tropical
tropical
polar
tropical
tropical
tropical
tropical
tropical
tropical
polar/temperate
temperate/tropical
polar/temperate
tropical
tropical
European Overseas Regions and Territories shaded in grey were not considered in the study.
2217 plant species have been introduced on Runion
Island, 628 have become naturalized, and 62 were considered as invasive in the 1990s (Gargominy, 2003;
Macdonald et al., 1991). There are currently around
200 invasive plant species. For all the French overseas
territories, Gargominy (2003) highlighted the negative
role of invasive species with respect to biodiversity
conservation. Weed control in EORTs is essentially
mechanical and/or chemical (Hivert, 2003) and never
succeeds in long-term regulation of populations (Brondeau and Triolo, 2007). Eradication appears to be an
efficient way (technically and economically) to control
aliens on islands but requires early invader detection
and rapid political decision-making before the plant
has time to spread throughout a large area (Loope et
al., 2006). Only a few biological control programmes
have been implemented in the EORTs, all of which
were local programmes without any between-EORT
collaboration. In this paper, we analyse exotic flora of
EORTs to identify species common to several EORTs.
We selected five species among those present in more
than five EORTs that are under efficient classical biological control in other parts of the world. Here we
present classical biological control programmes that
could be implemented as European collaborative ac-
tions between EORTs and international collaborations
with other countries that have already successfully directed such control programmes.
Methods and materials
EORT climates range from polar to tropical, according to their geographical location. We selected EORTs
with warm temperate, subtropical and tropical climates
for this analysis. The degree of EORT invasion by alien
plants was analysed on the basis of literature data and
personal knowledge of certain situations (e.g. Runion,
New Caledonia, French Polynesia). A list of alien invasive species in EORTs was compiled from several
databases, literature and ongoing synthesis projects in
UK overseas territories (Varnham, 2005), the Canaries
(Sanz-Elorza et al., 2005), Madeira (Medeiros, 2006),
Azores (Silva, pers. comm.) (Silva and Smith, 2004),
Antilles (Joseph, 2006), French Polynesia (Meyer,
2000, 2004), New Caledonia (de Garine-Wichatitsky
et al., 2004; Meyer et al., 2006), the Caribbean region
(Kairo et al., 2001) and the IUCN database of invasive
species in French overseas territories, (Soubeyran, unpublished data). A species/EORT matrix was built. The
nomenclature of plant species was verified according
477
XII International Symposium on Biological Control of Weeds
to the Global Compendium of Weeds (Randall, 2002).
We analysed the number of species mentioned in several EORTs. Species present in five or more EORTs
were selected. We compiled plant biological control research or action programmes implemented in EORTs,
and species that are already under biological control in
other countries (Julien and Griffiths, 1998). We considered the possibility of developing a biological control
programme through collaborations between EORTs for
each invasive species.
Table 2.
EORT
F French Guiana
NL Aruba
NL Netherland Antilles
UK Turk & Caicos
P Madeira
P Azores
UK BVI
F Guadeloupe
F Martinique
UK Pitcairn
UK Montserrat
UK Tristan da Cuna
F New Caledonia
F Wallis Futuna
UK Cayman
F French Polynesia
UK Ascension
ES Canaries
F Runion
F Mayotte
UK Anguilla
UK BIOT
UK St Helena
UK Bermuda
Results
Plant invasions in EORTs
From seven UPRs and 21 OCTs, we selected
22 EORTs with warm temperate, subtropical, or tropical climates (Table 1). Saint Helena, Ascension and
Tristan da Cuna were considered as three different entities, which means we included 24 different sites in
this study. A list of 1267 plant species was compiled
from invasive plant lists for the different EORTs. The
number of plants per site ranges from three for French
Guiana and Aruba to 410 for Bermuda (Table 2). There
are two explanations for this variation. The first explanation concerns the origin of the information. In
some lists, only environmental weeds are considered
to be the most important invasive species, while both
environmental and agricultural weeds are taken into
account in other lists. The second explanation is that
EORT invasion patterns differ markedly between sites.
For instance, Joseph (2006) recorded very few invasive plants (22) in Martinique compared to Runion
(178). It is also well known that continental sites such
as French Guiana are less invaded than oceanic islands. We found 75 species that invaded at least five
sites (Table 3). Leucaena leucocephala (Lam.) de Wit
(recorded at 21 different sites) appears to be the most
common and best-distributed species. Five other species are present at 10 sites at least (Lantana camara L.,
Psidium guajava L., Albizia lebbeck (L.) Benth., Casuarina equisetifolia L., Ricinus communis L.). There
are about 851 and 205 species present at only one or
two sites, respectively. Most of them are common
weeds present in other EORTs, but are not considered
as invaders or environmental threats and are thus not
listed. However, some of them, even though they are
only considered to be invasive at one site, are highly
invasive, e.g. Hiptage benghalensis (L.) Kurz, which
seriously threatens local vegetation in dry habitats of
Runion, and the small tree Miconia calvescens DC. in
the French Polynesian rainforest.
Biological control programmes in EORTs
Only a few classical biological control research or
action programmes of have been undertaken despite the
extent of the invasive plant problems in most EORTs.
umber of alien, invasive weeds per European
N
Overseas Regions and Territories (EORT).
Number of weeds
3
3
7
8
10
12
15
18
22
26
28
49
67
61
74
96
101
151
178
190
196
230
288
410
The first one was launched in the early 1900s, with the
introduction and release of Ophiomyia lantanae (Froggatt) for L. camara control in French Polynesia (1916)
and New Caledonia (1924). Then four other agents (Teleonemia scrupulosa Stal, Syngamia haemorrhoidalis Guen., Octotoma scabripennis Gurin-Mneville,
Uroplata girardi Pic.) were released on this island over
the next 50 years, with varying degrees of efficacy
against L. camara (Gutierrez, 1976, 1979). This plant
has also been biologically controlled in other places
(Saint Helena, Ascension) (Julien and Griffiths, 1998).
Finally, only seven EORTs have developed a biological control programme (New Caledonia, French Polynesia, Saint Helena, Ascension, Runion, Montserrat
and Cayman) and only nine plant species have been
considered for biological control research programmes
or release, including: L. camara (see above); Opuntia stricta (Haw.) Haw. (New Caledonia, Cayman),
O. triacanthos (Willd.) Sweet (Montserrat) and Opuntia sp. (New Caledonia, Saint Helena, Ascension), using Cactoblastis cactorum (Berg) with good success;
Acanthocereus pentagonus (L.) Britton & Rose (New
Caledonia), using Hypogeococcus festerianus (Lizar &
Trelles); Miconia calvescens DC. (French Polynesia),
using Colletotrichum gloeosporioides L. f. sp. miconiae; Rubus alceifolius Poir. (Runion), using Cibdela
janthina (Klug); Ulex europaeus L. (Saint Helena), us-
478
Opportunities for classical biological control of weeds in European overseas territories
Table 3.
List of weed species considered invasive at five
sites at least.
Species
Leucaena leucocephala
Lantana camara
Psidium guajava
Albizia lebbeck
Casuarina equisetifolia
Ricinus communis
Acacia farnesiana
Argemone mexicana
Bryophyllum pinnatum
Melaleuca quinquenervia
Panicum maximum
Schinus terebinthifolius
Eicchornia crassipes
Cynodon dactylon
Commelina diffusa
Mirabilis jalapa
Solanum mauritianum
Tabebuia heterophylla
Tecoma stans
Pinus caribaea
Catharanthus roseus
Furcraea foetida
Melia azedarach
Canna indica
Syzygium jambos
Achyranthes aspera
Agave americana
Ageratum conyzoides
Antigonon leptopus
Bidens pilosa
Chamaesyce hirta
Cyperus rotundus
Grevillea robusta
Oxalis corniculata
Passiflora suberosa
Pennisetum purpureum
Pittosporum undulatum
Prosopis juliflora
Solanum nigrum
Spathodea campanulata
Terminalia catappa
Urochloa mutica
Ziziphus mauritiana
Adenanthera pavonina
Agave sisalana
Asclepias curassavica
Bambusa vulgaris
Carpobrotus edulis
Cenchrus echinatus
Clidemia hirta
Conyza bonariensis
Cryptostegia grandiflora
Eleusine indica
Eriobotrya japonica
Table 3.
Species
Total
21
13
12
11
11
10
9
9
9
9
9
9
9
8
8
8
8
8
8
8
7
7
7
7
7
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
5
5
5
5
5
5
5
5
5
5
5
(continued)
Leucaena diversifolia
Manilkara zapota
Melinis repens
Mimosa pudica
Momordica charantia
Opuntia ficus-indica
Paspalum conjugatum
Passiflora foetida
Phoenix dactylifera
Physalis peruviana
Plantago major
Psidium cattleianum
Rubus rosifolius
Senna occidentalis
Sida acuta
Sorghum halepense
Sphagneticola trilobata
Sporobolus indicus
Stachytarpheta urticifolia
Tamarindus indica
Ulex europaeus
Total
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
ing Tetranychus lintearius Dufour; Prosopis juliflora
(Sw.) DC. (Ascension), using Heteropsylla reducta
Caldwell & Martorell (and Rhinochloa sp. accidentally introduced); and Ligustrum robustum (Roxb.)
Blume subsp. walkeri (Decne.) P.S.Green (Runion),
for which Epiplema albida (Cassino & Swett) has
been tested but not yet released (CABI c.p., Julien
and Griffiths, 1998; Meyer, 1998). The situation concerning L. leucocephala is interesting. This plant is
considered as invasive almost everywhere it occurs
and is the most widely distributed species throughout
all EORTs. From 1985 to 1991, Heteropsylla cubana
Crawford, a biological control agent, arrived naturally
or accidentally in French Polynesia, New Caledonia,
and later in Runion and subsequently controlled this
invasive plant. Because of a conflict of interest regarding this invasive species, which is also a forage plant,
it was decided to biologically control H. cubana using
the lady bird beetle Olla v-nigrum (Mulsant) (Chazeau
et al.,1989; Quilici et al., 1995). Some other biological
control actions were also accidental, e.g. Rhinochloa
sp. against P. juliflora in Ascension. For others, such
as U. europaeus in Saint Helena, the biological control
agent T. lintearius was introduced along with its predator Phytoseiulus sp., thus nullifying the biological control. Most biological agents released were arthropods.
The only pathogen was C. gloeosporioides f.sp. miconiae for control of M. calvescens in French Polynesia
(Meyer and Killgore, 2000). This review highlights the
very low number of biological control actions undertaken in EORTs despite the fact that invasive plants are
479
XII International Symposium on Biological Control of Weeds
highly numerous and damaging to the environment and
biodiversity. Nevertheless, many of these species are
already targets of biological control actions or research
in other parts of the world.
Biological control programmes that could
be implemented in different EORTs
Cochereau (1972), proposed several strategies for
classical weed biological control programmes in the
Pacific, including targets such as Psidium guajava L.,
Melaleuca quinquenervia (Cav.) T.Blake, Elephantopus mollis Kunth, Stachytarpheta jamaicensis (L.)
Vahl, Mimosa invisa C.Mart. ex Colla , L. leucocephala, Solanum torvum Sw., Cyperus rotundus L., Rubus
rosifolius Sm., and Ageratum conyzoides L. Developing a new biological research programme (without any
knowledge of the target plant or its natural enemies) is
a very long process (more than 10 years) and very expensive with regard to a typical EORT budget, whereas
transferring biological control technology from countries where programmes are already underway is much
more time- and cost-effective. As many invaders are
common to several EORTs, joint biological control programmes could easily be implemented at a Europeanoverseas level. If EORTs decide to work together to
solve the problem of alien plants, species should be selected that are common to several sites. We have noted
that 78 plant species are invasive at five or more sites.
It is clearly not possible to implement so many biological control programmes and most of these species are
not yet biologically controlled elsewhere in the world.
To illustrate opportunities for developing classical
biological control actions in EORTs, we selected five
species according to four criteria: (1) historical success
of biological control of this target in other countries
with ecological similarities, (2) taxonomic isolation of
these weeds from indigenous flora in EORTs, (3) good
knowledge of biological control agents that are suitable
for use in EORTs, and (4) species that are not sources
of any conflicts of interest, such as Schinus terebenthifolius Raddi for honey or spice production, Psidium
catleianum Sabine and P. guajava for fruit production,
or Acacia spp. for wood production. The authors understand that this selection cannot be considered a priority for every EORT, as each one has its own priorities
in controlling invaders and/or biodiversity conservation. Nevertheless, the common feature of the following five examples is that they could be implemented
easily, rapidly, with a high probability of success, and
at low cost.
Case 1: Eichhornia crassipes (Mart.) Solms (Pont
ederiaceae): Nine sites are affected (Canaries, Guadeloupe, Martinique, French Polynesia, New Caledonia,
Runion, Bermuda, BVI, Cayman). Water hyacinth is
widely recognized as the worlds worst aquatic weed.
Native to the Amazon basin, it was exported throughout the tropics and warm temperate regions for its
flower and for water treatment. It forms dense mats
on water bodies, thus limiting access to water, navigation, and fishing. It produces H2S in the water, reduces
the water pH, increases evaporation, and reduces light
penetration and oxygen content. This leads to dramatic
biological changes, with social and economic consequences. Physical and chemical controls are very expensive, temporary, and ecologically and economically
unsustainable. Classical biological control is the only
feasible way to manage such widespread infestations.
A number of biological control agents have now been
introduced in about 30 countries. The species most
widely used are Neochetina weevils, N. bruchi Hustache and N. eichhorniae Warner (Coleoptera, Curculionidae) (Julien et al., 1999). With a 30-year history,
the biologies, host ranges, rearing, release and monitoring techniques are well documented (Julien et al.,
1999), and the efficiency is fully recognized in many
countries. Other agents are also used, such as the two
moths, Niphograpta albigutalis Warren and Xubida infusellus (Walker) (Lepidoptera, Pyralidae), which have
been released in 13 and three countries, respectively
(Julien et al., 2001). The weevils are currently reared in
South Africa at PPRI and can be considered as the most
suitable agents to initially release on tropical islands,
with an expected high success rate within two to seven
years (Le Bourgeois and Lebreton, 2006).
Case 2: Ulex europaeus L. (Fabaceae): Five sites are
affected (Canaries, Runion, Azores, Ascension, Saint
Helena). Native to the Western coast of Europe (UK,
France, Portugal), gorse is a prickly, perennial, evergreen legume which grows up to 3 m in height. It reproduces mainly by seed and is spread by machinery,
soil movements, water and animals. It is a major weed
problem in pastures and natural habitats, increasing the
risk of brushfires, reducing land utilization by forming
dense thickets, dramatically reducing stocking rates
and competing with native species of subalpine shrublands. It is considered as a weed of national significance in Australia, New Zealand and Hawaii. Several
biological control agents have already been used for
gorse. The gorse seed weevil Exapion ulicis Forst. (Coleoptera, Curculionidae) was introduced into Australia
in 1939 after being released in New Zealand. Its impact
is limited because the larvae are not present during the
second period of seed production. In 1998, the gorse
spider mite T. lintearius was released in Australia and
New Zealand. It forms colonies on plants and spin a
tent-like white web and feed on the leaves and branches. This spider mite may have a substantial impact but
is regulated by other mites such as phytoseids (Acari,
Phytoseidae). Other agents are under study, including
the gorse thrips Sericothrips staphylinus Haliday. In
Monserrat, the introduction of T. lintearius had no impact on gorse populations of the island, likely due to
the concomitant, accidental introduction of its predator,
Phytoseiulus sp. Pure populations of such biological
control agents must be introduced from the beginning
480
Opportunities for classical biological control of weeds in European overseas territories
and studies should be conducted to determine if indigenous phytoseids already exist in the area of introduction (Anonymous, 2003; Davies et al., 2004; Krause et
al., 1988).
Case 3: Clidemia hirta (L.) D.Don (Melastomata
ceae): Five sites are affected (Canaries, Mayotte,
Runion, Wallis and Futuna, Ascension). Kosters curse
is native to tropical America (Mexico and the West Indies, and southward to central Brazil). This noxious
weedy shrub grows up to 2 m tall in pastures and open
forests. It is an aggressive invader which shades out
all underlying vegetation. The seeds are principally
dispersed by frugivorous birds but any organism moving through the thickets will carry seeds away with it.
It is probably not resistant to fire, which is unlikely in
its habitat, but it rapidly colonizes burned areas. Introduced in Runion during the 1970s, it now colonizes
the wet forest understorey on the southeast coast and
roadsides and agricultural fields on the east coast. Several expeditions to find potential biological control
agents have been carried out in Trinidad, and a number
of insects were collected and screened. In Hawaii, a
thrips, Liothrips urichi Karny (Thysanoptera, Phlaethripidae), which was introduced in 1953, works well in
open areas but not in the shade of forests; while the
fungus, Colletotrichum gloeosporioides (Penz) Sace.
f.sp. clidemiae (Coelomycetes, Melanconiales), introduced in 1986, is efficient in shady and wet places. Both
the insect and pathogen would be complementary in
Runion and Mayotte settings. The Hawaii Department
of Natural Resources and University of Hawaii are still
testing other agents such as Lius poseidon Napp, a beetle; and moths Antiblemma acclinalis Hubner, Carposina bullata Meyrick and Mompha trithalama Meyrick
(Julien and Griffiths, 1998; Nakahara et al., 1992;
PIER, 2006a; Trujillo, 2005)
Case 4: Pistia stratiotes L. (Araceae): Four sites are
affected (Martinique, Runion, Bermuda, New Caledonia). Its origin is unknown, but it is now pan-tropical.
Water lettuce is a common aquatic weed in countries
with hot climates. It is a floating, rosette-forming,
stemless, stoloniferous herb. This plant usually propagates by means of stolons which break easily from the
plant and it also propagates by seed. It causes similar
problems as E. crassipes on bodies of water. The weevil Neohydronomous affinis Hustache (Curculionidae,
Erirhinae), which was collected in South America,
substantially reduced growth of P. stratiotes in Australia and Zimbabwe. It has now spread to more than
six countries. This is the most sustainable method to
control this free-floating weed. It has been readily established in six countries and has provided substantial
to excellent control in all of them.For introduction in
Runion, N. affinis can be obtained from PPRI in South
Africa (DeLoach, 1978; Dray and Center, 2003; Foxcroft and Richardson, 2003).
Case 5: Ageratina riparia (Regel) R.M.King &
H.Rob.(Asteraceae): Three sites are affected (Canar
ies, Runion, Bermuda). Mistflower is native to
Mexico. It is a low-growing perennial with tiny, white,
daisy-like flowers. It rapidly invades disturbed areas
and tends to spread along gullies and river banks. It
is rather a hemisciaphilous species which is confined
to the forest margins, paths, and gullies in subtropical to temperate climates. Chemicals from its leaf-litter
suppress the growth of other plants, giving mistflower
a further competitive advantage.A plume moth, Oidematophorus beneficus Yano & Heppner (Lepidoptera,
Pterophoridae), a gall wasp, Procecidochares alani
Steyskal (Diptera, Tephritidae) anda smut fungus,
Entyloma ageratinae Barreto & Evans (Ustilaginales,
Basidiomycotina), were introduced in Hawaii to attack
this aggressive weed in the mid-1970s. Biological control of mistflower in Hawaii has been an outstanding
success. Of the three agents, the fungus was the most
effective and it achieved total control of the plant in
wet areas within 8 months, and in dry areas within
38 years.The plant has remained under control ever
since. Mistflower has increasingly become a problem
in northern New Zealand. A feasibility study showed
that infested areas of New Zealand were likely suitable
for the mistflower agents, so the smut fungus and the
gall wasp were released in New Zealand in 1998 and
2001, respectively. Both are establishing and spreading rapidly and it looks promising that the plant will
be successfully controlled there too. Technology transfer from Hawaii or New Zealand to Runion, Canaries
or Bermuda would be easy. In Runion, no endemic
species belonging to the Eupatoriae tribe are present,
and there are only two indigenous species of the Adenostemma genusthese should be tested to determine
the specificity of the potential agents (Frohlich et al.,
2000; Morin et al., 1997; PIER, 2006b).
Conclusion
Biodiversity is more threatened by alien invasive plants
in tropical European overseas departments, territories,
and countries than in continental European countries
but very few classical weed biological control programmes have been undertaken to date. Many invasive
species are common to several EORTs and many of
them are already under biological control programmes
in other countries. These biological control agents and
technologies could be easily transferred within collaborative European and international programmes. Prior
to any introduction, a review of host-specificity test
results is necessary to determine whether complementary tests should be done according to the indigenous
or endemic flora conservation concerns in each setting.
We have described some examples of invasive species
that could be controlled with a high probability of success by several existing and proven biological control
agents, such as E. crassipes, U. europaeus, C. hirta,
A. riparia, and P. stratiotes. Many other weed species
481
XII International Symposium on Biological Control of Weeds
could also be targeted in EORTs by classical biological control throughout inter-EORT collaboration, e.g.
C. gloeosporioides f.sp. miconiae from French Polynesia to control M. calvescens at an early invasion stage
in New Caledonia and the Canaries, or through international collaborations (e.g. with Hawaii). EORTs should
be highly suitable places to implement classical biological control of alien invasive plants.
References
Anonymous (2003) Gorse (Ulex europaeus), Weed Management Guide, 6. NHT, CRCWM, Australia. 6 p.
Brondeau, A. and Triolo, J. (2007) Etablir des stratgies de
lutte oprationnelles contre les plantes exotiques invasives : Exemples lle de La Runion. Forum des gestionnaires 2007Etablir des stratgies de lutte contre les
plantes invasives, Paris, France. 9 p.
Chazeau, J., Bouye, E. and Bonnet De Larbogne, L. (1989)
Lutte biologique contre le psylle Heteropsylla cubana,
ravageur du faux-mimosa Leucaena leucocephala en
Nouvelle Caldonie. ORSTOM, Nouma, Nouvelle Caldonie. 82 p.
Cochereau, P. (1972) La lutte biologique dans le Pacifique.
Cahiers de lORSTOM, srie Biologie 16, 89103.
Conservation-International. (2006) Biodiversity Hotspots.
Conservation International. Available at: http://www.
biodiversityhotspots.org/xp/Hotspots/ (accessed 07 March
2006).
Davies, J.T., Ireson, J.E. and Allen, G.R. (2004) The role of
natural enemies in regulating populations of biocontrol
agents on gorse (Ulex europaeus L.). In: Sindel, B.M.
Johnson, S.B. (eds) Proceedings of the 14th Australian
Weeds Conference. Charles Stuart University, Wagga
Wagga, Australia, pp. 101104
De Garine-Wichatitsky, M., Barr, N., Blanfort, V., Brescia, F., Chazeau, J., Fogliani, B., Jaffre, T., Jourdan, H.,
Meyer, J.Y., Papineau, C. and Tassin, J. (2004) Altration
de la biodiversit terrestre des les franaises du Pacifique:
effets de lanthropisation et des invasions biologiques.
In: Assises de la Recherche Franaise dans le Pacifique,
pp. 8996. Nouma, Nouvelle Caldonie.
Deloach, C.J. (1978) Considerations in Introducing Foreign
Biotic Agents to Control Native Weeds of Rangelands. In:
Freeman, T.E. (ed.) Proceedings of the IVth International
Symposium on Biological Control of Weeds. University of
Florida, Gainsville, USA, pp. 3950.
Dray, F.A. and Center, T.D. (2003) Watterlettuce - Biological control of invasive plants in the Eastern United States.
Available at: http://www.invasive.org/eastern/biocontrol/
5Waterlettuce.html (accessed 07 March 2006).
Foxcroft, L. and Richardson, D.M. (2003) Managing alien
plant invasions in the Kruger National Park, South Africa.
In: Child, L. E., Brock, J. H., Brundu, G., Prach, K., Pysek,
P., Wade, P. M. and Williamson, M. (eds) Plant Invasions:
Ecological Threats and Management Solutions Backhuis
Publishers, Leiden, The Netherlands, pp. 385403.
Frohlich, J., Fowler, S.V., Gianotti, A., Hill, R.L., Killgore,
E., Morin, L. and Sugiyama, L. (2000) Biological Control
of Mist Flower (Ageratina riparia, Asteraceae): Tranferring a Successful Program from Hawaii to New Zealand.
In: Spencer, N. (ed.) Proceedings of the Xth International
symposium on biological control of weeds. Bozeman,
Montana, USA, pp. 5157.
Gargominy, O. (2003) Biodiversit et conservation dans les
collectivits franaises doutre-mer. Comit franais pour
lUICN, Paris, France. 246 p.
Gutierrez, J. (1976) Proposition dintroduction dinsecte dprdateur du lantana (Lantana camara L. Verbenaceae),
en Nouvelle Caldonie. ORSTOM, Nouma, Nouvelle
Caldonie. 9 p.
Gutierrez, J. (1979) Participation of the laboratory of applied
entomology of the ORSTOM Center, Noumea, in various
biological control operations undertaken within the South
Pacific area. ORSTOM, Nouma, Nouvelle Caldonie.
22 p.
Hivert, J. (2003) Plantes exotiques envahissantesEtat des
mthodes de lutte mises en oeuvre par lOffice National des
Forts la Runion. ONF, Saint Denis, Runion. 319 p.
Joseph, P. (2006) Les Petites Antilles face aux risques
dinvasion par les espces vgtales introduites. Lexemple
de la Martinique. Revue dEcologie (Terre et Vie) 61,
209224.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds. A world catalogue of agents and their target
weeds. CABI Publishing, Wallingford, UK. 223 p.
Julien, M.H., Griffiths, M.W. and Wright, A.D. (1999) Biological control of water hyacinth. The weevils Neochetina
eichhorniae and N. bruchi: biologies, hot ranges, and rearing, releasing and monitoring techniques for biological
control of Eichhornia crassipes. ACIAR monograph 60.
Canberra, Australia. 87 p.
Julien, M.H., Griffiths, M.W. and Stanley, J.N. (2001) Biological control of water hyacinth 2. The moths Niphograpta
albiguttalis and Xubida infusellus: biologies, hot ranges,
and rearing, releasing and monitoring techniques for biological control of Eichhornia crassipes. ACIAR mongraph 79. Canberra, Australia. 90 p.
Kairo, M., Ali, B., Chessmann, O., Haysom, K. and Murphy, S. (2001) Invasive Species Threats in the Caribbean
RegionReport to the nature conservancy. CAB International, West Indies, UK. 134 p.
Krause, M.A., Beck, A.C. and Dent, J.B. (1988) Control
of Gorse in Hill Country: An Economic Assessment of
Chemical and Biological Methods. Agricultural Systems
26, 3549.
Le Bourgeois, T. and Lebreton, G. (2006). Expertise sur la
gestion des plantes aquatiques envahissantes Eichhornia crassipes (Jacinthe deau) et Pistia stratiotes (Laitue
deau) dans les tendues deau douce littorales de la Runion. Cirad, Saint Pierre, Runion. 29 p.
Loope, L., Sheppard, A.W., Pascal, M. and Jourdan, H. (2006)
Question 8 : Lradication : une mesure de gestion des populations allochtones. In: Bauvais, M.-L., Colno, A. and
Jourdan, H. (eds) Espces envahissantes dans larchipel
no-caldonien. IRD, Montpellier, France, pp. 401418.
Macdonald, I.A.W., Thbaud, C., Strahm, W.A. and Strasberg, D. (1991) Effects of Alien Invasions on Native
Vegetation remnants on La Runion (Mascarenes Islands,
Indian Ocean). Environmental Conservation 18, 5161.
Medeiros, C. (2006) Survey of the main plant invaders of the
Madeira Laurisilva. In: Programme and Abstracts of the
International Symposium Intractable Weeds and Plant Invaders, p. 26. Ponta Delgada, Azores, Portugal.
482
Opportunities for classical biological control of weeds in European overseas territories
Meyer, J.Y. (1998) Perspectives davenir pour la lutte centre
Miconia calvescens en Polynesie francaise : strategie gnrale et tactiques de terrain. In: Meyer, J.Y. and Smith,
C.W. (eds) Proceedings of the First Regional Conference
on Miconia Control. Papeete, Tahiti, French Polynesia.
Gouvernement de Polynesie franqaise/University of Hawaii at Manoa/Centre ORSTOM de Tahiti, p. 90.
Meyer J.Y. (2000) Preliminary review of the invasive plants
in the Pacific islands (SPREP Member countries). In:
Sherley, G. (ed.) Invasive species in the Pacific: A technical review and draft regional strategy. South Pacific Regional Programme 2000, Apia, Samoa, pp. 85114.
Meyer, J.Y. and Killgore, E. (2000) First and successful release of a bio-control pathogen agent to combat the invasive alien tree Miconia calvescens (Melastomataceae) in
Tahiti. Aliens 12, 8.
Meyer J.Y. (2004) Threat of Invasive Alien Plants to Native
Flora and Forest Vegetation of Eastern Polynesia. Pacific
Science 58, 357375.
Meyer, J.Y., Loope, L., Sheppard, A.W., Munzinger, J. and
Jaffr, T. (2006) Les plantes envahissantes et potentiellement envahissantes dans larchipel no-caldonien:
premire valuation et recommandations de gestion. In:
Beauvais, M.-L., Colno, A. and Jourdan, H. (eds) Les espces envahissantes dans larchipel no-caldonien. IRD,
Paris, France, pp. 50115.
Mittermeier, R.A., Da Fonseca, G.A.B., Hoffmann, M.,
Pilgrim, J., Brooks, T., Gill, P.R., Mittermeier, C.G. and
Lamoreux, J. (2005) Hotspots revisited: Earths biologically richest and most endangered terrestrial ecoregions.
CEMEX, Conservation International. 392 p.
Morin, L., Hill, R.L. and Matayoshi, S. (1997) Hawaiis
successful biological control strategy for mist flower
(Ageratina riparia)can it be transferred to New Zealand? BioControl 18, 7788.
Nakahara, L.M., Burkhart, R.M. and Funasaki, G.Y. (1992)
Review and status of biological control of Clidemia
in Hawaii. In: Stone, C. P., Smith, C. W. and Tunison,
J. T. (eds) Alien Plant Invasion in Native Ecosystems
of Hawaii. Cooperative National Park Resources
Studies Unit, University of Hawaii, Honolulu, Hawaii,
pp. 452465.
Pier. (2006a) Clidemia hirta. Pacific Island Ecosystems
at Risk (PIER). Available at: http://www.hear.org/pier/
species/clidemia_hirta.htm (accessed 07 March 2006).
Pier. (2006b) Ageratina riparia. Pacific Island Ecosystems
at Risk (PIER). Available at: http://www.hear.org/pier/
species/ageratina_riparia.htm (accessed 04 March 2006).
Quilici, S., Franck, A., Montagneux, B. and Tassin, J. (1995)
Successful establishment on Runion Island of an exotic
ladybird, Olla V-nigrum, for the biocontrol of Leucaena
psyllid, Heteropsylla cubana. In: FAO (ed.) Proceedings
of the Workshop of Leucaena psyllid: a threat to agroforestry in Africa. Dar-es-Salaam, Tanzania, pp. 147154.
Randall, R.P. (2002) A global compendium of weeds. Richardson, R.G., Richardson, F.J., Melbourne, Victoria, Australia. 905 p.
Sanz-Elorza, M., Dana, E.D. and Sobrino, E. (2005) Aproximacion al listado de plantas vasculares aloctonas invasoras reales y potenciales en las islas Canarias. LAZAROA
26, 5566.
Silva, L. and Smith, C.W. (2004) A characterization of the
Non-indigenous Flora of the Azores Archipelago. Biological Invasions 6, 193204.
Trujillo, E.E. (2005) History and success of plant pathogens
for biological control of introduced weeds in Hawaii. Biological Control 33, 113122.
Varnham, K. (2005) JNCC Report 372: Database of nonnative species occurring in UK Overseas Territories: a
review. JNCC, Peterborough, UK. 372 p.
483
Weed biological control regulation in
Europe: boring but important
R.H. Shaw1
Summary
If the Europe Union (EU) is to deliver on its Convention on Biological Diversity (CBD) commitments
and its Member States are to have any chance of achieving a good status classification for their water
bodies, classical biological control will be needed. A lack of history and considerable inertia impede
the development of biological control programmes but an inappropriate and ineffective regulatory
environment could prove more of a barrier to weed biological control implementation. This paper
reviews the current situation in Europe and considers the codes and regulations for the release of
arthropods and fungi before considering the suitability of pest risk assessments.
Keywords: legislation, EU, pest risk assessment.
Introduction
As signatories to the Convention on Biological Diversity, European Union Member States (MS) have an
obligation to prevent the introduction of, control or
eradicate those alien species which threaten ecosystems, habitats and species (Decision VI/23 in 1992).
They are also encouraged to invest in research and assessment of biological control as a control option. The
European Strategy on Invasive Alien Species (ESIAS)
came in to being in 2003, and calls for a regional approach to the problem highlighting the need for cost:
benefit analyses of long-term control measures. Exotic weeds are amongst the most problematic of invasive species but are amenable to biological control
(Cruttwell McFadyen, 1998). In parallel many herbicides have been lost as a result of changes in registration
requirements and there is no shortage of invasive plant
species in Europe and more potential invasive alien
weeds arrive daily. A recent review by EPPO recorded
the shipments of aquatic plants into one French airport
and revealed an average of more than one shipment per
day of exotic plants. In the month of May 2006 alone,
almost 100,000 plants arrived through this port (EPPO,
2007) including ten species that are known problematic
invaders in Europe and another nine that are not yet
present in Europe but known to be invasive elsewhere.
As exotic species with a lack of specialist natural en CABI E-UK Centre, Bakeham Lane, Egham, Surrey TW20 9TY, UK
<[email protected]>.
CAB International 2008
1
emies, many of the invasive weeds in Europe would
be amenable to classical biological control (Sheppard
et al., 2006) and some have off-the-shelf agents that
have been thoroughly tested elsewhere and are likely
to be safe for release. It would seem that the stage is
set for the expansion of classical biocontrol of weeds
into Europe.
In much the same way as the weed invades, biological control activity could be expected to increase rapidly from a slow start, and the current lack of precedent
could be a reason in itself for a lack of take-up. In reality, a major factor is public and political perception.
Plants are perceived by the public and their representatives as less threatening than insect pests since the
latter more obviously threaten our food supply. Thus, a
proposal to engage in an irreversible introduction of yet
another alien species can seem counterintuitive. Fear of
what might go wrong is certainly a consideration when
a new concept is proposed, but other factors may prove
to be more important. Though public perceptions and
governmental needs are changing, in part thanks to ongoing programmes against Japanese knotweed, Fallopia japonica (Djeddour et al., these proceedings), the
crucial regulatory environment is far from ideal.
The regulations
The regulatory framework in Europe is characterised by
a lack of clarity and accompanied by varied interpretation. The lack of history of weed biological control in
Europe is almost certainly the cause of the legislative
gaps and similar situations existed in more experienced
484
Weed biological control regulation in Europe: boring but important
countries prior to their development of tailored legislation such as the Australian Biocontrol Act 1986, and the
New Zealand Hazardous Substances and new Organism
Act 1996. Thus, many countries find themselves using
regulations that were not designed for the purpose of
dealing with the introduction and release of exotic biological control agents such as those provided for plant
quarantine, genetically modified organisms and wildlife conservation. The responsible Government department is often unclear about how to proceed, and this
was a significant factor in the failure to make positive
decisions to release agents in the bracken, Pteridium
aquilinum (L.) Kuhn, programme in the 1980s (S.V.
Fowler, personal communication, 1998).
To assess the situation in Europe, a letter was sent
out to the Plant Protection Department of each European country in June 2005 requesting information on
their regulation of classical weed biological control
agents in general and more specifically those relating
to plant pathogenic fungi. Of the 25 recipients of the request, 11 responded. What became clear was that classical agents, both insect and fungal, cause considerable
problems not only in practical regulatory terms but also
in understanding. Some countries clearly had no understanding of the meaning and implications of classical
biological control, let alone the regulatory paths for the
two taxa. In some cases the responsible authority for
initial import and actual release differed, as did the departments dealing with fungi and insects.
The confusion comes about because, even though
the ultimate goal and strategy of classical biological
control is the same whether insect or fungal agents are
considered, they are regulated very differently by the
EU, and each country has to fit the requirements of its
own legislation. For example, one of the early milestones in a biological control programme is agreeing to
the test plant list to be used during host range testing.
This is carried out by the Technical Advisory Group
(TAG) in North America, but in the UK, for example,
there appears to be no mechanism for agreeing to such
lists prior to the application for release.
Arthropods
European directive 2000/29/EU protects Europe
against the introduction and spread of blacklisted
known pests and harmful macroorganisms by their
prohibition. However, any species not on the list (including all potential classical biological control agents
and many potentially invasive plants) can therefore be
introduced without any formal risk assessment. This
suggests that there is no EU level provision for assessing releases of beneficial exotics.
However, each EU member state transposes and interprets the ED 2000/29/EU into their national legislation and as a result can, and should, develop regulatory
procedures for the releases of non-indigenous macroorganisms such as arthropod classical biological control
agents (Genovesi and Shine, 2004). Currently there
remains a wide divergence in such regulatory requirements from virtually none in many EU countries, to an
incipient risk assessment process being developed with
appropriate regulatory bodies in, for example, the UK
and Portugalthe two countries most advanced in developing classical weed biological control programmes.
In the commercial world a lack of clear legislation is
often exploited. However, that is generally not the case
in the public-interest, public-investment research sector, and consequently, weed biological control releases
are conspicuous by their absence.
There have been over a thousand releases of biological control agents for weeds worldwide (Julien
and Griffiths, 1998) yet no full classical weed biological control programme has been carried out in any EU
Member State. This contrasts strongly with the repeated
use of exotic insects to control insect pests in Europe.
The BIOCAT database (Greathead and Greathead,
1992 updated to end-2004) lists a staggering 137 species that have been included in a total of 276 releases
since 1901.This extensive use of exotic natural enemies
in Europe has been poorly regulated in some countries
and, as a result, has not been without problems. This is
not surprising since, for example, of the 65 species to
have been introduced since 1908 into Spain alone, only
four are considered monophagous (Jacas et al., 2006).
Thus, it is highly likely that extensive non-target effects are occurring without being noticed. The recent
problem with the predatory ladybird Harmonia axyridis (Pallas) (Majerus et al., 2006) highlights the dangers and may be another factor to hinder the acceptance
of classical weed biological control in Europe.
There is an inherently greater general precaution
over releases of herbivorous vs entomophagous arthropods because, as in the United States, the release
agency is likely to be legally responsible should any
biological control agent cause economic loss or significant environmental damage (Delfosse, 2005; Miller
and Aplet, 1993).
EU countries would be expected to use the EPPO
standards on the safe use of biological control (EPPO,
2000) as well as and the newly revised international
advisory Guidelines for the Export, Shipment, Import
and Release of Biological Control Agents and Organisms Claimed to be Beneficial otherwise called the
International Standards for Phytosanitary Measures,
publication No. 3 or ISPM 3 (IPPC, 2005) [reviewed
by Genovesi and Shine (2004); Kairo et al., (2003)].
These documents also highlight the need for consultation between relevant neighbouring countries. Interestingly, ISPM 3 recommends that the National
Plant Protection Organization should conduct a pest
risk assessment (PRA) either before import or before
release of biological control agents and other beneficial organisms. It is also inclusive of fungal agents, i.e.
non-formulated micro-organisms that are released with
the expectation of establishment. It follows, therefore,
485
XII International Symposium on Biological Control of Weeds
that any national regulatory procedures developed by
EU member states for the release of classical biological
control agents, as a result of EU directive 2000/29/EU,
and based on ISPM 3, should be technically capable of
including both macro- and microorganism agents.
Another organisation that became involved in the
regulation of arthropod biological control agents was
the Organisation for Economic Cooperation and Development who produced a set of guidelines (OECD,
2004). These guidelines are very thorough and deal with
the likelihood of establishment and efficacy as well as
direct non-target effects. Seeing the threat posed to the
commercial biological control industry if these guidelines became the template for national regulation, the
International Biocontrol Manufacturers Association
(IBMA) proposed that the International Organisation
for Biological Control (IOBC) should oversee the harmonization of all these documents into one. This harmonization document was published in 2005 (Bigler
et al., 2005) with the stated aim of providing advice to
the competent national authorities on the information
required to perform risk assessments with respect to the
import and release of an insect biological control agent
(IBCA). Most of the guidelines remain applicable to
weed biological control agents but these have been specifically excluded from consideration. The sections are
divided as follows:
Information requirements for the importation of
non-native IBCAs for research
Information requirements for the deliberate release
of non-native IBCAs
not previously released
previously authorized for release
Information requirements for the release of native
IBCAs
Despite this mass of non-binding documentation it is
yet to be shown how any nation will actually implement
the guidelines in their national legislation when it comes
to the release of an invertebrate classical weed biological control agent. Amazingly, the situation is even more
bizarre when one considers plant pathogens.
Plant pathogens
The regulatory framework that exists in the EU
for the use of plant pathogens in classical biological
control is officially driven by one central EU directive
which effectively hinders and may actually prevent
their use. Although aimed at minimizing the use of
chemicals by regulating the placing of plant protection
products on the market, the EU directive for chemical
pesticide regulation 91/414/EU, as updated by Council
Directive 2005/25/EU, has been written in such a way
as to include, by default, microorganisms as classical
biological control agents. For example, the definitions
section of the directive begins with a reference to the
form in which the products are supplied to the user and
the whole process revolves around label claims. This
is contrary to classical biological control as there is no
supply to users and there are no labels. That no specific
consideration was given to micro-organisms as classical biological control agents is perhaps not unexpected,
since it is a technique with no history in Europe. The
situation, however, has been described as totally inadequate as it hinders progress (Seier, 2005).
A highly-specific obligate plant pathogen, released
once in order to provide permanent control of the target weed, generates no subsequent sales, but must still
go through a regulatory assessment application designed for non-specific herbicides. Such applications
require large amounts of scientific data inappropriate
for the use of pathogens as classical biological control
agents (e.g. mammalian toxicity and efficacy versus
current chemical alternatives). In addition, the application costs are high, although some EU countries appear
willing to reduce the costs of application. The UK currently has a Biopesticide Scheme with dossier assessment costs of about 23,000 (33,400 Euro) and allows
a reduced data package. Nonetheless, these costs and
requirements are even considered prohibitive to the
commercial biological control manufacturers who are
producing products with labels and generating subsequent income.
Classical biological control agents are normally used
in line with the EU goal of reducing chemical inputs to
the environment. Yet such agents are blocked by inadequacies in the legislation aimed at stricter assessment
of new pesticides. The consequences of 91/414/EU are
wide reaching and could prevent the use of plant pathogens in classical biological control in Europe, despite
an impeccable worldwide safety record (Barton, 2004)
and high levels of effectiveness (Charudattan, 2005).
Recent reviews of the 91/414/EU directive have
separated consideration of chemicals from microorganisms but there remains a need for a new directive or for
revisions to cover classical biological control agents.
As this will take considerable time, an interim measure
might be to indicate that Member States apply the directive only to formulated products, thereby distinguishing between those microorganisms considered
for commercialization (i.e. those requiring labels with
storage, application and safety information) and those
considered as classical fungal agents to be released
once, or a few times, in the public interest.
One glimmer of hope is the current EU-funded project looking at the regulation of biological control agents
in Europe (REBECA). The groups involved come from
all fields of biological control from scientists, through
producers to regulators, and are expected to make recommendations to the Commission. At a recent REBECA
microbials subgroup meeting it was accepted that classical weed biocontrol agents that are fungi should be
excluded from 91/414 (B. Ritchie, 2006, personal communication). It will be a step forward if such a recommendation is accepted by the EU.
486
Weed biological control regulation in Europe: boring but important
The Japanese knotweed programme (see Djeddour
et al., these Proceedings) may be the first to test the regulatory framework, since both an insect and a leafspot
fungus are primary candidate agents. Communications
with various government departments since the late
1990s have attempted to clarify the pathway for each
type of agent with little progress until 2005. Then initial
meetings were held with the Pesticide Safety Directorate through whom the UK Governments Department
of Environment Food and Rural Affairs (Defra) implements 91/414. Negotiations continue with an interdepartmental committee which was established just to
deal with this particular issue. All parties are taking a
pragmatic approach to the issues and mutually agreeable and practical solutions are expected.
Pest Risk Assessment (PRA)
As intimated above, it would seem that some version of a PRA will be favoured by EU Member States.
Historically the treatment of potential crop pests has
differed from that of uncultivated plants but these are
now coming together in the key area of risk analysis
(Baker et al., 2005). The International Plant Protection
Convention (IPPC) was signed in 1951 as a response to
increasing pest invasions, in particular that of the Colorado potato beetle, Leptinotarsa decemlineata Say. The
IPPC has since produced 21 International Standards for
Phytosanitary Measures (ISPMs) which are recognized
by the World Trade Organisation (WTO). If a pest is
of potential economic importance then it is considered
a quarantine pest and joins the lists of such organisms
posted by nations and trading blocks. To avoid unnecessary barriers to trade, both the IPPC and WTO stipulate that quarantine status and controls can only be put
in place once a real threat is identified through a pest
risk analysis. The tools for this are ISPM2 (FAO, 1996)
and ISPM 11 (FAO, 2001).
According to the ISPM 2 revision document, which
is currently out for member country consultation, biological control agents are intended to be beneficial to
plants or plant products without causing harm. This is
true from an ecosystem perspective but quite the opposite of the intention when one considers the target plant
to which harm is most definitely intended. It goes on to
say that, when performing a PRA or monitoring their
release, the main concern is unanticipated harm to nontarget organisms in the PRA area. The most appropriate current tool for assessing the risks associated with a
biological control release is probably ISPM 11, which
was established in 2001 and was updated in 2004 to
include analysis of environmental risk and living modified organisms (FAO, 2004). Unhelpfully, it refers to
pests throughout which from a biological control perspective is confusing since biological control agents
must then be called beneficial pests. Nonetheless, the
IPPC defines a pest as any species strain or biotype of
plant, animal or pathogenic agent, injurious to plants
or plant products (FAO, 2004). As such, classical biological control agents rest firmly in ISPM 11. Furthermore, in section 1.1.2 of the latest version of ISPM 11,
it is stated that a PRA may be initiated if a request is
made to import an organism.
The general requirements of a PRA are fairly similar country to country but since they are written for
pests, much of the information required is not necessarily appropriate for weed biological control. An
analysis of the questions posed found that half are inappropriate, mainly because they deal with the risk of
the pests arrival and the prospects for its control. It
seems increasingly likely that this form of PRA will
be applied to biological control agents, at least in the
UK, and it will probably be under Plant Health regulations, as suggested above. This is much the same as
the situation in South Africa where the Directorate of
Plant Health and Quality regulates any biological control agent release.
Discussion
Biological control of weeds in Europe is just beginning and has all the teething troubles associated with
such a period. The unclear funding and regulatory
situation coupled with a general inertia has hindered
the development of classical biological control and in
some areas Europe reflects the situation in countries
such as Australia and New Zealand as it was decades
ago. Europe has an advantage in that it can learn from
history and avoid the painful mistakes made by our
pioneering ancestors, especially since a lot of biological control expertise is already present in European
countries.
It is clear that at least those regulations governing
fungal biological control agents are in need of revision.
My proposal to precede description of Plant Protection
Product with the word formulated in the 91/414 Directive could solve most of the problems surrounding
the presumably mistaken inclusion of classical fungal
weed agents. As is often the case with authorities, simply trying to avoid regulations by requesting an exemption is not the best course, and the inclusion into Plant
Health legislation would seem to be the best solution.
However, it is likely that the early applications will
need to be considered by all interested parties in the
absence of tailored regulations.
As Europe comes to terms with its CBD commitments and the growing scale of invasions by environmental weeds, classical biological control should
become more commonplace. This is particularly so in
aquatic and riparian systems, where most of Europe
completely bans the use of chemical herbicides and no
real alternative exists. The driving force in such delicate habitats may well turn out to be the Water Framework Directive (Dec. 2000) which requires parties
to ensure that all their waterways reach good status
by 2015. The presence of invasive alien species in or on
487
XII International Symposium on Biological Control of Weeds
these waterways logically would prevent this goal being
achieved and should focus attention on alternative approaches to their management, ones that are commonplace in countries in other continentsthat is, classical
biological control.
Acknowledgements
I would like to thank the Defra Japanese knotweed
Project Board for their advice and counsel as well as
the project funders for supporting the research.
References
Baker, R., Cannon, R., Bartlett, P. and Barker, I. (2005) Novel
strategies for assessing and managing the risks posed by
invasive alien species to global crop production and biodiversity. Annals of Applied Biology 146, 177191.
Barton, J. (2004) How good are we at predicting the field
host-range of fungal pathogens used for classical biological control of weeds? Biological Control 31, 99122.
Bigler, F., Bale, J.S., Cock, M.J.W., Dreyer, H., Greatrex,
R., Kuhlmann, U., Loomans, A.J.M. and Lenteren, J.C.v.
(2005) Guidelines on information requirements for import and release of invertebrate biological control agents
in European countries, CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources 1, 10.
Charudattan, R. (2005) Ecological, practical, and political inputs into selection of weed targets: what makes a
good biological control target? Biological Control 35,
183196.
Cruttwell McFadyen, R.E. (1998) Biological control of
weeds, Annual Review of Entomology 43, 369393.
Delfosse, E.S. (2005) Risk and ethics in biological control.
Biological Control 35, 319329.
EPPO. (2007) Pathways analysis: aquatic plants imported in
France. EPPO Reporting Service 16, 1824.
EPPO. (2000) Safe use of biological control: Import and release of exotic biocontrol agents. PM6/2(1), 14.
FAO. (2004) Pest risk analysis for quarantine pests, including analysis of environmental risks and living modified
organisms. International Standards for Phytosanitary
Measures No. 11. Food and Agriculture Organisation,
Rome.
FAO. (2001) Pest risk analysis for quarantine pests, International Standards for Phytosanitary Measures No. 11.
Food and Agriculture Organisation, Rome.
FAO. (1996) Guidelines for pest risk analysis. International
Standards for Phytosanitary Measures No. 2. Food and
Agriculture Organisation, Rome.
Genovesi, P. and Shine, C. (2004) European Strategy on Invasive Alien Species. Strasbourg: Council of Europe Publishing.
Greathead, D.J. and Greathead, A.H. (1992) Biological control of insect pests by insect parasitoids and predators: the
BIOCAT database. Biocontrol News and Information 13,
61N68N.
IPPC. (2005) Guidelines for the Export, Shipment, Import
and Release of Biological Control Agents and Organisms
Claimed to be Beneficial. Secretariat of the International
Plant Protection Council, Food and Agriculture Organisation, Rome.
Jacas, J.A., Urbaneja, A. and Viuela, E. (2006) History and
future of introduction of exotic arthropod biological control agents in Spain: a dilemma? Biocontrol 51, 130.
Julien, M.H. and Griffiths, M.H. (1998) Biological control
of weeds; A world catalogue of agents and their target
weeds (Fourth edition). CABI Publishing, Wallingford,
UK, 223 p.
Kairo, M.T.K., Cock, M.J.W. and Quinlan, M.M. (2003)
An assessment of the use of the Code of Conduct for the
Import and Release of Exotic Biological Control Agents
(ISPM No. 3) since its endorsement as an international
standard. Biocontrol News & Information 24,1527.
Majerus, M., Strawson, V. and Roy, H. (2006) The potential
impacts of the arrival of the harlequin ladybird, Harmonia
axyridis (Pallas) (Coleoptera: Coccinellidae), in Britain.
Ecological Entomology 31, 207215.
Miller, M. and Aplet, G. (1993) Biological control: a little
knowledge is a dangerous thing. Rutgers Law Review 45,
285334.
OECD. (2004) Guidance for information requirements for
regulation of Invertebrates as biological control agents
(IBCAs). OECD Environment, Health and Safety Publications Series on Pesticides, Organisation for Economic
Co-operation and Development, 22 p.
Seier, M.K. (2005) Exotic beneficials in classical biological
control of invasive alien weeds: friends or foes? In: Plant
Protection and Plant Health in Europe: Introduction and
Spread of Invasive Species. Held at Humboldt University,
Berlin, Germany, 911 June 2005, 191196. (Online
proceedings available at: http://dpg-bcpc-symposium.de/
fileadmin/alte_Webseiten/Invasive_Symposium/articles/
articles.htm)
Sheppard, A.W., Shaw, R.H. and Sforza, R. (2006) Top 20
environmental weeds for classical biological control in
Europe: a review of opportunities, regulations and other
barriers to adoption. Weed Research 46, 125.
488
Abstracts: Theme 7 Opportunities and Constraints for the Biological Control of Weeds in Europe
Field evaluation of Fusarium oxysporum as a
biocontrol agent for Orobanche ramose
E. Kohlschmid, D. Mller-Stver and J. Sauerborn
University of Hohenheim, Institute for Plant Production and Agroecology in the Tropics and
Subtropics (380), 70593 Stuttgart, Germany
Under the changing agro-climatic conditions of western Europe, the root parasitic weed Orobanche
ramosa infests at a progressing rate host crops such as hemp, tobacco and to an increasing degree
oilseed rape in France. Fusarium oxysporum (FOG) was isolated from O. ramosa tubercles, parasitizing tobacco in Germany. The fungus was formulated in wheat flour kaolin (Pesta) granules and
showed promising results in controlling O. ramosa under greenhouse conditions, reducing number
and dry matter of the parasite by up to 90 %. Consequently FOG was tested under field conditions
using different application techniques. In-furrow application and broadcasting of the inoculum after tobacco planting as well as subsoil application pre-planting decreased the number of Orobanche shoots.
In a further experiment, in-furrow application of FOG markedly reduced number and dry matter of
O. ramosa. However, no distinct further reduction could be noticed when biocontrol was combined with
a resistance reducer. The results revealed the potential of plant pathogens for Orobanche ramosa control
and future experiments should work on enhancing the observed effect under natural conditions.
Potential for biological control of
Hydrocotyle ranunculoides in Europe
R. Shaw1 and J.R. Newman2
CABI Bioscience UK (Ascot), Silwood Park, Buckhurst Road, Ascot, SL5 7TA
CEH Wallingford, Maclean Building, Crowmarsh Gifford, Wallingford, OX10 8BB
1
Hydrocotyle ranunculoides is an invasive aquatic macrophyte, present in several European countries
and elsewhere outside its native range of southern North America. The plant has spread from single
introductions to occupy marginal habitats of over 30 miles in at least three rivers in the UK in the last
4 years. Attempts have been made to control this plant using mechanical and chemical means, combined with manipulation of the environment. Most of these have proved unsuccessful due to rapid
recolonisation. Given the restrictions on chemical use for aquatic weed control in Europe, we propose
a novel combination of mechanical control and biological control as a technique to prevent the spread
of the plant within catchments where it is already established, and to eradicate relatively new small
infestations. Data will be provided on collection of potential biological control agents from Argentina,
the initial screening and preliminary host specificity studies.
CAB International 2008
489
XII International Symposium on Biological Control of Weeds
Alien poisonous weeds: a challenge for a biological
control of weeds program in Europe
R. Sforza,1 M. Cristofaro2 and W. Jones1
USDA-ARS European Biological Control Laboratory, CS 90013 Montferrier sur Lez,
34988 St Gly du Fesc, France
2
ENEA-Casaccia, Rome, Italy and BBCA, via del Bosco 10, 00060 Sacrofano (Rome), Italy
1
Considering weeds as harmful organisms threatening ecosystems is acknowledged worldwide, however
there are exotic weeds that have a second serious impact on human beings and livestock due to their
high allergenic or poisonous capability. We review here the opportunity to apply a European biocontrol program against the North American ragweed (Ambrosia artemisiifolia) and silverleaf nightshade
(Solanum elaeagnifolium), and the West-Asian hogweed (Heracleum mantegazzianum). Ragweed is
an annual weed now spreading quickly, infesting at least eight Eurasian countries. The species represents a peculiar case because it is a serious issue for human health with pollen causing severe allergies.
Silverleaf nightshade foliage and unripe fruits contain dangerous levels of solanine, while mature berries contain also high levels of asteroid alkaloids such as solanine and solanosine, which are toxic to
livestock. Animals should be removed from infested areas until control is achieved. Direct contact with
hogweed leaves and/or stems may cause severe blisters on human skin, when glucoside phototoxins in
the plants are activated. This review reports lists of natural enemies known for each species and their
known use in biological control programs outside of Eurasia.
Using augmentative biocontrol against Euphorbia esula:
an innovative program in France
R. Sforza,1 J. Le Maguet,1 B. Gard1 and L. Curtet2
USDA-ARS European Biological Control Laboratory, CS 90013 Montferrier sur Lez,
34988 St Gly du fesc, France
2
ONCFS - CNERA AM Station de la Dombes 01330 Birieux, France
Native from Eurasia, Euphorbia esula ssp. esula, leafy spurge, is usually found in grasslands and disturbed areas in France. In the flooded meadows of the Sane Valley (Ain, Sane-et-Loire), an unusual
spread has been observed since the 1990s affecting the management of grasslands. The increase of this
E. esula could be a threat for this ecosystem of European importance (Natura 2000 site), and for plant
and bird diversity. Chemical control, commonly used by farmers since the end of the 1990s, remained
uneffective as previously observed in the US. In that regard, since 2003, we made extensive surveys
for collecting indigenous natural enemies in leafy spurge stands. Seven indigenous natural enemies,
including a rust, were observed on E. esula, of which the cerambicyd beetle, Oberea erythrocephala,
and the cecidomyid fly, Spurgia sp. were particularly studied. Natural infestations of O. erythrocephala
larval stage ranged 11 to 26% in 2006. Attempts to rear both species are described, particularly with
the beetle on an artificial medium. In Spring 2006, 100 field-collected and artificially reared adults
were released on a 10-sq.m. plot naturally covered with leafy spurge. Damage and infestation rates
were compared to controls. In addition, choice and no-choice tests were estimated with E. esula and
E. palustris, a local protected native plant. Considerations about the future use of O. erythrocephala for
augmentative release is discussed.
490
Abstracts: Theme 7 Opportunities and Constraints for the Biological Control of Weeds in Europe
The biological control of Impatiens glandulifera Royle
R.A. Tanner and H.C. Evans
CABI Bioscience, Silwood Park, Ascot, Berks SL5 7TA, UK
Impatiens glandulifera Royle is a highly invasive weed which has successfully invaded almost every
riparian system in the UK after its introduction from the Himalayas as a garden ornamental in 1839.
Now one of the tallest annual plants in Europe, I.glandulifera is able to outcompete native species due
to its vigorous growth rates, large seed banks and prolific seed dispersal. When I. glandulifera forms
monocultures in riparian habitats, the effects on the ecosystem can be severe, potentially causing bank
erosion, biodiversity loss and increased risk of flooding. The environment agency estimates it would
cost between 150300 million to eradicate the plant in the UK. In August 2006 CABI scientists
conducted a survey of the natural enemies of I. glandulifera in its native range (foothills of the Himalayas, Pakistan). High levels of damage caused by arthropods and fungal pathogens were observed in
all populations sampled, and the most interesting agents were collected and shipped back to the UK
for further testing. This paper presents the results the work conducted in 2006 and explores further the
potential for developing this work into a full biocontrol program against I. glandulifera in Europe.
491
This page intentionally left blank
Theme 8:
Release Activities and
Post-release Evaluations
Session Chair:
Rosemarie De Clerck-Floate
493
This page intentionally left blank
Keynote Presenter
Release strategies in weed biocontrol:
how well are we doing and is there
room for improvement?
S.V. Fowler,1 H.M. Harman,2 J. Memmott,3 P.G. Peterson4 and L. Smith1
Summary
A large number of factors associated with the agent and habitat can be manipulated when making a
biocontrol release. Yet, it is only recently that an experimental approach has been taken, particularly
with the release of arthropod weed biocontrol agents, for example examining factors such as release
size to test ideas originating from theoretical or retrospective studies. For release size, both retrospective analyses and experimental studies with arthropod agents usually show that larger releases have a
higher probability of establishing. However, a limited number of large releases at a few sites also run
the risk of chance extinction from locally major environmental effects. These risks can be minimized
by releasing at many widespread sites, so there is often a trade-off between a few large releases and
many smaller one. Unless these risks and relationships are known, a mixed strategy with arthropods,
using a range of release sizes, is likely to be optimal at least at the start of a release programme. Recent theoretical attention has been applied to specific mechanisms creating lower individual fitness in
small populations, termed Allee effects. It appears that Allee effects, including genetic inbreeding and
problems in finding mates, may have been underestimated: They can be powerful, as witnessed by
deliberate extinctions using sterile male release. Genetic issue with biocontrol releases, such as strain
selection, inbreeding depression, drift and creation of laboratory-selected strains, have also received
considerable, mostly theoretical, analysis. However, the actual rate of establishment success per agent
species in weed biocontrol programmes is now 80100%. These overall analyses do hide problems
in establishing particular agent species: Programmes in New Zealand targeting recalcitrant environmental weeds have been hampered by partial or complete failure to get potentially critical agent
species established. A case study of heather beetle in New Zealand is used to illustrate the frustrations
and research challenges in a release programme.
Keywords: release size, Allee effects, genetics of biocontrol releases, establishment
rates, heather beetle.
Introduction
Classical biological control offers just about the only opportunity to make multiple experimental releases of an
exotic organism into a new environment. Despite this,
most studies of the effect of different release strategies
Landcare Research, PO Box 40, Lincoln, New Zealand.
Landcare Research, Private Bag 92170, Auckland 1142, New Zealand.
3
University of Bristol, School of Biological Sciences, Woodland Road,
Bristol BS8 1UG, UK.
4
Landcare Research, Private Bag 11052, Palmerston North 4442, New
Zealand.
Corresponding author: S.V. Fowler <[email protected]>.
CAB International 2008
1
2
until recently relied on retrospective analyses (Hall and
Ehrler, 1979; Beirne, 1985; Simberloff, 1989; Hopper
and Roush, 1993). Many of these studies called for a
more rigorous experimental approach in classical biological control programmes, and recently, this call has
been heeded (Memmott et al., 1998; Grevstad, 1999a;
Memmott et al., 2005), although there are usually constraints in operational programmes because of the additional expense of experimental releases (McFadyen,
1998). This paper reviews developments with release
strategies, for example research on the effect of release
size on establishment success. Inevitably, our review
is biased towards arthropods used for weed biocontrol,
largely because more research has been done regarding
495
XII International Symposium on Biological Control of Weeds
release strategies with arthropod agents. However,
many of the issues also affect pathogen agents, and
we highlight areas where pathogen release strategies
have received research attention. We also ask what the
establishment rate is for new agent releases in wellresourced biocontrol programmes and whether there is
room for improvement. Lastly, we use the release of
heather beetle in New Zealand to illustrate some of the
ecologically interesting challenges that a partially successful release programme can face.
How to make releases of weed
biological control agents?
In biological control releases, a multitude of factors
can be varied, some of which may have a dramatic effect on the probability of the agent establishing a selfperpetuating population (Day et al., 2004). Factors that
can be manipulated in biocontrol release programmes
can broadly be divided into the following:
1. Agent stage/quality/release timing: Releases can often be made at a variety of seasons, using a variety
of life history stages of an agent (e.g. eggs, larvae,
pupae and adults for insects, various spore types or
hyphal suspensions for pathogens and releases using
entire plants for pathogens or insects). The number
of individuals, and their genetic make-up, is obviously open to manipulation, but these areas have received sufficient research attention to warrant their
own section below.
2. Habitat/environmental quality: Release sites are se
lected in biocontrol invasions, and apart from the
requirement for the host weed to be present, many
other factors may be important (including some
form of climate match if the weed occupies a wider
range of climatic zones than are optimal for the
agent). Releases may be made initially into cages
or other protected environments. This may be to
protect low initial release numbers from abiotic factors, such as bad weather events, from biotic factors, such as predators or competitors, or simply to
prevent unwanted dispersal of mobile stages. Other
mechanisms to discourage dispersal might involve
supplemental food for agents or treating the target
plants with fertilizer to improve the habitat.
Given the ease with which many of these factors
could be explored experimentally, it is surprising that
so little practical science has been done. However, recent studies have experimentally varied release size
(see section below), and in other cases, the effect of
simulated rain (Norris et al., 2002) or rain protection
(Hill et al., 1993) on releases has been quantified. There
are also examples where different release strategies for
pathogens have been compared (e.g. Sheppard et al.,
1993). Recent research, including quantitative field experimentation, under the CRC for Weed Management
in Australia has been reviewed by Day et al. (2004).
Small populations are perceived as vulnerable to
poor performance that may result in extinction. If this
vulnerability is not a feature of large populations, then
this is often termed an Allee effect, after this issue was
first raised by Allee et al. (1949). Allee effects could
become particularly important if the population size
drops because of say bad weather soon after the release. Allee effects could also cause, or be enhanced by,
a released population that fails to grow rapidly because
of poor individual performance. A corollary of this is
that, for a biocontrol agent to be successful, any poor
performance at low densities will need to be improved
at higher densities. For example, if predation causes
major problems for agent releases, then continued effort with that agent might only be worthwhile if this
predation is suspected to be an Allee effect (and thus
will decline as an issue once populations rise above a
certain threshold size).
With any new agent species, there will be uncertainty over the optimal way to release. Consequently,
mixed strategies are probably best at the early stages
of a release programme. Ideally, the effect on establishment rates of various release strategies will be explored
using rigorous experimental techniques (and/or adaptive management approaches), but political or funding
realities may interfere (McFadyen, 1998). Uncertainty
in optimal release strategies has been dealt with explicitly in modelling optimal release sizes (see below).
Optimal release size
Retrospective analyses of arthropod agents show
that establishment rates improve with increased size
of releases (Hall and Ehler, 1979; Beirne, 1985; Simberloff, 1989; Hopper et al., 1993), but even large releases can be exterminated by unusually severe, but
local, stochastic environmental effects such as floods,
droughts or inadvertent weed control. Therefore, there
is a trade-off between making few large releases, and
risking chance local extinction, and spreading this risk
(over time and/or space) among a larger number of
smaller releases (Grevstad, 1999b). However, small
populations are also clearly at much more risk of extinction from chance events (both environmental and
demographic stochasticity) than larger populations.
Demographic stochasticity refers to chance changes
in say birth and deaths, which are unlikely to cause
extinction in anything other than very small populations. In contrast, Allee effects on small populations
are considered to involve specific mechanisms reducing individual fitness at small population sizes. The
most significant for biological control are likely to be
genetic inbreeding and loss of heterozygosity in small
populations leading to decreased individual fitness (see
the next section) and problems in finding mates when a
population is small. In general, the importance of Allee
effects has only recently been recognized (Courchamp
et al., 1999; Stephens and Sutherland, 1999). Allee ef-
496
Release strategies in weed biocontrol: how well are we doing and is there room for improvement?
fects can be powerful, as demonstrated by deliberate
extinctions of pest species by mass release of sterile
males, which induces major Allee effects in relatively
large populations (Courchamp et al., 1999).
Simulation models (Grevstad, 1999b) suggest that
demographic stochasticity alone is not likely to be important in the establishment of biocontrol agents, while
Allee effects and environmental variability are crucial.
In a variable environment, a large number of very small
releases will maximize the chance of overall establishment because environmental variability reduces the
likelihood of establishment over the entire range of colony sizes (Grevstad, 1999b). In contrast, when an Allee
effect is present (in a constant environment), a single
large release is optimal because the presence of an Allee effect results in a population establishment threshold. For colony sizes below the threshold, a population
will become extinct, while those above the threshold
will establish (Grevstad, 1999b).
Memmott et al. (1998) investigated the relationship
between release size and the probability of establishment of gorse thrips, Sericothrips staphylinus Haliday,
in New Zealand. A higher proportion of small releases
became extinct: Thrips were recovered from 100% of
releases of 270 and 810 thrips but from only 33% of
releases of 10, 30 and 90 thrips. This suggested that
the optimum release size for gorse thrips in New Zealand might be fewer than 100 (Memmott et al., 1998),
in contrast to the previous strategy of 1000 thrips per
release.
In a 5-year experiment in New Zealand, different release sizes of the broom psyllid Arytainilla spartiophila (Frster), were monitored (Memmott et al., 2005).
Local extinction was greatest in the first year, and although the probability of extinction in the first year was
related to release size, several releases of just one pair
of psyllids resulted in established populations. Similar
results were obtained by Grevstad (1999a) using two
chrysomelid beetles in the United States. Thus, releases
of even very small numbers can result in establishment,
albeit with a higher probability of early extinction compared with releases of larger numbers of individuals.
These conflicts were further investigated using a
modelling approach by Shea and Possingham (2000).
They found that it was always possible to find one optimal release size that was better than any mixed release
size strategy but only if the relationship between establishment rate and release size was known. Early in a
release programme, it is better to use a range of release
sizes to help determine the optimum. However, spatial
variability in suitability of release sites, if unknown a
priori, encourages a more risk-adverse release strategy
(i.e. more small releases).
Overall, the issue is of inherent uncertainty, whether
in defining an optimal release size or in the problems
likely to be encountered with spatial variation between
sites or just plain bad luck in the stochastic events that
could cause local extinction of a release. Consequently,
current release strategies for insects in New Zealand
use a range of release sizes early in a programme.
This maximizes efficiency while reducing the risk of
complete failure in case there is a currently unknown
threshold release size below which establishment is
unlikely. As a result, a dataset of release size and establishment success across a range of species is slowly
being accumulated.
Genetics and releases of
biological control agents
There have been several reviews on genetic issues
concerning rearing/culturing and releasing biocontrol
agents, concentrating mainly on arthropods (Mackauer,
1976; Roush, 1990; Hopper et al., 1993; Hufbauer and
Roderick, 2005). Genetic issues, which are acute in
small populations, include the loss of alleles through
genetic drift and inbreeding depression, especially
where the mating of siblings increases the likelihood
of homozygous deleterious recessive alleles. Another
genetic issue, not related to population size, is the possible selection for laboratory-adapted populations, particularly when colonies are maintained for too long in
artificial conditions.
The significance of inbreeding depression in field
populations has been contentious, but in a landmark
study, Nieminen et al. (2001) found that the egghatching rate in laboratory experiments with the butterfly, Melitaea cinxia (L.), was significantly lower in
egg batches laid by inbred females compared to crossbred females. In field studies, using released inbred
and crossbred populations, fewer inbred populations
survived after 1 year (Nieminen et al., 2001).
The only experimental test on how laboratory rearing affected a biocontrol agent that we could find used
the geometrid moth, Chiasmia assimilis (Warren), a
biocontrol agent against Acacia nilotica (L.) in Australia. In this study, Wardill et al. (2004) reared a range of
isofemale lines from several sites. From generation 3,
they mixed the lines and sites, creating several control
lines to mimic the bulk-rearing typical of biocontrol
programmes. At generation 8, they created replicatemixed isofemale lines within and across sites. Genetic
variation was reduced in single isofemale lines in generation 8 compared with generation 1. However, the
within- and across-site hybrid lines maintained high
genetic variability. Fecundity in generations 8 to 11
showed a similar pattern. Thus, hybridizing isofemale
lines appeared to restore genetic variation and fitness
(Hopper at al., 1993). However, bulk-rearing over
four to six generations also maintained genetic variability and fitness, perhaps because of the minimum
population size of 30 females and 30 males.
Overall, the significance of genetic issues in releasing biological control agents remains unresolved.
Crucially, we have little evidence that releases fail to
497
XII International Symposium on Biological Control of Weeds
establish, or do well, because of genetic factors such as
strain/provenance selection, inbreeding depression or
lack of adaptability. Experimental research during the
release and establishment of biological control agents
could offer considerable insights. In the meantime, it
is probably prudent to follow some or all of the guidelines suggested by earlier comprehensive reviews (e.g.
Hopper et al., 1993) such as:
Collect from a wide geographic range and match
weed biotype and agent pathotype/provenance where
known.
Maintain separate lines in laboratory rearing (but
release as hybrid mixtures).
Rear for a limited number of generations in artificial
conditions (e.g. no more than five).
Keep rearing/culturing conditions similar to those in
the field (e.g. fluctuating temperatures) to minimize
adaptation to laboratory conditions.
Avoid genetic bottlenecks or ensure that they are
short (one generation) and then expand the population rapidly to maintain heterozygosity.
Collecting from a range of sites to maximize the
genetic variability of agents has often been used with
arthropod biocontrol agents, but care is needed to ensure that, for example, host range is consistent between
source populations. In contrast, releases of pathogens
for biocontrol of weeds have typically used material
from single lines in part because of unjustified concerns over the potential evolution of expanded host
ranges (Morin et al., 2006b). As a result, the biocontrol
programme against blackberry, Rubus fruticosus L., in
Australia had, until recently, only used rust strains that
were effective against a limited number of biotypes of
the weed (Morin et al., 2006a). Matching weed biotypes and agent pathotypes is clearly a vital issue with
selection of pathogens as weed biocontrol agents (e.g.,
Morin et al., 2006a,b). With both arthropod and pathogens agents, there is a need to show that the tactic of
releasing genetically variable weed biocontrol agents
to encourage adaptation (e.g. to additional biotypes of
a target weed or to a wider climatic range) is not going
to increase the risk of agents expanding their host range
to include non-target plant species. Indeed, a range of
studies outside biological control show that rapid evolutionary changes are possible (Orr and Smith, 1998;
Thompson, 1998). However, the role of adaptation in
the success of biological control agents remains unresolved and open to debate (Hufbauer and Roderick,
2005). From retrospective studies, we can say that expansion of the host range of weed biocontrol agents as
a result of evolutionary adaptation does not seem to
have happened (van Klinken and Edwards, 2002), and,
for example, the substantial surveys to check for attack on non-target plants in New Zealand are showing
that host-range testing procedures dating back to the
late 1920s were generally very reliable (Paynter et al.,
2004; Landcare Research, unpublished data).
What establishment rates are
programmes achieving?
There are different ways of measuring establishment
rates, each subject to biases. For example, the quoted
establishment rates per weed biocontrol agent species of
71% (Julien et al., 1984) is a simple and probably quite
robust statistic, but it ignores whether establishment of
an agent has failed at different sites or within countries.
Rates per agent/country can also be calculated, with
Julien et al. (1984) showing this to be 64%. This takes
into account different experiences in different countries
but again misses establishment failure in areas within
countries. Rates of establishment per release appear
to be seldom known, although the specific release size
experiments discussed earlier are an exception to this.
Syrett et al. (2000a) also report that establishment rates
per release for nine agent releases across a range of
pasture weeds in Australia varied from 12% to 92%.
In New Zealand, it is apparent that the effort put into
monitoring of any given agent can vary through time,
adding variability to overall establishment rates per release. In addition, if some strains were better than the
others or local adaptation was important, then differences in establishment rates might emerge over time
within a biocontrol agent species.
In New Zealand, the overall establishment rate per
species of weed biocontrol agent has increased from a
reported 44% (Cameron et al., 1993) to 76% (Fowler
et al., 2000; Syrett et al., 2000a). Indeed, if the establishment rate is calculated for releases after the nationwide technology transfer programme was set up in the
1990s, then it increases to 95% (Hayes, 2000). Another
region that has invested strongly in release and redistribution strategies is Oregon, USA, where the establishment rate per species is 81% (McEvoy and Coombs,
1999). If pathogens are considered separately, then the
establishment rate is 100% for both New Zealand and
Australia (Morin et al., 2006b). On the basis of these
statistics, it would seem that there is not a great deal of
room for improvement in establishment rates for agent
species for weed biocontrol in areas where substantial
effort is invested into release, redistribution and monitoring. However, failure to establish any agent species
that is released in large numbers can be frustrating. In
New Zealand this has been the case with the sawfly,
Monophadnus spinolae Klug., which is one of the few
adequately host-specific agents available for the serious environmental weed Clematis vitalba L. The species establishment rate can also hide examples where
establishment has been achieved but only at a poor rate
per release: The next section discusses such a case example. Establishment per se is not the aim of weed biocontrol programmes, and it appears that despite high
rates of agent establishment, the proportion of agents
that appear to contribute to weed suppression remains
stubbornly low. For example, McFadyen (1998) gives
498
Release strategies in weed biocontrol: how well are we doing and is there room for improvement?
agent establishment rates of 74% (in programmes that
were rated as successful overall), but a proportion of
agents established and contributing to control of only
55%. In New Zealand, establishment rates for all agents
released exceed 80%, but of those established, we rate
only around a third as contributing to weed suppression
(Landcare Research, unpublished data). A similar pattern is emerging for pathogens, with 100% establishment in Australia, but only 33% of species contributing
to weed control (Morin et al., 2006b).
A case study: heather beetle
releases in New Zealand
Heather, Calluna vulgaris (L.), was deliberately introduced in Tongariro National Park in 1912. It now infests 50,000 ha, including one third of the park. Heather
displaces native vegetation (Williams and Keys, 1994),
and biological control is the only practical management
option in the park. The biocontrol programme (Syrett
et al., 2000b) relied on just one agent species, the heather
beetle Lochmaea suturalis (Thompson), because this
beetle was such a well-studied species, which causes
major damage to valued indigenous heather in Europe
(Cameron et al., 1944; Brunsting, 1982). In 1995, beetles were collected from a range of sites in the UK to
ensure a climatic match to the national park (Emberson,
1986). A debilitating microsporidian disease was commonly found in heather beetles imported from the UK.
To eliminate this disease, beetles in quarantine were
reared from isofemale lines, with lines being destroyed
if sampled offspring were diseased (Smith et al., 1998).
Lines were mixed before main set of 16 releases in the
national park in Summer 1997/98, but two releases in
summer 1995/96 were from one or two lines (second
generation after field collection) each from the single
geographic areas of Oakworth, England and Glencoe,
Scotland. These first two releases were from lines that
were doing particularly well, so beetles were available without disrupting the progress of the main rearing effort. All release sites were sampled intensively
for beetle adults or larvae at 1- to 2-year intervals. No
recoveries were made until summer 1999/2000, when
five adult beetles and 20 larvae were found at Te Piripiri, a site that had received a covering of volcanic ash in
June 1996, 5 months after the release. Despite intensive
collecting efforts, no other recoveries have been made
from any of these releases. Thus, our establishment rate
Table 1.
from releases made from captive-reared heather beetle
imported in 1995 is only 6% (Table 1).
The microsporidian pathogen was detected in one
captive-rearing line in May 1999. All populations of
heather beetle in captivity were then destroyed. At Te
Piripiri, heather beetle numbers increased exponentially from 1999/2000 to 2001/2002 (Fig. 1), causing
severe damage to heather. We urgently redistributed
beetles from this localized, established population and
also set up a secondary captive-rearing facility nearby.
We trialled other release methods, including releasing
larvae (in one case with a release size of 6000) and, in
one case, used a much larger number of releases of just
ten adults per release. We also made three releases in
heather infestations near Rotorua, about 130 km north
of Tongariro National Park, and at an altitude of 400 m
(compared to 6501300 m for sites in and around the
national park).
Unfortunately, the releases in the national park using beetles collected from the one site where beetles
were causing severe damage to heather suffered from
a similar very low establishment rate as the original
releases (Table 2). However, all three releases at Rotorua established and caused severe damage to heather
within 2 years. Varying release parameters (caging,
early vs late summer, using larvae, release size) did not
appear to improve establishment in the national park.
After 2001/2002, the population numbers of beetles at
the Te Piripiri declined markedly (Fig. 1). This collapse
in numbers occurred over an area of about 1 ha and
included outlying heather patches that were not noticeably damaged, so food shortage was not a viable explanation. The population of beetles was being carefully
monitored, and this, together with a set of experimental
studies, showed that predation, parasitism or diseases
could not be implicated in the population collapse (Peterson et al., 2004). Although we had done simple climate matching before collecting and releasing heather
beetle, local weather conditions were the most likely explanation for both the collapse in numbers at Te Piripiri
and the dramatically improved establishment rate at
the lower altitude Rotorua sites. The spring weather in
the year when the beetle populations at Te Piripiri collapsed was unusual, with what appeared to be a relatively warm end to the winter in September, followed
by an exceptionally cold October/November with late
snow. Nearby weather station data confirmed that October 2002 was the coldest since records began 18 years
previously (Peterson et al., 2004). We hypothesized
The first releases of heather beetle in Tongariro National Park, using captive-reared beetles from the UK.
Release season/year
Summer 1995/96
Summer 1997/98
Summer 1998/99
Totals
Number of releases
2
13
2
17
Size of releases
250
100800
250
5700
499
Number of releases established
1
0
0
1 (6%)
XII International Symposium on Biological Control of Weeds
Number of adult beetles
160
140
120
100
80
60
40
20
2005/06
2004/05
2003/04
2002/03
2001/02
2000/01
1999/00
1998/99
1997/08
1996/07
Summer Season
Figure 1.
Table 2.
The numbers of adult heather beetles in a standardized sweep net
sample at Te Piripiri. Numbers increased exponentially from the
first detection in 1999/2000 until 2001/2002 but then crashed, apparently due to winter/spring weather conditions (further details in
text).
The second set of releases of heather beetle in New Zealand, using beetles redistributed and/or captive-reared
from the first establishment site.
Release region
Tongariro
Rotorua
Release season/year
2000/20012005/2006
2000/01
Number of releases
49
3
that beetles emerging from overwintering and their offspring would not be able to survive such major temperature fluctuations. In support, a Danish study reported
that heather beetles had difficulty surviving winters
with fluctuating temperatures, and individuals with
smaller than average body size suffered particularly
high mortality (Jensen and Nielsen, 1985). When we
measured the body size of beetles collected from Tongariro National Park or Rotorua and compared them to
field-collected beetles from a range of sites in the UK,
to our surprise, we found that the New Zealand beetles were significantly smaller than those from the UK
(Peterson et al., 2007). Perhaps, small body size also
increases beetle vulnerability to prolonged winters; furthermore, winters in Tongariro National Park are longer
than those in Rotorua and Oakworth, England where
surviving beetles were sourced from (Peterson et al.,
2007). Currently, we do not know whether this size difference is genetic or phenotypic. If genetic, then it is
likely to be caused by genetic drift (founder effects)
or inbreeding depression. A recent check of rearing
records showed that the Te Piripiri release of 250 adult
heather beetles was of second-generation beetles from
Size of releases
Number of releases established
106000 (total, 7500)
2 (4%)
250 (total, 750)
3 (100%)
just two isofemale lines, and one of these lines only
contributed 29 beetles to the release. Unfortunately,
we do not have preserved specimens of the original females or of their offspring in subsequent generations. If
the size difference is phenotypic, then we have to find
why. One hypothesis is that the leached volcanic soils
typical of heather infestations in North Island, New
Zealand, lead to nutritionally poorer heather for the
beetle. Increased soil nitrogen (e.g. from air pollution)
has been implicated as a factor causing heather beetle
outbreaks in Europe, resulting in conversion of valued
heather heathland into grassland (Heil and Diemont,
1983). As an experimental test of the importance of
nitrogen, we have been releasing heather beetle into
paired plots at Tongariro National Park with or without fertilizer application since Spring 2005. By summer 2006/2007, we had the second outbreak of heather
beetle in Tongariro National Park, which so far has destroyed heather in an area of about 1 ha. It may be a coincidence, but this outbreak started at the first fertilized
release site we set up. This outbreak has spread beyond
the small fertilized patch, so there could a threshold
effect whereby high beetle populations act to increase
500
Release strategies in weed biocontrol: how well are we doing and is there room for improvement?
host plant nutritional quality as suggested for in Salvinia molesta Mitchell biocontrol in the 1980s (Room and
Thomas, 1985). It may also be significant that the original establishment site, Te Piripiri, was unique among
all release sites in receiving a heavy fall of volcanic
ash in the winter after the release, which might have
produced a nutrient flush. It is now over 11 years since
the first releases of heather beetle in New Zealand, and
a multitude of research questions have emerged; plus,
to date, we have only suppressed heather over a minute
percentage of the infested area.
Conclusions
Experimental biocontrol releases are being carried out
in some cases, after a plethora of papers calling for this.
With arthropod agents at least, this has resulted in some
neat practical science to add to the range of retrospective analyses and theoretical models aimed at optimizing release strategies. Multiple strategies to reduce risk
of getting something wrong are the norm with arthropod agents, and commonsense and modelling both suggest this is a sensible approach. However, establishment
success rates per agent species are quite high, so in
terms of getting species established, this research may
not help much (although some failed establishments
can still be very frustrating). The major log-jam appears not to be establishment per se but getting agents
that are effective at weed suppression. The key issues
with release strategies (used in a broad sense to include
provenance/strain and rearing/culturing) are those that
might improve agent performance. These are probably
mostly genetic factors, such as strain/provenance selection, avoiding inbreeding depression, allowing for
sufficient genetic variability to allow for post-release
selection (although its importance has yet to be demonstrated) and ensuring that laboratory adaptation does
not impinge on agent performance. Individual release
programmes will nearly always begin with uncertainty
regarding the best way to make releases, and some of
the scientific challenges are illustrated by the case study
of heather beetle in New Zealand.
References
Allee, W.C., Park, O., Emerson, A.E., Park, T. and Schmidt,
K.P. (1949) Principles of Animal Ecology. Saunders, Philadelphia, USA.
Beirne, B.P. (1985) Avoidable obstacles to colonization in
classical biological control of insects. Canadian Journal
of Zoology 63, 743747.
Brunsting, A.M.H. (1982) The influence of the dynamics
of a population of herbivorous beetles on the development of vegetational patterns in a heathland system. In:
Visser, J.H. and Minks, A.K. (eds) Proceedings of the 5th
International Symposium on InsectPlant Relationships.
Centre for Agricultural Publication and Documentation,
Wageningen, The Netherlands, pp. 215223.
Cameron, A.E., McHardy, J.W. and Bennett, A.H. (1944) The
heather beetle (Lochmaea suturalis). British Field Sports
Society, Petworth, UK, 69 pp.
Cameron, P.J., Hill, R.L., Bain, J. and Thomas, W.P. (1993)
Analysis of importations for biological control of insect
pests and weeds in New Zealand. Biocontrol Science and
Technology 3, 387404.
Courchamp, F., Clutton-Brock, T. and Grenfell, B. (1999) Inverse density dependence and the Allee effect. Trends in
Ecology & Evolution 14, 405410.
Day, M.D., Briese, D.T., Grace, B.S., Holtkamp, R.H., Ireson, J.E., Sheppard, A.W. and Spafford Jacob, H. (2004)
Improving release strategies to increase the establishment
rate of weed biocontrol agents. In: Sindel, B.M. and Johnson, S.B. (eds) Proceedings of the 14th Australian Weeds
Conference. Weed Society of New South Wales, Sydney,
Australia, pp. 369373.
Emberson R. (1986) Factors limiting heather beetle populations. In: Green, P. (ed) Heather Control Workshop, Whakapapa, Mount Ruapehu, Tongariro National Park, 12 May
1986. Unpublished report. Department of Conservation,
Tongariro District, Turangi, New Zealand, 68 pp.
Fowler, S.V., Syrett, P. and Hill, R.L. (2000). Success and
safety in the biological control of weeds in New Zealand.
Austral Ecology 25, 553562.
Grevstad, F.S. (1999a) Experimental invasions using biological control introductions: the influence of release size on
the chance of population establishment. Biological Invasions 1, 313323.
Grevstad, F.S. (1999b) Factors influencing the chance of population establishment: implications for release strategies
in biocontrol. Ecological Applications 9, 14391447.
Hall, R.W. and Ehler, L.E. (1979) Rate of establishment of
natural enemies in classical biological control. Bulletin of
the Entomological Society of America 25, 280282.
Hayes, L.M. (2000) Technology transfer programmes for biological control of weeds the New Zealand experience.
In: Spencer, N.R. (ed) Proceedings of the X International
Symposium on Biological Control of Weeds. Montana
State University, Bozeman, USA, pp. 719727.
Heil, G.W. and Diemont, W.H. (1983) Raised nutrient levels
change heathland into grassland. Vegetatio 53, 113120.
Hill, R.L., Gourlay, A.H. and Winks, C.J. (1993) Choosing
gorse spider mite strains to improve establishment in different climates. In: Prestidge, R.A.(ed) Proceedings of
the 6th Australasian Grassland Invertebrate Ecological
Conference. AgResearch, Hamilton, New Zealand, pp.
377383.
Hopper, K.R. and Roush, R.T. (1993) Mate finding, dispersal,
number released, and the success of biological control introductions. Ecological Entomology 18, 321330.
Hopper, K.R., Roush, R.T. and Powell, W. (1993) Management of genetics of biological control introductions. Annual Review of Entomology 38, 2751.
Hufbauer, R.A. and Roderick, G.K. (2005) Microevolution
in biological control: mechanisms, patterns and processes.
Biological Control 35, 227239.
Jensen, K-M.V. and Nielsen, B.O. (1985) Overvintringsbiologien hos lyngens bladbille (Lochmaea suturalis Thoms.)
(Coleoptera: Chrysomelidae). Flora og Fauna 91, 412.
Julien, M.H., Kerr, J.D. and Chan, R.R. (1984) Biological
control of weeds: an evaluation. Protection Ecology 7,
325.
501
XII International Symposium on Biological Control of Weeds
McEvoy, P.B. and Coombs, E.M. (1999). Biological control
of plant invaders: regional patterns, field experiments, and
structured population models. Ecological Applications 9,
387401.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
Mackauer, M. (1976) Genetic problems in the production of
biological control agents. Annual Review of Entomology
38, 2751.
Memmott, J., Fowler, S.V. and Hill, R.L. (1998) The effect of
release size on the probability of establishment of biological control agents: gorse thrips (Sericothrips staphylinus)
released against gorse (Ulex europaeus) in New Zealand.
Biocontrol Science and Technology 8, 103115.
Memmott, J., Craze, P.G., Harman, H.M., Syrett, P. and
Fowler, S.V. (2005) The effect of propagule size on the
invasion of an alien insect. Journal of Animal Ecology 74,
5062.
Morin, L., Aveyard, R., Batchelor, K.L., Evans, K.J., Hartley,
D. and Jourdan, M. (2006a) Additional strains of Phragmidium violaceum released for the biological control of
blackberry. In: Preston, C., Watts, J.H. and Crossman,
N.D. (eds) Proceedings of the 15th Australian Weeds
Conference. Weed Management Society of South Australia, Adelaide, Australia, pp. 565568.
Morin, L., Evans, K.J. and Sheppard, A.W. (2006b) Selection
of pathogen agents in weed biocontrol: critical issues and
peculiarities in relation to arthropod agents. Australian
Journal of Entomology 45, 349365.
Nieminen, M., Singer, M.C., Fortelius, W., Schps, K. and
Hanski, I. (2001) Experimental confirmation that inbreeding depression increases extinction risk in butterfly populations. The American Naturalist 157, 237244.
Norris, R.J., Memmott, J. and Lovell, D.J. ( 2002) The effect
of rainfall on the survivorship and establishment of a biocontrol agent. Journal of Applied Ecology 39, 226234.
Orr, M.R. and Smith, T.B. (1998) Ecology and speciation.
Trends in Ecology and Evolution 13, 502506.
Paynter, Q.E., Fowler, S.V., Gourlay, A.H., Haines, M.L.,
Harman, H.M., Hona, S.R., Peterson, P.G., Smith, L.A.,
Wilson-Davey, J.R.A., Winks, C.J. and Withers, T.M.
(2004) Safety in New Zealand weed biocontrol: a nationwide survey for impacts on non-target plants. New Zealand Plant Protection 57, 102107.
Peterson, P., Fowler, S.V. and Barrett, P. (2004) Is the poor
establishment and performance of heather beetle in Tongariro national park due to the impact of parasitoids,
predators or disease? New Zealand Plant Protection 57,
8993.
Peterson, P., Fowler, S.V., Harman, H., Barrett, P. and Merrett, M. (2007). Biological control of heather. Unpublished report. Landcare Research, Palmerston North, New
Zealand, 38 pp.
Room, P.M. and Thomas, P.A. (1985). Nitrogen and establishment of a beetle for biological control of the floating
weed Salvinia in Papua New Guinea. Journal of Applied
Ecology 22, 139156.
Roush, R.T. (1990) Genetic variation in natural enemies:
critical issues for colonisation in biological control. In:
MacKauer, M., Ehler, L.E. and Roland, J. (eds) Critical
Issues in Biological Control. Intercept Books, Andover,
UK, pp. 262288.
Shea, K. and Possingham, H.P. (2000) Optimal release strategies for biological control agents: an application of
stochastic dynamic programming to population management. Journal of Applied Ecology 37, 7786.
Sheppard, A.W., Lewis, R.C. and Delfosse, E.S. (1993).
The establishment of Uromyces heliotropii Sred., a biological control agent of Heliotropium europaeum L. In:
Swarbrick, J.T., Henderson, C.W.L., Jettner, R.J., Streit,
L. and Walker, S.R. (eds) Proceedings of the 10th Australian and 14th Asian-Pacific Weed Conference. Weed Society of Queensland, Brisbane, Australia, pp. 8993.
Simberloff, D. (1989). Which insect introductions succeed
and which fail? In: Drake, J.A., Mooney, H.A., di Castri,
F., Groves, R.H., Kruger, F.J., Rejmanek, M. and Williamson, M. (eds) Biological Invasions: A Global Perspective.
Wiley, Chichester, UK, pp. 6175.
Smith, L., Harris, R.J., Peterson, P. and Syrett, P. (1998) Introduction of heather beetle Lochmaea suturalis (Thomson) (Coleoptera: Chrysomelidae) into Tongariro National
Park as a biological control agent for heather Calluna vulgaris (Ericaceae). Landcare Research Contract Report:
LC9798/133. Landcare Research, Lincoln, New Zealand,
26 pp.
Stephens, P.A. and Sutherland, W.J. (1999) Consequences of
the Allee effect for behaviour, ecology and conservation.
Trends in Ecology & Evolution 14, 401405.
Syrett, P., Briese, D.T. and Hoffmann, J.H. (2000a) Success
in biological control of terrestrial weeds by arthropods. In:
Gurr, G. and Wratten, S. (eds) Biological Control: Measures of Success. Kluwer, Dordrecht, The Netherlands, pp.
189230.
Syrett, P., Smith, L.A., Bourner, T.C., Fowler, S.V. and Wilcox, A. (2000b) A European pest to control a New Zealand weed: investigating the safety of heather beetle,
Lochmaea suturalis (Coleoptera: Chrysomelidae) for biological control of heather, Calluna vulgaris. Bulletin of
Entomological Research 90, 169178.
Thompson, J.N. (1998) Rapid evolution as an ecological process. Trends in Ecology and Evolution 13, 329332.
Van Klinken, R.D. and Edwards, O.R. (2002) Is host-specificity
of weed biological control agents likely to evolve rapidly
following establishment? Ecology Letters 5, 590596.
Wardill, T.J., Graham, G.C., Manners, A., Playford, J., Zalucki, M., Palmer, W.A. and Scott, K.D. (2004) Investigating genetic diversity to improve the biological control
process. In: Sindel, B.M and Johnson, S.B. (eds) Proceedings of the 14th Australian Weeds Conference. Weed Society of New South Wales, Sydney, Australia, pp. 364367.
Williams, K. and Keys, H. (1994) Proceedings of the Second
Heather Control Workshop, Turangi, 1921 August 1993.
Unpublished DOC Report. Department of Conservation,
Taupo, New Zealand, 101 pp.
502
Feeding impacts of a leafy spurge
biological control agent on a native plant,
Euphorbia robusta
J.L. Baker and N.A.P. Webber1
Summary
The biological control agent, Aphthona nigriscutis (Foudras), has been established in Fremont County,
WY, since 1992. Near one release site, a mixed stand of leafy spurge, Euphorbia esula (L.), and a
native plant, Euphorbia robusta (Engelm.), was discovered in 1998. During July 1999, A. nigriscutis
was observed feeding on both leafy spurge and E. robusta. Thirty-six E. robusta plants were located
and staked within a 2-ha area, which had a visually estimated leafy spurge canopy of more than
50%. Eighty-eight percent of the E. robusta plants had feeding damage. By August 2001, the leafy
spurge canopy had declined to 6%, and the E. robusta had increased to 450 plants with just 5.7%
having feeding damage. In subsequent years, the data followed the same pattern; however, in 2007,
no feeding damage was observed. At that time, leafy spurge groundcover was just 3%. For the 9-year
period, leafy spurge canopy was inversely correlated to E. robusta density and positively correlated
to A. nigriscutis feeding damage, suggesting that as leafy spurge density declined so did A. nigriscutis
feeding on E. robusta.
Keywords:Aphthona nigriscutis, non-target feeding, Euphorbia esula.
Introduction
Site 1, a parcel of land 5 km southwest of Lander, Fremont County, WY, has been infested with leafy spurge,
Euphorbia esula (L.), for over 30 years. Aphthona nigriscutis (Foudras) was redistributed to this site in 1996
from locally established populations. While monitoring
A. nigriscutis at site 1, we observed a small colony of a
native spurge, Euphorbia robusta (Engelm.)
Early in the leafy spurge biological control effort,
E. robusta had been identified as a species of interest because it is closely related to leafy spurge, both
belonging to the subgenus Esula, is a perennial that
could support the long life cycle of Aphthona beetles
and is sympatric with leafy spurge in North America
(Pemberton, 1985). In 1997 and 1998, we observed
E. robusta plants with feeding scars on the leaves and
occasionally saw A. nigriscutis feeding on the plants. A
Fremont County Weed and Pest Control District, 450 N. 2nd Street,
Room 315, Lander, WY 82520, USA.
Corresponding author: J.L. Baker <[email protected]>.
CAB International 2008
1
few of those plants were marked for future monitoring,
and the next year, the marked plants were gone. This
feeding activity by A. nigriscutis could have been anticipated. An examination of the petitions to introduce
Aphthona flava (Pemberton and Rees, 1990), Aphthona
cyparissiae (Pemberton, 1986) and Aphthona czwalinae (Pemberton, 1987) in the United States showed
that acceptance of E. robusta was almost as high as for
leafy spurge. E. robusta was not used for host testing
A. nigriscutis (Pemberton, 1989).
Early host-plant testing was designed to demonstrate that new biological control of weed agents would
not attack economically valuable crop species. In recent years, testing also encompasses the impacts that
biological agents might cause to native species. Rhinocyllus conicus (Frlich), a biological control agent
for musk thistle, Carduus nutans (L.), has been found
to impact a wide variety of native thistles, some endangered (Gassmann and Louda, 2001). Increased concern has stimulated a call for greater scrutiny of new
biological control agents, more thorough study of the
target species before release and post release tracking
of host range under field conditions (Waage, 2001). It
is in the spirit of post-release evaluation that these data
are offered.
503
XII International Symposium on Biological Control of Weeds
Materials and methods
Between May and August of 1999, site 1 was visited
several times, and E. robusta plants were located,
marked and photographed. The soils there are red
loam, 50 to 150 cm deep. The site slopes 10 to 20 to
the northeast. Average annual precipitation is 33 cm,
although over the last 5 years, rainfall has been 50% to
75% of the normal rate.
The 2-ha study site included 36 E. robusta plants
whose latitude and longitude had been recorded, and
each was numbered using a wooden stake. Leafy spurge
ground cover was estimated in 1999 and 2000. In 2000,
the leafy spurge density was determined by taking
36 samples across the site using a randomly placed
metre-square frame. In 2001, a permanent grid was established across the site at 15-m intervals, and ground
cover for leafy spurge, E. robusta, annual and perennial
grass and forbs, trash and bare ground were sampled
annually at each grid intersection using a point frame
and recording only the first contact with each pin (Levy
and Madden, 1933). Plant density was determined after
centring a metre-square frame over each grid intersection. Sampling dates were timed to capture maximum
vegetation growth in late July to late August. Each
marked E. robusta plant was revisited several times
each summer from 2000 to 2007 and examined for
feeding damage. Newly discovered E. robusta plants
were marked with numbered wooden stakes, and the
latitude and longitude were recorded and added to the
list of plants to monitor.
Leafy spurge and E. robusta plants were dug and
roots inspected for larvae and larval feeding damage.
Ten leafy spurge and two E. robusta plants from near the
study site were inspected in fall of 2000, and 12 plants
per species from inside the site boundaries were inspected in 2001. In 2003, 30 plants per species were
collected, with soil intact and held in cages, in a glass-
Table 1.
Results
Ground cover by plant class has been relatively constant since 2001. Leafy spurge ground cover fell from
50% in 1999 to 10% in 2000, 6% in 2001, and 6%,
9%, 9%, 5%, 2% and 3% through to 2007, respectively.
Leafy spurge density did not show the same declining
trend as leafy spurge groundcover (Table 1). This was
explained by the reduced size in the individual plants.
In 1999, the leafy spurge was 25 to 50 cm tall and heavily branched, while in later years, plants are mostly less
than 20 cm tall, single stemmed and non-flowering.
The E. robusta population increased 15-fold by
2002. The 36 plants marked in 1999 at site 1 increased
to 194 in 2000, 479 in 2001 and 542 in 2002. Thereafter, plant numbers declined to 411 in 2003, 456, 441,
391 and finally 307 in 2007 (Fig. 1). More than 600
plants were marked and evaluated over the study period 1999 to 2007. Of the original 36 plants, 15 (44%)
remained to 2007, although two were dug up for evaluation. Three of the original 36 plants that were missing
in 2000 reappeared in later years. In contrast, the E. robusta population at site 2 comprised 81 plants in 2000,
101 in 2001 and 76 in 2007. Of the original 81 plants,
54% or 67% survived in 2007.
Adult-feeding damage to E. robusta by A. nigriscutis
declined on a percentage basis through the years from
87% in 1999, 48% in 2000, 5% in 2001, 2.6% in 2002
and then to 0%, except in 2006 when 1% (four plants)
lant density of leafy spurge, Euphorbia esula and Euphorbia robusta, and ground cover by plant class and
P
species at site 1.
Year
Density
Leafy
spurge
(per m2)
1999a
2000a,b
2001
2002
2003
2004
2005
2006
2007
house and later inspected for emerging adults, larval
presence and root damage.
For comparison with site 1, a second population of
E. robusta was selected far from either leafy spurge
infestations or A. nigriscutis releases. At this second
site, located 33 km southeast of Lander, WY, E. robusta plants were marked and counted in 2000, 2001
and 2007.
11.9
7.0
9.2
13.8
24.8
10.4
6.3
7.0
Euphorbia
robusta
(per m2)
0.05
0.07
1.50
1.50
0.09
0.09
0.04
Percent Ground Cover
Perennial
grass
51.0%
45.4%
46.1%
45.7%
51.3%
48.6%
38.2%
Shrub
5.0%
8.2%
7.1%
8.9%
8.9%
7.3%
12.1%
Perennial Annual Annual
forb
grass
forb
2.0%
2.5%
2.7%
2.7%
2.5%
1.6%
3.9%
Ground cover visually estimated.
Density determined by randomly placed m2 frame.
504
1.0%
0.7%
0.9%
0.0%
0.4%
0.4%
1.8%
0.0%
0.0%
1.4%
1.4%
3.2%
0.0%
3.0%
Trash
14.0%
17.9%
8.2%
12.5%
6.4%
16.8%
17.1%
Bare
Leafy Euphorbia
ground spurge
robusta
21.0%
19.6%
25.0%
20.2%
22.3%
24.0%
21.0%
50%
10%
6.0%
5.7%
8.6%
8.8%
4.8%
1.8%
2.9%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Feeding impacts of a leafy spurge biological control agent on a native plant, Euphorbia robusta
Figure 1.
The changes over time for leafy spurge ground cover (%), Euphorbia robusta population size, feeding incidence (%) and number of E. robusta plants with feeding.
had minor feeding. An increase in actual numbers of
plants with feeding temporarily followed the increase
in total plant numbers but essentially fell to zero after
2002 with 2006 being the only exception (Fig. 1).
No larval feeding by A. nigriscutis on E. robusta
was observed. In the fall of 2000, larvae could be found
on the roots of three out of ten leafy spurge plants dug
up at site 1. None were found on two E. robusta plants
dug near this study site. In 2001, ten more plants for
each species were dug up. Two of the E. robusta were
plants with recorded adult feeding that year. No larvae
were found on either species. In 2003, of the 30 plants
of each species that were removed to the glasshouse
with soil intact, only the leafy spurge produced any
A. nigriscutis (Wacker and Butler, 2006).
Discussion
As the E. robusta population increased over time, the
number of plants with feeding damage decreased in numbers and percentages (Fig. 1). It appears that E. robusta
was competitively suppressed by leafy spurge because,
over the same period while the leafy spurge groundcover
declined, the E. robusta population increased.
The feeding damage on E. robusta lagged a year behind the groundcover decline, suggesting that the adult
feeding by A. nigriscutis in 2000 was more closely related to the leafy spurge ground cover in 1999 than in
2000 (Figure 1). The population at site 2 remained relatively constant over the same period with a similar turn
over in plants as site 1, suggesting that E. robusta is
a short-lived perennial with regular death and recruitment. Only where there was a decline in competition
from the leafy spurge was there an increase in population recruitment.
Even though host-specificity testing predicted that
E. robusta should be a good host for the Aphthona
beetles, with 80% to 100% feeding acceptance and
completed reproduction in the laboratory (Pemberton,
1986, 1987, 1989), observations at site 1 indicate that
A. nigriscutis only fed heavily on E. robusta when its
primary host leafy spurge was plentiful and able to
support the biological control agent in large numbers.
Even with a 15-fold increase, E. robusta did not show
up in the ground cover measurements nor did it fill the
space vacated by leafy spurge (Table 1).
It is not known if A. nigriscutis can complete its life
cycle on E. robusta in the field. However, the strong
correlation between the decline in leafy spurge with the
decline in beetle damage to E. robusta and the absence
of A. nigriscutis larvae on the E. robusta roots suggests
that we observed a transient adult-feeding effect. A
comparison of root morphology could explain this. The
leafy spurge root is smaller in diameter and has more
root hairs close to the surface, which may provide a
better food source for developing larvae than E. robusta
(Wacker and Butler, 2006). If E. robusta was a good
developmental host for the beetle, then it would have
been unlikely for the adult feeding to decline and the
density of the E. robusta plants to increase with leafy
spurge decline (R.W. Pemberton, 2003, personal communication).
This is in keeping with observations made in 1998
and 2001 at Camels Hump west of Medora, ND. In
1998, this site was heavily infested with leafy spurge,
which supported populations of A. nigriscutis and Aphthona lacertosa (Rosenhauer). The insects were superabundant, and millions were collected for redistribution
in just a few hours. Every blade of grass had notches
in the leaves, and the insects could be observed feeding
505
XII International Symposium on Biological Control of Weeds
on every plant species present. By 2001, leafy spurge
was nearly gone. At that time, no Aphthona sp. was observed feeding activity on any species other than the few
remaining leafy spurge plants. Waage (2001) reported
two parallel occurrences where weed biological control
agents attacked non-target species during the epidemic
period of agent development when the host plants were
abundant. A lace bug, Teleonemia scrupulosa, released
against Lantana camara in sesame crops in Uganda attacked the crop at peak populations (Davies and Greathead, 1967), and a leaf beetle, Zygogramma bicolorata,
released against Parthenium hysterophorum, attacked
sunflowers in India during population explosions (Jayanth et al., 1993). In both cases, a decline in host-plant
numbers resulted in a decline in the biological control
agent and the non-target feeding stopped (Davies and
Greathead, 1967; Jayanth et al., 1993).
The Aphthona beetles are proving to be excellent
biological control agents that severely impact their target weed, leafy spurge, in the United States (Nowierski
and Pemberton, 2002). Their reputation can only be
enhanced by these recently observed modest transient
effects on their most likely non-target host, E. robusta.
Acknowledgements
I thank Drs Robert Pemberton, Mic Julien and Joseph
DiTomaso for their guidance and insight in preparing
this manuscript.
References
Davies, J.C. and Greathead, D.J. (1967) Occurrence of Teleonemia scrupulosa on Seasamum indicum Lin. in Uganda.
Nature 2123, 102103
Gassmann, A. and Louda, S.M. (2001) Rhinocyllus conicus: intial evaluation and subsequent ecological impacts in North
America. In: Wajnberg, E., Scott, J.K. and Quimby, P.C.
(eds) Evaluating Indirect Ecological Effects of Biological
Control. CABI Publishing, Wallingford, UK, pp. 147176.
Jayanth, K.P., Mohandus, S., Asokan, R. and Visalakshy,
P.N.G. (1993) Parthenium pollen induced feeding by Zy-
gogramma bicolorata (Coleoptera:Chrysomelidae) on
sunflower (Helianthus annuus) (Compositae). Bulletin of
Entomological Research 83, 595598.
Levy, E.B. and Madden, E.A. (1933) The point method for
pasture analysis. New Zealand Journal of Agriculture 46,
267279.
Nowierski, R.M. and Pemberton, R.W. (2002) Leafy spurge
(Euphorbia esula L.). In: Van Driesche, R., Blossey, B.,
Hoddle, M., Lyon, S. and Reardon, R. (eds) Biological
Control of Invasive Plants in the Eastern United State.
US Forest Service Forest Health Technology Enterprise
Team-200204, Morgantown, WV, USA, pp. 181207.
Pemberton, R.W. (1985) Native plant considerations in biological control of leafy spurge. In: Delfosse, E.S. (ed.)
Proceedings of the VI International Symposium on Biological Control of Weeds. Agriculture Canada, Ottawa,
pp. 5771.
Pemberton, R.W. (1986) Petition for the release of Aphthona
cyparissiae, a new flea beetle for leafy spurge control in
the United States. In: Spencer, N. (ed.) Purge Spurge Database. USDA/ARS, Sidney, MT, USA.
Pemberton, R.W. (1987) Petition for the release of Aphthona czwalinae Weise against leafy spurge (Euphorbia
esula L.) in the United States. In: Spencer, N. (ed.) Purge
Spurge Database. USDA/ARS, Sidney, MT, USA.
Pemberton, R.W. (1989) Petition to introduce Aphthona nigriscutis Foudras (Chrysomelidae) to the United States
for leafy spurge (Euphorbia esula L.) control. In: Spencer, N. (ed.) Purge Spurge Database. USDA/ARS, Sidney, MT, USA.
Pemberton, R.W. and Rees, N.E. (1990) Host specificity and
establishment of Aphthona flava Guill. (Chrysomelidae), a
biological control agent for leafy spurge (Euphorbia esula
L.) in the United States. Proceedings of the Entomological Society of Washington 92, 351357 [Reproduced in:
Spencer, N. (ed.) Purge Spurge Database. USDA/ARS,
Sidney, MT, USA].
Waage, J.K. (2001) Indirect ecological effects in biological
control: the challenge and the opportunity. In: Wajnberg,
E., Scott, J.K. and Quimby, P.C. (eds) Evaluating Indirect
Ecological Effects of Biological Control. CABI Publishing, Wallingford, UK, pp. 111.
Wacker, S.D. and Butler, J.L. (2006) Potential impact of two
Aphthona spp. on a native, nontarget Euphorbia species.
Rangeland Ecology & Management 59, 468474.
506
Variation in the efficacy of a mycoherbicide
and two synthetic herbicide alternatives
G.W. Bourdt, G.A. Hurrell and D.J. Saville1
Summary
Field experiments testing the efficacy of plant pathogenic fungi as inundative biological control agents
usually reveal a wide spatial and temporal variation in the response of the target weed population.
Such variation is typically thought to be much greater than that of synthetic herbicides. This perception may impede the commercial development of these fungi as mycoherbicides. In this study, we
subjected this notion to an objective test using data from a total of 75 independent field experiments
that quantified the responses of dairy pasture populations of giant buttercup, Ranunculus acris L., in
New Zealand to a novel mycoherbicide utilizing the plant pathogen, Sclerotinia sclerotiorum (Lib.)
de Bary, or to two synthetic phenoxy herbicides, 2-methyl-4-chlorophenoxyacetic acid (MCPA) and
4-chloro-2-methylphenoxybutyric acid (MCPB). After excluding extreme and/or unreliable response
estimates (one from 29 MCPA estimates, one from 14 MCPB estimates, and six from 32 S. sclerotiorum estimates), the mean percent reduction (standard deviation) of the R. acris was 64% (23), 39%
(22) and 51% (20) for MCPA, MCPB and the S. sclerotiorum mycoherbicide, respectively. While
these data indicate that the MCPA is on average more effective than both MCPB and the mycoherbicide, they provide no support for the notion that these two synthetic herbicides are less variable in
efficacy than the mycoherbicide.
Keywords: biocontrol, Sclerotinia sclerotiorum, Ranunculus acris, weed control.
Introduction
Plant pathogenic fungi used as mycoherbicides typically
result in spatially and temporally variable reductions
in their target weed populations. This phenomenon is
well illustrated by experiments with Sclerotinia sclerotiorum (Lib.) de Bary (Brosten and Sands, 1986; Hurrell et al., 2001) and, in the case of this and of other
pathogens, is an impediment to their commercial development as mycoherbicides. This is particularly so
when synthetic herbicide alternatives are available in
the market place that are perceived to be more reliably
efficacious. This problem has characterized and delayed the development of S. sclerotiorum as a mycoherbicide in New Zealand.
In this paper, we explore how well our perception,
that the synthetic herbicides marketed for the selec-
tive control of Ranunculus acris L. (giant buttercup)
in dairy pastures are generally more reliably efficacious than applications of S. sclerotiorum, matches
up to the empirical evidence. We do this by comparing
the variability and mean response of field populations
of the weed to the two phenoxy herbicides, 2-methyl4-chlorophenoxyacetic acid (MCPA) and 4-chloro2-methylphenoxybutyric acid (MCPB), in published
experiments, with the variability and mean response
to a novel formulation of S. sclerotiorum in a recently
completed series of experiments. The two phenoxy
herbicides were chosen for the comparison because
they are the only ones for which data from individual
experiments have been published; average responses
only have been published for the two other herbicides
registered for use against R. acris (thifensulfuron and
flumetsulam; Lamoureaux and Bourdt, 2007).
Methods and materials
AgResearch Lincoln, Private Bag 4749, Christchurch, New Zealand
<[email protected], [email protected],
[email protected]>.
Corresponding author: G.W. Bourdt <graeme.bourdot@agresearch.
co.nz>.
CAB International 2008
1
Response of R. acris to MCPA and
MCPB in historical experiments
In this paper, we use the responses of R. acris to
MCPA in 29 published experimental tests (Tuckett,
507
XII International Symposium on Biological Control of Weeds
1961; Popay et al., 1984, 1989; Bourdt and Hurrell,
1990; Butcher et al., 1993; Sanders et al., 1994; Harris
and Husband, 1997) and to MCPB in 14 experimental
tests (Thompson, 1983; Popay et al., 1989; Bourdt and
Hurrell 1990; Harris and Husband 1997) as reported in
a wider study of the effectiveness of herbicides against
R. acris in dairy pastures in New Zealand (Lamoureaux
and Bourdt, 2007). The percentage reduction in the
measured response variable (either number of plants
or percent ground cover) was derived from the mean
responses (across field replicates) given in tabular or
graphical form by the authors. Mean responses per experiment were used, as we were interested in the variability between, not within, experiments. We restricted
our attention to applications made by the authors during the time of year recommended on the product labels
[AugustNovember (late winterlate spring) for MCPA
and SeptemberDecember (early springearly summer)
for MCPB]. This gave 29 and 14 individual tests for
MCPA and MCPB, respectively, with mean application rates of 1.2 kg ai/ha for MCPA and 1.5 kg ai/ha for
MCPB and mean assessment times of 25 and 23 weeks,
respectively, after application. The mean application
rates are lower than the current mid-point rates on the
labels, 1.875 and 2.310 kg ai/ha for MCPA and MCPB,
respectively (Young, 2008), but are likely to approximate the rate typically applied to dairy pastures.
Response of R. acris to a
S. sclerotiorum mycoherbicide
A series of experiments was conducted during 2004
to 2005 in which the fungal pathogen, S. sclerotiorum,
was applied to small plots in R. acris-infested dairy
pastures using a precision granule spreader (with infested farms randomly selected using standard stratified survey methodology). Three formulations were
compared, but we consider only the most effective of
them here: NF361, a dry polymer-coated flake containing mycelium adhering to spent barley (brewers
waste). There were 33 sites in total, 23 in the Golden
Bay region and ten in the Taranaki region. Half of the
sites were treated in early November 2004 (12 and 5 in
Golden Bay and Taranaki, respectively) and half in
mid-late November 2004 (11 and 5 in Golden Bay and
Taranaki, respectively). There were two non-treated
control plots at each site. The percentage of the pasture
surface covered by R. acris was measured on each of
three occasions (8, 14 and 20 weeks after application)
using laser-assisted point analysis. In this paper, we report only the data recorded 20 weeks after application
for a more valid comparison with the synthetic herbicides. The data for this assessment was missing for one
of the sites where the farmer had sprayed the paddock
with glyphosate before the assessment. For each of the
remaining 32 sites, the percentage reduction in cover of
R. acris due to the mycoherbicide was calculated using
a BeforeAfter, ControlImpact calculation (Green,
1979). In this way, the impact of the mycoherbicide was
assessed by comparing the changes in cover on treated
and control plots between the assessments before and
after treatment impact. The percentage reduction
100 - [(At/Bt)/(Ac/Bc) 100]
was calculated for each site. Here, B and A are R. acris
groundcover at the time of treatment (before the impact) and after treatment (after impact), respectively.
Subscripts t and c denote treated and control plots, respectively. This improvement in calculating the weeds
response was not possible in the case of the historical
data for MCPA and MCPB because the before impact
values were typically not reported.
Statistical analysis
The mean percent reduction in R. acris, standard deviation and 95% confidence interval for the mean were
calculated for each of MCPA, MCPB and the mycoherbicide. In addition, normal distribution curves were
fitted to the means and standard deviations to illustrate
the variation in the effects of the three herbicides. The
three mean percent reductions were statistically compared using analysis of variance, and the standard deviations were compared using an F test.
Results and discussion
Histograms of the percentage reductions of R. acris
populations caused by the synthetic herbicides, MCPA
and MCPB, and the mycoherbicide utilizing S. sclerotiorum are given in Fig. 1. The median percentage reduction for MCPA of 69.0 was substantially higher than for
both MCPB and S. sclerotiorum (39.4 and 45.2, respectively; Table 1).
An alternative, more traditional statistical analysis
of these data, based on the normal distribution, is problematical because of the occurrence of several extreme
negative values. In one experiment involving MCPA
and MCPB (Bourdt and Hurrell, 1990), the control
plot mean groundcover was exceptionally low, and this
resulted in negative percentage reductions of -55% and
-177%, respectively (Fig. 1). We omitted these values
from the statistical analysis. Another problem was that,
at four of the 32 sites involving S. sclerotiorum (two in
each of Taranaki and Takaka), the initial cover of R. acris was very low (<6%) on the treated plot, resulting in
unreliable estimates of the percent reduction (-367%,
-314%, 45% and 100%). These data were also omitted
from the statistical analysis. In addition, we omitted extreme values from two other sites involving S. sclerotiorum (-61% and -70%). In summary, we omitted all
six negative responses and two positive responses.
It is interesting to note that, out of the total number of 75 experiments, we alone reported negative estimates (6 out of a total of 52); no negative estimates
were reported by the other authors (in 23 experiments).
508
Variation in the efficacy of a mycoherbicide and two synthetic herbicide alternatives
Table 1.
ercentage reductions in groundcover of Ranunculus acris in dairy pastures in New Zealand resulting from appliP
cations of either of two phenoxy herbicides, MCPA (at an average of 1.2 kg/ha) or MCPB (average of 1.5 kg/ha)
or a novel Sclerotinia sclerotiorum-based mycoherbicide (at 100 kg/ha).
Sample size
MCPA
MCPB
Sclerotinia
sclerotiorum
29
14
32
Mediana
69.0
39.4
45.2
Statistical analysis using reduced samplesb
Reduced
sample size
28
13
26
Mean
63.6
38.7
51.0
Standard
deviation
23.3
21.9
19.8
95% confidence limits
for the mean
63.6 9.0
38.7 13.2
51.0 8.0
Medians were calculated using the full sample.
Sites were excluded from the statistical analysis if the initial cover of R. acris was very low (<6%) on either the treated or control plots, as
these sites yielded unreliable estimates of the percent reduction. One such site involved MCPA and MCPB, and four sites involved S. sclerotiorum. Two other sites (involving S. sclerotiorum) giving unusually low estimates were also excluded.
Figure 1.
Histograms of the percentage reductions of Ranunculus acris populations
caused by the synthetic herbicides MCPA and MCPB and the mycoherbicide
utilizing Sclerotinia sclerotiorum (using size classes of 25% reduction). For
MCPA and MCPB, these histograms display the data summarized in Table
1 of Lamoureaux and Bourdt (2007). For S. sclerotiorum, the data have not
been reported elsewhere. Note that the bottom class includes all values that
are less than -100. Also displayed are normal curves using the means and
standard deviations (based on reduced samples) given in Table 1.
509
XII International Symposium on Biological Control of Weeds
We hypothesize that this is an example of the wellknown phenomenon of publication bias (Rothstein
et al., 2005), with negative estimates less likely to
be published. In the present paper, our own research
yielded all of the S. sclerotiorum estimates, 10 out of
29 MCPA estimates and 10 out of 14 MCPB estimates.
As it seems likely that negative estimates for MCPA
and MCPB have occurred in field experiments but not
been published, the MCPA and MCPB histograms in
Fig. 1 probably convey an overly favourable picture
of the efficacy of these herbicides, while the S. sclerotiorum estimates, having been fully reported, convey a
more realistic picture. As a consequence, we consider
that our omission of the two negative values, -61% and
-70%, from the S. sclerotiorum data leads to a fairer
comparison with the synthetic herbicides.
Using the reduced data sets, the mean percentage reduction for MCPA of 63.6 was again substantially and
significantly higher than for MCPB (38.7, P < 0.01)
and S. sclerotiorum (51.0, P < 0.05; see means and
their 95% confidence limits in Table 1). This is perhaps
why MCPA is considered by dairy farmers to be a better choice than MCPB for the control of R. acris. The
mean percent reduction for S. sclerotiorum was greater
than for MCPB, but the difference was not statistically
significant.
Interestingly, and contrary to popular perception, the
S. sclerotiorum mycoherbicide was no more variable in
its efficacy than the two synthetic herbicides (F = 1.38,
P > 0.05; also see the standard deviations in Table 1
and the normal curves in Fig. 1). However, overall,
variability was high. The causes of this wide variation
in results for all three treatments are not evident from
the experiments. Possible causes are (1) variation in
weather conditions after application, (2) errors in application rates, (3) variation between experiments in
the level of herbicide resistance (Bourdt et al., 1990)
and resistance to S. sclerotiorum (Pottinger, 2006) and
(4) variation in plant regeneration (from rhizome buds)
and seedling recruitment after spraying (Brown, 1993;
W. Brown, 1993, unpublished results).
Conclusions and outlook
In conclusion, the mycoherbicide, utilizing the fungus
S. sclerotiorum, would, if commercially available, be
almost as effective against R. acris in New Zealand
dairy pastures as the synthetic herbicide, MCPA (51%
cf. 64% reduction, Table 1). This conclusion rests upon
the assumption that resistance in R. acris to MCPA has
not changed in intensity or spatial extent in New Zealand since the experiments we have reviewed in this paper were conducted. Our comparison also showed that,
contrary to common perception, MCPA and MCPB
are not necessarily more reliable against R. acris than
S. sclerotiorum. Further testing of these preliminary results could be conducted through experiments that include both MCPA (and other synthetic herbicides) and
S. sclerotiorum treatments. Given the wide variation in
response that can be expected from both synthetic and
biological control options (Fig. 1) and considering that
the drivers of these variations are probably different
between the synthetic and biological control options, it
would be necessary to repeat such comparative experiments in both time and space. Experiments to determine
the causes of the variation could, furthermore, lead to
improvements in the effective use of both the synthetic
and biological control alternatives and help determine
the specific environmental conditions and market niches,
which best suit the alternative approaches.
Acknowledgements
We acknowledge the financial support provided by
Dairy InSight New Zealand Limited for the experiments conducted with the mycoherbicide in Golden
Bay and Taranaki, New Zealand, during 20042005,
and the cooperation of the dairy farmers who provided
sites for the experiments. We also thank the biosecurity staff of the Tasman District and Taranaki Regional
Councils for providing comprehensive lists of R. acrisinfested sites to serve as study populations for the
series of experiments involving the S. sclerotiorum
mycoherbicide.
References
Bourdt, G.W. and Hurrell, G.A. (1990) Effects of annual
treatments of MCPA and MCPB on giant buttercup (Ranunculus acris L.) in dairy pastures. In: Popay, A.J. (ed)
Proceedings of the 43rd New Zealand Weed and Pest
Control Conference, vol. 43. New Zealand Weed and Pest
Control Society, Pacific Park Hotel, Dunedin, New Zealand, pp. 233236.
Bourdt, G.W., Hurrell, G.A., and Saville, D.J. (1990) Variation in MCPA-resistance in Ranunculus acris L. subsp. acris and its correlation with historical exposure to MCPA.
Weed Research 30, 449457.
Brosten, B.S. and Sands, D.C. (1986) Field trials of Sclerotinia sclerotiorum to control Canada thistle (Cirsium arvense). Weed Science 34, 377380.
Brown, W. (1993) Giant buttercup poster paper. Massey
Dairy Farming Annual 1993, 186190.
Butcher, M.R., Strachan, C.M., and Field, R.J. (1993) Giant
buttercup (Ranunculus acris) control on Golden Bay dairy
farms. Massey Dairy Farming Annual, 1993, 93103.
Green, R. (1979) Sampling Design and Statistical Methods
for Environmental Biologists. Wiley, New York, 257 pp.
Harris, B.M. and Husband, B.M. (1997) Flumetsulam for
control of giant buttercup in pasture. New Zealand Plant
Protection 50, 472476.
Hurrell, G.A., Bourdt, G.W., and Saville, D. (2001) Effect
of application time on the efficacy of Sclerotinia sclerotiorum as a mycoherbicide for Cirsium arvense control in
pasture. Biocontrol Science and Technology 11, 317330.
Lamoureaux, S. and Bourdt, G.W. (2007) A review of the
ecology and management of Ranunculus acris L. in pasture. Weed Research 47, 112.
510
Variation in the efficacy of a mycoherbicide and two synthetic herbicide alternatives
Popay, A.I., Edmonds, D.K., Lyttle, L.A., and Phung, H.T.
(1984) Timing of MCPA applications for control of giant
buttercup. In: Hartley, M.J. (eds) Proceedings of the 37th
New Zealand Weed and Pest Control Conference. New
Zealand Weed and Pest Control Society, Russley Hotel,
Christchurch, New Zealand, pp. 1719.
Popay, A.I., Edmonds, D.K., Lyttle, L.A., and Phung, H.T.
(1989) Chemical control of giant buttercup (Ranunculus
acris L.). New Zealand Journal of Agricultural Research
32, 299303.
Pottinger, B.M. (2006) Determining the key pathogenicity
factors in Sclerotinia sclerotiorum to improve its potential
as a mycoherbicide. PhD thesis, Lincoln University, Canterbury, New Zealand.
Rothstein, H., Sutton, A.J., and Borenstein, M., eds. (2005)
Publication Bias in Meta-analysis: Prevention, Assessment and Adjustments. Wiley, New York, 374 pp.
Sanders, P., Rahman, A., and Popay, A.J. (1994) Evaluation
of thifensulfuron for control of some pasture weeds. In:
Popay, A.J. (ed) Proceedings of 47th New Zealand Plant
Protection Conference. The New Zealand Plant Protection Society, Waitangi, New Zealand, pp. 6267.
Thompson, A. (1983) Pasture weed control by rope wick applicator. In: Hartley, M.J. (ed) Proceedings of the 36th
New Zealand Weed and Pest Control Conference. The
New Zealand Plant Protection Society, Rotorua, New
Zealand, pp. 9698.
Tuckett, A.J. (1961) Giant buttercup. In: Hartley, M.J. (ed)
14th New Zealand Weed and Pest Control Conference.
New Zealand Weed and Pest Control Society, War Memorial Hall, New Plymouth, New Zealand, pp. 124
126.
Young, S. (ed) (2008). New Zealand Novachem Agrichemical
Manual. Agrimedia Limited, Christchurch (652 p).
511
Ten years after the release of the water
hyacinth mirid Eccritotarsus catarinensis
in South Africa: what have we learnt?
J.A. Coetzee,1,2 M.P. Hill1 and M.J. Byrne2
Summary
Water hyacinth, Eichhornia crassipes (Mart.) Solms, is the worst aquatic weed in South Africa, and
biological control offers the most sustainable control option. The mirid, Eccritotarsus catarinensis
(Carvalho) (Hemiptera: Miridae), was released against water hyacinth in South Africa in 1996 and
shown to be damaging to the plant and host-specific within the Pontederiacae. Feeding, oviposition and nymphal development were recorded on pickerelweed, Pontederia cordata L., an important
aquatic plant in North America but a potential weed in South Africa. The release of this agent allowed
us to test in the field that pickerelweed was not part of the mirids realized host range. The agent
subsequently established at 15 sites around South Africa, including those where climatic modeling
had indicated that it would not due to low winter temperatures, calling into question the usefulness
of climate-matching techniques in the absence of microclimate and behavioural data. Hypertrophic
nutrient conditions also reduced the effectiveness of E. catarinensis due to rapid proliferation of the
plant, but the mirid reduced both the vigour and competitive ability of water hyacinth in mesotrophic
and eutrophic water. E. catarinensis is emerging as an effective agent in areas of medium to lownutrient status with a warm climate and should be considered for release in other areas of the world,
particularly Africa, where few Pontederiaceae occur. This programme shows the value of considering
fundamental vs realized host ranges but suggests that more data are needed to increase confidence in
climate compatibility predictions.
Keywords: host specificity, realized host range, climate-matching, post-release evaluation, agent impact.
Introduction
Water hyacinth, Eichhornia crassipes (Mart.) Solms,
remains the worlds worst aquatic weed, even though
up to seven biological control agents have been released
against it in at least 30 countries (Julien and Griffiths,
1998). The effects of these agents are spatially and temporally variable such that water hyacinth still causes
problems in many regions, including South Africa (Hill
and Olckers, 2000). One of the newer agents against
water hyacinth is the mirid, Eccritotarsus catarinensis
(Carvalho) (Hemiptera: Miridae), which was screened
Rhodes University, Department of Zoology and Entomology, P.O. Box
94, Grahamstown, 6140, South Africa.
2
University of the Witwatersrand, School of Animal, Plant and Environmental Sciences, Private Bag X3, Wits 2050, South Africa.
Corresponding author: J.A. Coetzee <[email protected]>.
CAB International 2008
1
and released in South Africa in 1996 as a new natural
enemy of water hyacinth (Hill et al., 1999).
Hill et al. (1999) found that E. catarinensis had potential as a control agent of water hyacinth in South
Africa due to its host specificity within the Pontederiaceae and because it has long-lived, mobile adults that
are obviously damaging to the plant the four nymphal
instars and the adults feed gregariously, resulting in
chlorosis and ultimately death of the leaves (Hill et al.,
1999). Since 1996, the mirid has been released at least
18 sites in South Africa (Hill et al., 1999) and has established at 15.
In this paper, we review the results of the last 10 years
of research since the mirids release. We have conducted
a range of laboratory and field experiments to (1) further
evaluate the realized host range of the mirid, (2) determine its thermal physiology and potential distribution
in South Africa and (3) assess the impact it is likely to
have on water hyacinth. Based on these findings, we
512
Ten years after the release of the water hyacinth mirid Eccritotarsus catarinensis in South Africa
Host specificity
Before the release of E. catarinensis, host-specificity
trials demonstrated that pickerelweed, Pontederia cordata L., an important, native, littoral plant of waterways
in the United States, may be at risk because feeding,
oviposition and nymphal development were recorded
on it in the laboratory (Hill et al., 1999). This did not
prevent the release of the mirid in South Africa, as
pickerelweed is neither indigenous nor economically
important.
Hill et al. (2000) predicted that the results of the
laboratory host-specificity tests were indicative of an
artificially expanded host range, and despite feeding on
non-target pickerelweed under laboratory conditions,
the mirid would have minimal, if any, non-target effects
on this species in the field, where natural host-selection
cues would prevail. Since E. catarinensis had already
been released in South Africa, we were presented with
an ideal opportunity to test its realized host range in the
field. We first attempted to force the mirid to establish
on pickerelweed plants in the absence of water hyacinth by sleeving them onto leaves to prevent their initial dispersal (Coetzee et al., 2003). The mirids fed on
the leaves and produced offspring within the sleeves.
Once the sleeves were removed after 5 weeks, the pickerelweed was monitored for establishment of the mirid.
Under field conditions, E. catarinensis did not sustain
a population on pickerelweed in the absence of water
hyacinth (Coetzee et al., 2003).
We also conducted choice tests in the field by placing
pickerelweed plants among water hyacinth plants in a
heavily infested river that had a large, well-established
population of mirids. Monitoring of the pickerelweed
plants showed that, although feeding damage was evident, it was far less than on water hyacinth and was
indicative of spillover feeding damage because of the
high mirid population levels (Hill et al., 2000).
Therefore, the prediction of Hill et al. (1999) that,
under restricted laboratory conditions, pickerelweed
was a more suitable host for E. catarinensis than under
field conditions, was correct.
Thermal physiology
In South Africa, where at least five biological control
agents have been released, water hyacinth control is not
as successful as that in tropical areas (Hill and Olckers,
2000), and it was assumed that low winter temperatures
play a crucial role in the successful control of water
hyacinth in South Africa. Many of the worst water hyacinth infestations in South Africa occur at high-altitude
sites that are typified by cold winters (Hill and Olckers,
2000). At these high elevations, water hyacinth infesta-
tions are subject to frost and winter dieback. Biological control in these areas is not as successful as that in
frost-free areas because, in colder areas, regrowth of
water hyacinth occurs during spring, whereas the insect
populations only reach significant levels during midsummer (Hill and Cilliers, 1999). This lag period may
allow the plant populations to increase unchecked and
could be responsible for the variable results achieved
by water hyacinth biological control agents in these
regions.
When biological control agents are released into a
new country, they should ideally be species or strains
from a climatically matched area (Williamson, 1996).
We therefore investigated various aspects of the mirids
thermal physiology to determine whether it might be
limited by cold winter temperatures in South Africa.
We determined the critical thermal minimum (CTMin,
a point short of death where locomotory impairment
occurs, but from which recovery is possible) of E. catarinensis to be 1.2C and the lower lethal limit (LT50,
the temperature at which 50% of the population dies)
to be -3.5C (Coetzee et al., 2007a). Neither of these
limits is particularly low, and they might prevent the
mirid from establishing in areas that receive considerable winter frost.
Another method available for climate matching
is degree-day modeling, which uses temperature and
time to predict the number of generations that an insect can complete at a given locality. We calculated that
the mirids thermal constant K was 342-degree days,
above a developmental threshold t of 10.2C (Coetzee
et al., 2007a). These values were then used to calculate
accumulated degree days according to the methods of
Campbell et al. (1974) for 128 South African localities,
using the equation:
K=
(Tmax + Tmin)
S{
were able to predict what might occur in the field once
it was released, and over time, we have been able to
assess these predictions.
The mean annual degree days accumulated for each
location was then calculated, which predicted the number of generations that E. catarinensis could complete
at each locality. The number of generations that the
mirid could complete during the winter months (April
to August) was also calculated. The results of both
of these calculations indicated that fewer generations
were possible at high altitudes, as was expected (Coetzee et al., 2007a).
From these data, we predicted that low winter temperatures will limit the establishment of E. catarinensis
in the field. This explained the failure of mirid establishment at Delta Park in Johannesburg, a high-altitude
site, where it has been released at least three times in
early summer, established, but then has not persisted
through the winter as a consequence of heavy frosts.
However, our prediction could not explain why the mirid did establish at another high-altitude site on the Vaal
River, which experiences similar climatic conditions as
Delta Park. It is likely that the effects of microclimates
513
XII International Symposium on Biological Control of Weeds
provided by an abundance of overhanging vegetation
played a major role in providing thermal refuges for
E. catarinensis at the Vaal River site, thereby allowing
it to persist through the winter. Furthermore, the mirids behaviour might allow it to escape extreme temperatures in the canopy of the plant by moving down
towards the crown in cold weather. Therefore, the assumption that standard meteorological data can be used
to represent the conditions actually experienced by the
insects in the field is a generalization (McClay and
Hughes, 1995; van Klinken, et al., 2003), highlighting
the fact that we cannot ignore the effects of microclimates and behaviour on establishment patterns of the
mirid in South Africa.
Impact
Another factor contributing to the variable results
achieved by water hyacinth biological control agents
in South Africa is the effect of eutrophication of bodies of water (Hill and Olckers, 2000). Water hyacinth
proliferation is usually closely linked to increases in
eutrophication in these systems (Hill, 1999), and as a
result, the effect of feeding by biological control agents
is often insufficient to retard water hyacinth growth
(Hill and Cilliers, 1999).
A previous study investigated the effects of herbivory by the mirid on water hyacinth grown at high, medium and low nitrogen (N) and phosphorus (P) nutrient
concentrations under laboratory conditions (Coetzee
et al., 2007b). The results showed that water-nutrient
concentration affected plant growth parameters of water hyacinth significantly more than did herbivory by
the mirid. At high-nutrient concentrations, leaf and
daughter plant production were more than double than
at low-nutrient concentrations, while stems were twice
as long at high-nutrient concentrations compared to
low concentrations. Chlorophyll content was also twice
as high at high-nutrient concentrations as at low concentrations. Although herbivory by E. catarinensis did
not have as great an effect on water hyacinth vigour
as nutrient concentration, it significantly reduced the
length of second petioles, chlorophyll content of water
hyacinth leaves and the production of daughter plants
(Coetzee et al., 2007b). These results are important
because water hyacinth populations increase rapidly
by vegetative reproduction through the production of
daughter plants (Edwards and Musil, 1975), so any reduction in daughter plant production will have negative
consequences for the rate of spread of water hyacinth.
From these results, we predicted that the mirid would
have the greatest impact on water hyacinth infestations
under mesotrophic conditions. At Clairwood quarry in
KwaZulu-Natal, a eutrophic site, E. catarinensis has
had a major impact on the infestation and is responsible
for clearing the weed. Initially, large, brown, circular
patches appeared in the mat, which gradually began to
sink as the plants started to die. Eventually, most of
the water hyacinth sank, and the majority of the impoundment remains clear, with fringing populations of
water hyacinth. Based on examination of the plants in
the field, the mirid is having an impact on the plants, as
the leaves are brown and clearly chlorotic.
Discussion and conclusion
In the 10 years that this agent has been established in
South Africa, it has been shown to have a wider distribution than was first predicted (Coetzee et al., 2007a),
and while prediction of the insects thermal resilience
was reasonably accurate, the thermal buffering of microclimates considerably underestimated the eventual
distribution. Inability to include local microclimates is
an inevitable failing of models, which use climate envelopes as a basis for predicting distributions (Sutherst,
2003), while presence of the weed is not necessarily
a predictor of climatic suitability for the agent (van
Klinken et al., 2003).
Laboratory host-specificity trials can produce ambiguous results, as the differences between the fundamental/physiological and realized/ecological host
ranges are difficult to determine (van Klinken, 2000).
Undertaking field host-range studies in the region of
origin of the weed can often resolve these ambiguities, but in most cases, only plants common to both
countries can be tested (Olckers et al., 2002). Pickerelweeds presence in South Africa allowed validation of
laboratory host-specificity results in the field, including
assessment of the mirids suitability for release in the
United States. Hill et al. (1999) predicted that at best
pickerelweed would be an inferior host in comparison
to water hyacinth. Field trials confirmed that the mirid
would not establish on pickerelweed in the absence of
water hyacinth, but where the two grow sympatrically,
some spillover feeding is expected (Hill et al., 2000;
Coetzee et al., 2003). However, this spillover feeding
on pickerelweed has not been quantified and possibly
should be, before the agent is considered for release in
the United States. In this case, the laboratory results
overestimated the potential impact on the non-target
species.
E. catarinensis was expected to contribute to the
control of water hyacinth (Hill et al., 1999), but its
overall impact was predicted to be subtle in comparison
to the Neochetina spp. weevils that are the mainstay of
water hyacinth biological control (Coetzee et al., 2005)
and likely to be negligible in highly eutrophic conditions (Coetzee et al., 2007b). Five to six years after release, mirid populations were generally low and their
impact slight. However, in the past 2 to 3 years, several
outbreaks of the mirid have been seen, resulting in water hyacinth mats collapsing, even at eutrophic sites. It
is uncertain if these high populations of the mirid will
persist, and this aspect warrants further study.
514
Ten years after the release of the water hyacinth mirid Eccritotarsus catarinensis in South Africa
Since its first introduction to South Africa in 1992,
there have been ten scientific publications on the biology, host specificity and impact of E. catarinensis on
water hyacinth. Several of the predictions made about
the agent proved to be accurate, while some were underestimates and others overestimates. Despite this
close examination, we are still unclear of the interaction this agent will have with the other agents released
against water hyacinth in South Africa (e.g. Ajuonu et
al., 2007) and its long-term impact on water hyacinth
populations. Selection of a good agent retains the elements of art, even as we improve the science (Hoelmer
and Kirk, 2005).
References
Ajuonu, O., Byrne, M., Hill, M., Neuenschwander, P. and
Korie, S. (2007) Survival of the mirid Eccritotarsus catarinensis as influenced by Neochetina eichhorniae and
Neochetina bruchi feeding scars on leaves of water hyacinth Eichhornia crassipes. BioControl 52, 193200.
Campbell, A., Frazer, B.D., Gilbert, N., Gutierrez, A.P. and
MacKauer, M. (1974) Temperature requirements of some
aphids and their parasites. Journal of Applied Ecology 11,
431438.
Coetzee, J.A., Byrne, M.J. and Hill, M.P. (2003) Failure of
Eccritotarsus catarinensis, a biological control agent of
waterhyacinth, to persist on pickerelweed, a non-target
host in South Africa, after forced establishment. Biological Control 28, 229236.
Coetzee, J.A., Center, T.D., Byrne, M.J. and Hill, M.P. (2005)
The impact of Eccritotarsus catarinensis, a sap-feeding
mirid biocontrol agent, on the competitive performance of
waterhyacinth. Biological Control 32, 9096.
Coetzee, J.A., Byrne, M.J. and Hill, M.P. (2007a) Predicting
the distribution of Eccritotarsus catarinensis, a natural
enemy released on water hyacinth in South Africa. Entomologia Experimentalis et Applicata 125, 237247.
Coetzee, J.A., Byrne, M.J. and Hill, M.P. (2007b) Impact
of nutrients and herbivory by Eccritotarsus catarinensis
on the biological control of water hyacinth, Eichhornia
crassipes. Aquatic Botany 86, 179186.
Edwards, D. and Musil, C.J. (1975) Eichhornia crassipes in
South Africa a general review. Journal of the Limnological Society of Southern Africa 1, 2327.
Hill, M.P. (1999) What level of host specificity can we expect
and what are we prepared to accept from new natural enemies for water hyacinth? The case of Eccritotarsus catarinensis in South Africa. In: Hill, M.P., Julien, M.H. and
Center, T.D. (eds) Proceedings of the First IOBC Global
Working Group Meeting for the Biological and Integrated
Control of Water Hyacinth, 1619 November 1998,
Zimbabwe, pp. 6266.
Hill, M.P. and Cilliers, C.J. (1999) A review of the arthropod
natural enemies, and factors that influence their efficacy,
in the biological control of water hyacinth, Eichhornia
crassipes (Mart.) Solms-Laubach (Pontederiaceae), in
South Africa. In: Olckers, T. and Hill, M.P. (eds) Biologi-
cal Control of Weeds in South Africa (19901998). African Entomology Memoir 1, 103112.
Hill, M.P. and Olckers, T. (2000) Biological control initiatives
against water hyacinth in South Africa: constraining factors, success and new courses of action. In: Julien, M.H.,
Hill, M.P., Center, T.D. and Jianqing, D. (eds) Biological and Integrated Control of Waterhyacinth, Eichhornia
crassipes. Proceedings of the 2nd Meeting of the Global
Working Group for the Biological and Integrated Control
of Waterhyacinth. Beijing, China, 912 October 2000.
Australian Centre for International Agricultural Research,
Canberra, Australia, pp. 3338.
Hill, M.P., Cilliers, C.J. and Neser, S. (1999) Life history and
laboratory host range of Eccritotarsus catarinensis (Carvalho) (Heteroptera: Miridae), a new potential natural
enemy released on water hyacinth (Eichhornia crassipes
(Mart.) Solms-Laub.) (Pontederiaceae) in South Africa.
Biological Control 14, 127133.
Hill, M.P., Center, T.D., Stanley, J., Cordo, H.A., Coetzee,
J.A. and Byrne, M.J. (2000) The performance of the water hyacinth mirid, Eccritotarsus catarinensis, on water
hyacinth and pickerel weed: a comparison of laboratory
and field results. In: Spencer, N.R. (ed) Proceedings of the
Xth International Symposium on the Biological Control
of Weeds. Bozeman, MT, USA, 414 July 1999, Montana
State University, Bozeman, MT, USA, pp. 357366.
Hoelmer, K.A. and Kirk, A.A. (2005) Selecting arthropod
biological control agents against arthropod pests: can the
science be improved to decrease the risk of releasing ineffective agents? Biological Control 34, 255264.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds. A World Catalogue of Agents and their Target
Weeds. Fourth Edition. CABI Publishing, Wallingford,
UK, 223 pp.
McClay, A.S. and Hughes, R.B. (1995) Effects of temperature on developmental rate, distribution, and establishment of Calophasia lunula (Lepidoptera: Noctuidae), a
biocontrol agent for Toadflax (Linaria spp.). Biological
Control 5, 368377.
Olckers T., Medal, J.C. and Gandolfo, D.E. (2002) Insect herbivores associated with species of Solanum (Solanaceae)
in northeastern Argentina and southeastern Paraguay, with
reference to biological control of weeds in South Africa
and the United States of America. Florida Entomologist
85, 254260.
Sutherst, R. (2003) Prediction of species geographical ranges.
Journal of Biogeography 30, 112.
van Klinken, R.D. (2000) Host specificity testing: why we
do it and how we can do it better. In: Spencer, N.R. (eds)
Proceedings of the Xth International Symposium on the
Biological Control of Weeds. Bozeman, MT, USA, 414
July 1999, Montana State University, Bozeman, MT,
USA, pp. 5468.
van Klinken, R.D., Fichera, G. and Cordo, H. (2003) Targeting biological control across diverse landscapes: the release, establishment and early success of two insects on
mesquite (Prosopis spp.) in Australian rangelands. Biological Control 26, 820.
Williamson, M. (1996) Biological Invasions. Chapman and
Hall, London.
515
Release and establishment of the Scotch
broom seed beetle, Bruchidius villosus,
in Oregon and Washington, USA
E.M. Coombs,1 G.P. Markin2 and J. Andreas3
Summary
We provide a preliminary report on the Scotch broom seed beetle, Bruchidius villosus (F.) (Coleoptera: Bruchidae). This beetle was first recorded as an accidental introduction to North America in
1918. Host-specificity tests were completed before the beetle was released as a classical biological
control agent for Scotch broom, Cytisus scoparius L. (Fabaceae), in 1997 in the western USA. Beetles
were collected in North Carolina and shipped to Oregon in 1998. More than 135 releases of the beetle
have been made throughout western Oregon and Washington. Nursery sites have been established,
and collection for redistribution began in 2003. The bruchids initial establishment rate is higher in interior valleys than at cooler sites near the coast and in the lower Cascade Mountains. Seed-pod attack
rates varied from 10% to 90% at release sites that were 3 years old or older. Seed destruction within
pods varied from 20% to 80%, highest at older release sites. B. villosus may compliment the impact
of the widely established Scotch broom seed weevil, Exapion fuscirostre (F.) (Coleoptera: Curculionidae). B. villosus populations were equal to or more abundant than the weevil at seven release sites
in Oregon. At sites where the bruchids were established, they made up 37% of the seed-pod beetle
population, indicating that they are able to compete with the weevil and increase their populations. At
several release sites older than 5 years, bruchid populations have become equal to or more abundant
than the weevils.
Keywords: Cytisus scoparius, Exapion fuscirostre, post-release assessment,
pre-dispersal seed predation, biological control.
Introduction
Scotch broom, Cytisus scoparius (L.) Link, Fabaceae,
is native to central and southern Europe. It is a nitrogenfixing leguminous perennial shrub 13 m tall with
showy yellow pea-like flowers. Each plant may produce
thousands of seeds that are long-lived, up to 80 years in
the soil (Bossard and Rejmanek, 1994). Scotch broom
was introduced into North America as an ornamental
plant, but it escaped cultivation and became naturalized. Without its natural enemies to keep it in check,
Oregon Department of Agriculture, 635 Capitol St., Salem, OR, 97301,
USA.
2
US Forest Service, Rocky Mountain Research Station, Bozeman, MT,
59717, USA.
3
Washington State University, King County Extension, 200 Mill Ave. S.,
Ste 100, Renton, WA, 98057, USA.
Corresponding author: E.M. Coombs <[email protected]>.
CAB International 2008
1
Scotch broom flourished and spread, particularly in
disturbed habitats (Andres et al., 1967). It occurs from
British Columbia south to California and along the
eastern seaboard of the USA, with major infestations in
North Carolina (Andres and Coombs, 1995; Redmon
et al., 2000; Coombs et al., 2004). Scotch broom has
spread throughout the Pacific Northwest, west of the
Cascade Mountains (Andres and Coombs, 1995), and
is considered noxious in California, Hawaii, Idaho, Oregon and Washington. Most infestations occur where
the natural plant communities have been disrupted by
logging, construction, overgrazing and rights of way.
In Oregon, where the plant infests the western third of
the state, an annual economic loss was estimated at $47
million (Radtke and Davis, 2000). Much of the loss
was associated with competition affecting reforestation
after logging and the cost of control.
Because Scotch broom is a serious invasive pest in
several countries, including Australia, New Zealand,
Canada and the United States, an international program
516
Release and establishment of the Scotch broom seed beetle, Bruchidius villosus, in Oregon and Washington, USA
has been established to explore the possibilities for classical biological control (Syrett et al., 1999). Biological
control of Scotch broom in the United States began in
1960 with the release of the twig-mining moth, Leucoptera spartifoliella Huebner (Lepidoptera: Lyonetiidae) (Frick, 1964; Andres and Coombs, 1995). Upon
additional investigation, the moth was found to occur
in the Pacific Northwest before its intentional introduction (Andres and Coombs, 1995). The seed weevil,
Exapion fuscirostre F. (Coleoptera: Apionidae), was
released in 1964 in California (Andres et al., 1967), in
1983 in Oregon and about 1989 in Washington (Andres
and Coombs, 1995; Coombs et al., 2004). Large-scale
implementation programs for collection and redistribution of the seed weevil have been conducted by the
Oregon Department of Agriculture (ODA) (Isaacson et
al., 1995). The seed weevil is now widely distributed,
occurring throughout the Pacific Northwest (Coombs
and Piper, 2002; Coombs et al., 2004). ODA surveys
showed that the degree of seed reduction by the weevil
alone (2060%) was insufficient for long-term control,
based on studies by Rees and Paynter (1977). Numerous other insects are associated with Scotch broom in
the United States, many of which are adventive introductions (Waloff, 1968).
To further reduce Scotch broom seed production, an
additional seed-feeding insect was sought to compliment the impact of the weevil. The Scotch broom seed
beetle, Bruchidius villosus F. (Coleoptera: Bruchidae),
was accidentally introduced in the eastern USA during the early 1900s (Redmon et al., 2000). It was first
reported in 1918 in Massachusetts and is abundant in
North Carolina, where seed reduction has been measured at more than 80% (Redmon et al., 2000). The
bruchid was introduced in Oregon in 1998 and in
Washington in 1999 (Coombs et al., 2004). This was
the first adventive natural enemy in the United States
to undergo host-specificity testing according to the US
Department of Agricultures (USDA) protocol to become a sanctioned classical weed biological control
agent.
Biological control agent
The life history of B. villosus, as provided by Southgate
(1963) and Andres et al., (1967), is a univoltine insect
that overwinters as an adult and is native to western
and central Europe. Adults emerge from the duff in the
spring and feed on Scotch broom pollen. After mating,
gravid females may oviposit 212 eggs on the outside
of Scotch broom seed pods. Larvae hatch and tunnel
into the pod and begin feeding on the seeds. Usually,
each larva develops in a single seed, completing development in 1530 days. Pupation lasts 1020 days and
occurs within the seed coat. Adults are dark grey and
about 23 mm in length. Adult feeding has little impact
on the plant. Other plant species attacked include Portuguese broom, Cytisus striatus (Hill) Rothm., Spanish
broom, Spartium junceum L., and French broom, Genista monspessulana (L.) L. Johnson.
Methods
After the appropriate USDA permits were obtained,
adult B. villosus were collected in the spring of 1998
in North Carolina and shipped to Oregon and later to
Washington. Three main habitat types (coastal, valley
and mountain) were chosen as release habitats where
Scotch broom occurs. Releases of 100400 adults per
release were made, and site data were recorded. Releases
of adult beetles during the mating and oviposition periods were preferred over releasing teneral adults. Selected release sites in each of the three main habitat
types were sampled to evaluate establishment 4 or more
years after release and to estimate numbers of adult B.
villosus. Site evaluations included the use of beating
sheets one to three times per year to verify presence of
adults and sampling of three to ten plants per site, with
three samples per plant. Seed pods (25100 per site)
were harvested to determine presence of eggs, larvae,
adult seed beetles and numbers of damaged seeds. B.
villosus was determined to be established when adults
were recovered 3 or more years after release. When
collection at nursery sites exceeded 200 adults per hour
sampled, surplus beetles were harvested for redistribution to other sites.
Results and discussion
The first release of B. villosus as a classical biological control agent for Scotch broom in the United States
was made in western Oregon in August 1998. Unfortunately, the release cage used was either vandalized or
blown away by strong winds, and the beetle was never
recovered. Numerous releases were made in other areas from 1999 to 2001. In 2003, an established population west of Salem, OR, supported surplus beetles for
local redistribution (Table 1). We anticipate that major
collection and redistribution projects will occur during
the coming decade.
Table 1.
517
Year
1998
1999
2000
2001
2002
2003
2004
2005
2006
Total
ear, number of releases and total number of
Y
Bruchidius villosus (F.) released in Oregon. First
collections for redistribution began in 2003.
Number of releases
2
17
8
30
3
1
1
7
14
83
Number released
280
3580
1600
7550
450
100
200
1300
3100
18,160
XII International Symposium on Biological Control of Weeds
In western Oregon, more than 83 releases of B.
villosus have been made throughout the core area of
Scotch broom infestations (Fig. 1) and at least 52 sites
in Washington. We attempted to include a wide variety of habitats that were infested with Scotch broom,
including coastal, inland valley and mountainous. We
sampled for adult beetles at 21 of 57 sites that were 4 or
more years old. In 2005, we were unable to find B. villosus at seven of the 21 sites where a beating sheet was
used, and three of the 26 sites where ripe seed-pods
were collected. E. fuscirostre was found at all but one
site in each of these surveys. We suspect that, as populations of B. villosus become better adapted to local climates and conditions, the number of sites without the
bruchid will decrease.
Our observations showed that B. villosus populations
required at least 4 years before sites were amenable to
collection for redistribution. The bruchid does not appear to aggregate at new release sites and thus are difficult to recover until populations build up at a given site
for at least 3 years. We were unable to find release sites
that were not already inhabited by E. fuscirostre at the
time of releasing B. villosus because of the earlier intensive collection and redistribution program conducted
by ODA (Fig. 1). We hope to conduct tests to determine
the population potentials and impacts of both beetles
alone and in combination. Of the 21 release sites moni-
Figure 1.
tored for adult beetles, the mean ratio of the bruchid to
the weevil increased with time, and at three 6-year-old
sites was 50% or greater. Our data show that, despite
competition with E. fuscirostre, B. villosus should continue to spread and increase (Fig. 2).
It is not yet clear which insect in which habitat is the
better competitor without long-term monitoring. Data
from the seed-pod analysis (Fig. 3) show that, with respect to E. fuscirostre, the overall proportion of B. villosus regardless of habitat sampled remains about the
same when years-since-release is not factored in. Our
sample size is inadequate at this time to clearly identify
the trends. However, when we looked at establishment
rates as a function of general habitat region, it was evident that B. villosus had a higher rate of establishment
at the inland sites in the Willamette Valley of western
Oregon (Fig. 3). The valley climate is drier and warmer
than the coastal and mountainous sites.
Seed-pod attack rates were highly variable, from
10% to 90% at release sites that were 3 years old or
older. Seed destruction within pods varied from 20%
to 80%, being highest at older release sites. Analysis
of seed-pod attack rates and seed destruction will be
analyzed in future work.
There may be inherent problems in our early analysis
that compared active adults collected by beating plants
to the number of adults sampled in the seed-pod col-
Release locations of Bruchidius villosus (F.) (circles; n = 83) and Exapion
fuscirostre (F.) (squares; n = 529) on Scotch broom in Oregon, USA.
518
Release and establishment of the Scotch broom seed beetle, Bruchidius villosus, in Oregon and Washington, USA
BRVI Population Increase
50
40
30
20
4
Years since release
Figure 2.
Mean increase of the 14 established Bruchidius villosus (F.) populations (from ripe seed-pod
analysis) by years since release, as a percentage of the total seed-pod beetle population, including
Exapion fuscirostre (F.), in western Oregon.
lections. Free-roaming adult beetles can move about;
therefore, behaviour and abundance of adults of each
species may differ in relation to plant phenology. This
appears to be evident when comparing populations
sampled on the same date from the cooler coastal and
mountainous sites, which were phenologically behind
those from the warmer and drier valley sites. Future
analysis should be adjusted to specific degrees of plant
phenology, e.g. percent in flower, average size or colour of pods. However, because of the variation that we
have observed among and between sites, this may not
be practical.
In the course of sampling the production of beetles
in seed-pods, we found that a small percentage of beetle larvae were parasitized by chalcidoid wasps. Representative wasps were collected and sent to a USDA
taxonomist for identification. The most common of the
four species of pteromalids and the only one identified
to species was Pteromalus sequester Walker, a natural
enemy of B. villosus (Syrett et al., 1999) and the gorse
BRVI by Habitat
100
BRVI
80
ESTAB
60
40
20
0
Coastal
Valley
Mountain
Habitat type
Figure 3.
Percent of Bruchidius villosus (F.) releases, based on adult surveys, that were established
(open bars) and the percentage of seed-pod beetle populations consisting of B. villosus (grey
bars) by regional habitat type. The establishment rate of B. villosus was higher in the Willamette Valley region than in coastal or mountainous regions, but their population percentages
in relation to Exapion fuscirostre (F.) were not.
519
XII International Symposium on Biological Control of Weeds
seed weevil, Exapion ulicis Forster, in Europe (Davies,
1928). Our observations showed that parasitoid attacks
on B. villosus and E. fuscirostre did not exceed 4% at
any of the sites. We do not know how this parasitoid
arrived in the United States, but we are concerned that
they have the potential to render both seed-pod beetles
ineffective as biocontrol agents.
Our studies show that B. villosus populations can
establish and increase at the Scotch broom habitat
types that we surveyed in the western USA. It is still
too early to determine the degree of seed reduction that
will be achieved by the combination of B. villosus and
E. fuscirostre. Additional long-term studies are needed
to document the spread and impact of both seed-pod
beetles. As both species should become widely established, we hope that their impact will be additive, or at
the very least, non-competitive. Based on studies by
Rees and Paynter (1977), we expect that a 95% reduction of seed production will be necessary to achieve
long-term control of Scotch broom infestations. At the
present time, B. villosus appears to be better adapted
to the warmer and drier climate of the interior western
valleys of the American west coast. Additional studies
are needed to determine the combined impacts of all
natural enemies established on Scotch broom in the
western USA.
Acknowledgements
The authors wish to express their thanks to Dr. Timothy Forrest, University of North Carolina, Asheville,
for providing the original releases of B. villosus for the
western USA and to the many agency (ODA, BLM and
USFS), county and university personnel who helped
at various stages of this study. A special thanks is extended to Carol Horning, who spent many a long day
making collections in the field.
References
Andres, L.A., Hawkes, R.B. and Rizza, A. (1967) Apion seed
weevil introduced for biological control of Scotch broom.
California Agriculture 21, 13.
Andres, L. A. and Coombs, E.M. (1995) Scotchbroom, Cytisus scoparius (L.) Link (Leguminosae). In: Nechols, J.R.,
Andres, L.A., Beardsley, J.W., Goeden, R. and Jackson,
C.G. (eds) Biological Control in the U.S. Western Region:
Accomplishments and Benefits of Regional Research
Project W-84, 19641989. University of California Di-
vision of Agriculture and Natural Resources. Pub. 3361.
Oakland, USA, pp. 303305.
Bossard, C.C. and Rejmanek, M. (1994) Herbivory, growth,
seed production, and resprouting of an exotic invasive
shrub, Cytisus scoparius. Biological Conservation 67,
193200.
Coombs, E.M. and Piper, G.L. 2002. Biological control of
weeds a tool for forest management. American Forester
47(3), 810.
Coombs, E.M., Markin, G.P. and Forest, T.G . (2004) Scotch
broom. In: Coombs, E.M., Clark, J.K., Piper, G. L and
Cofrancesco, A.F., Jr. (eds) Biological Control of Invasive Plants in the United States. Oregon State University
Press, Corvallis, USA, pp. 160168.
Davies, W.M. (1928). The bionomics of Apion ulicis (gorse
weevil) with special reference in the control of Ulex europaeus in New Zealand. Annals of Applied Biology 15,
263286.
Frick, K.E. (1964) Leucoptera spartifoliella, an introduced
enemy of Scotch broom in the western United States.
Journal of Economic Entomology 57, 589591.
Isaacson, D.L., Miller, G.A. and Coombs, E.M. (1995) Use of
geographic systems (GIS) distance measures in managed
dispersal of Apion fuscirostre for control of Scotch broom
(Cytisus scoparius). In: Delfosse, E.S. and Scott, R.R.
(eds) Proceedings of the VIII International Symposium on
Biological Control of Weeds. Lincoln University, Canterbury, New Zealand. DSIR/CSIRO, Melbourne, Australia,
pp. 695699.
Radtke, H. and Davis, S. (2000) Economic analysis of containment programs, damage, and production losses from
noxious weeds in Oregon. Technical report prepared for
Oregon Department of Agriculture, Salem, USA, 40 pp.
Redmon, S.G., Forrest, T.G. and Markin, G.P. (2000) Biology
of Bruchidius villosus (Coleoptera: Bruchidae) on Scotch
broom in North Carolina. The Florida Entomologist 83,
242253.
Rees, M. and Paynter, Q. (1997) Biological control of Scotch
broom: modeling the determinants of abundance and the
potential impact of introduced insect herbivores. Journal
of Applied Ecology 34, 12031221.
Southgate, B.J. (1963). The true identity of the broom bruchid (Coleoptera) and synonymic notes on other species
of Bruchidius. Annals of the Entomological Society of
America 56, 795798.
Syrett, P., Fowler, S.V., Coombs, E.M., Hosking, J.R., Markin, G.P., Paynter, Q.E. and Sheppard, A.W. (1999) The
potential for biological control of Scotch broom (Cytisus
scoparius (L.) Link) (Fabaceae) and related weedy species. Biocontrol News and Information 20, 1734.
Waloff, N. (1968) Studies on the insect fauna on Scotch
broom Sarothamnus scoparius (L.) Wimmer. Advances in
Ecological Research 5, 87208.
520
Biological control of Mediterranean sage
(Salvia aethiopis) in Oregon
E.M. Coombs,1 J.C. Miller,2 L.A. Andres3
and C.E. Turner4
Summary
Mediterranean sage, Salvia aethiopis L., is a serious naturalized invasive plant of rangelands in the
sagebrush steppe in the Pacific Northwest area of the United States. Two species of weevils, Phrydiuchus tau Warner and Phrydiuchus. spilmani Warner, were introduced from Europe as classical
biological control agents. Only P. tau established and was widely redistributed throughout the region.
Our observations show that, after the establishment and population increase of the weevil, densities
of Mediterranean sage decreased at three of the four initial release sites and subsequently at 17 of
25 weevil release sites where plant densities dropped 25 orders of magnitude from >1/m2. Level of
control appears to be associated with a combination of plant community type, disturbance and grazing
intensity. The decline in weed density was most apparent in the sagebrush steppe community with
light to no grazing. In comparison, salt desert scrub, annual grass dominated and heavily grazed communities showed little change in Mediterranean sage density over 25 years. This is the first report of
successful biological control against Mediterranean sage.
Keywords: successful management, range improvement, weevil, Phrydiuchus tau.
Introduction
Target plant
Mediterranean sage, Salvia aethiopis L. (Lamiaceae), is a pungent herbaceous monocarpic, biennial
weed, naturalized in the United States from its native
range in southern and southeastern Europe (Tutin et al.,
1972; Mijatovic, 1973; Davis, 1975; Roche and Wilson, 1999). The plant has been observed and studied in
xeric ruderal habitats in Yugoslavia, Bulgaria, Greece,
Turkey and Iran, and was occasionally observed in
similar habitats in France, Italy and Spain (Andres and
Drea, 1963).
Mediterranean sage and several of its allies have
been imported and cultivated in the United States for
use as ornamentals and medicinal herbs (Bailey, 1935).
Mediterranean sage was first reported in North America
Oregon Department of Agriculture, 635 Capitol St., Salem, OR 97301,
USA.
2
Oregon State University, Rangeland Ecology and Management, Corvallis, OR 97331, USA.
3
Retired (USDA-ARS, 1324 Arch St., Berkley, CA 94708, USA)
4
Deceased (USDA-ARS)
Corresponding author: E.M. Coombs <[email protected]>.
CAB International 2008
1
from California in 1892 (Howell, 1942) and was suspected of being introduced as a contaminant in alfalfa
seed (Dennis, 1980). It has since spread and become
a serious range pest in Oregon, California and Idaho
and a minor problem in Washington, Colorado, Texas
and Arizona (Wilson et al., 1994; Coombs and Wilson,
2004). The major infestation in the United States occurs in the Goose Lake Basin in southern Lake County,
OR, and northern Modoc County, CA (Andres et al.,
1995). Overall infestations were estimated at 510,000
ha in seven Oregon counties (Radtke and Davis, 2000),
2800 ha in northeastern California, 1600 ha in northern Idaho (Wilson et al., 1994) and 120 ha in Colorado
(Mowrer, 1996). Infestations have also been recorded
in South Dakota, Arizona and Texas.
Mediterranean sage reproduces only by seed, sprouting in the fall or spring after rains provide adequate soil
moisture for seedling survival. Young plants overwinter
as rosettes and typically bolt in late spring and flower
mid-June to early July. Smaller rosettes may overwinter for a second year without flowering (Wilson and
McCaffrey, 1993). Mature plants grow to 2090 cm
tall, producing up to 100,000 seeds. Seeds are dispersed during the late summer and fall after the stalks
break off and are blown about as tumbleweeds.
In Oregon, the majority of Mediterranean sage occurs in early successional communities in the sagebrush
521
XII International Symposium on Biological Control of Weeds
steppe, upper salt desert shrub and disturbed sites dominated by annuals (e.g. rangelands and pastures overgrazed by livestock, abandoned farm land, construction
sites, rights-of-way and burned areas). This non-toxic
unpalatable plant displaces livestock forage and was
estimated to cause $1.05 million in annual losses in Oregon (Radtke and Davis, 2000).
Mediterranean sage can be a difficult plant to manage. Foliar herbicides are less effective due to the woolly
leaves, which can act as barriers reducing chemical absorption. Furthermore, the plant occurs on low-value
land, and in conjunction with high costs and difficulty of
control, it was targeted for biological control by the US
Department of Agriculture Agricultural Research Service (USDA-ARS) in 1958 (Andres and Drea, 1963).
The project was initiated at a time when it was thought
that the weed would have a major impact on rangelands
throughout the Great Basin in the American northwest.
Biological control agent
Foreign exploration to search for natural enemies of
Mediterranean sage was conducted in Turkey and Iran.
A weevil, later described as Phrydiuchus tau Warner
(Coleoptera: Curculionidae), was found (Andres and
Drea, 1963; Andres and Riza, 1965; Andres, 1966;
Warner, 1969). A second species, P. spilmani Warner
(Warner, 1969), was found feeding on Salvia verbenacea L. in Italy. During feeding trials, P. spilmani also
accepted S. aethiopis as a host.
Both species of weevils were introduced in the
United States as classical biological control agents for
Mediterranean sage (Andres et al., 1995). A total of
631 adult P. spilmani were collected in Italy and released in Oregon from 1969 to 1970. Although several
recoveries of adults were made, the weevil failed to
establish (Andres et al., 1995). A total of 1650 adult
P. tau were introduced from Yugoslavia and released in
Oregon from 1971 to 1973 (Andres et al., 1995). P. tau
readily established, and the sites provided surplus weevils that were shipped to California in 1972 (Andres
et al., 1995), Idaho in 1979 (Wilson and McCaffrey,
1993) and Colorado in 1992 (Mowrer, 1996). A limited
local redistribution began in Oregon and California in
1976 (Andres et al., 1995), followed by an intensive redistribution program in Oregon by the Oregon Department of Agriculture (ODA) from 19791983 (Coombs
et al., 1996; Turner and Coombs, 1996).
The life history of P. tau was studied as part of the
process of assessing its efficacy as a biological control agent. The weevil is univoltine, producing one
generation per year. In the fall, females oviposit along
the midribs and petioles on the undersides of rosette
leaves (Andres, 1966; Wilson and McCaffrey, 1993;
Coombs and Wilson, 2004). The newly hatched larvae
penetrate the plant epidermis and tunnel down the leaf
petiole into the root crown. Larval feeding can sever
vascular tissues of the root crown, which results in the
suppression of flowering and seed production. Severe
larval feeding can result in plant death (Andres, 1966).
In the spring (June), the mature larvae exit the rosette
and construct a spherical chamber of soil particles in
which they pupate. Adults emerge after several weeks
and feed on the foliage and flowers before entering
a period of summer dormancy (aestivation) until fall
rains stimulate plant growth. In the fall, adults feed,
mate and deposit eggs on rosettes (Andres, 1966; Wilson and McCaffrey, 1993).
Biological control has come under some scrutiny because of the lack of monitoring efficacy and non-target
effects (McEvoy and Coombs, 2000). To document the
overall efficacy of the Mediterranean sage biological
control project, we determined that there was a need
to conduct additional evaluation beyond the original
release site studies. Our project objectives were to:
(1) continue monitoring the original four ARS release
sites, (2) assess efficacy at early ODA regional release
sites, (3) document trends of Mediterranean sage density, (4) document the relationship of P. tau numbers to
plant density and (5) compare trends of plant density to
site characteristics and land use.
Methods and materials
Original release sites
Initial observations of plant and insect interactions
were conducted by Andres and cooperators in southern
Lake County, OR, near the town of Lakeview. The climate is characterized by wet winters and dry summers.
The average annual precipitation in the Lakeview area
is 36 cm. Four sites served as the original release sites
for P. tau. The criteria for selection were heavy infestations of Mediterranean sage (>1 m2), and representative of the variety of plant communities and land use
in the area. Three sites: (1) Geyser, about 1.5 km north
of Lakeview; (2) Cottonwood, 13 km west of Lake
view and (3) Killer Hill, 6.5 km south of Lakeview, were
classified as sagebrush steppe, encompassing dense to
open grassland intermixed with dense to open shrub.
The fourth site, Dicks Creek, 24 km north of Lake
view, is in a yellow pine-shrub forest that was logged
10 years before the release of P. tau. Unfortunately, several years of data from the ARS studies were lost during
a relocation of the USDA-ARS offices at the Albany,
CA facility. Furthermore, the Killer Hill site was infrequently visited because it was difficult to relocate.
Periodic visits were made to the four original release sites, usually in early November to (1) estimate
the density of rosettes per square metre, (2) conduct a
timed count of the number of adult weevils observed
per observer hour and (3) estimate percent ground cover
by rosettes. No data were taken between 1980 and
1991. Plant density and percent cover were determined
by using a 1/4-m2 quadrat frame placed on the ground
every 4 m along established transects. Twelve readings
522
Biological control of Mediterranean sage (Salvia aethiopis) in Oregon
were made along each cardinal direction for a total of
48 plots per site per visit. Weevil counts were obtained
by counting the number of adults observed per minute
in each transect to determine number of weevils per
observer hour.
Regional release sites
After the establishment of P. tau at the original release sites, a programme of collection, redistribution
and monitoring was conducted by ODA. By November
of 1976, populations of P. tau at the Cottonwood site
reached >100 adults collected per hour. Weevils from
that site were redistributed throughout infestations in
Oregon and California. In the following years (1979 to
1983), ODA and cooperators made releases of P. tau
throughout the region. The objective was to release
weevils at all heavily infested townships and in each
major drainage system. Some collection and redistribution efforts were made by ODA during later years at
new and isolated infestations and for releases in other
states (Coombs and Wilson, 2004).
The collections of P. tau adults for redistribution were
primarily conducted during the mating period in November. Weevils were sorted into groups of 100 adults
and placed into containers provisioned with tissue paper and host plant foliage. Releases occurred within
several days of collection at Mediterranean sage infestations 0.5 ha and a minimum density of 1.0 m2.
In the summer of 1995, a field survey was conducted
at 25 sites where P. tau had been previously released by
ODA from 1979 to 1983. Only sites that could be positively located from maps on the original release forms
Figure 1.
were selected. Each site was inventoried for (1) estimating plant density by using 1 m2 quadrats or actual
counts at low densities within 1000 m2, (2) determining
the number of adult P. tau observed per minute, (3) assessing present land use and degree of grazing, rated as
none, light (<50% use) and heavy (>50% use) and (4)
documenting plant community type. The apparent degree of control of Mediterranean sage based on density
of plants square metre was categorized as: poor, >1.0;
fair, 0.990.1; good, 0.090.01; and excellent, <0.009.
The sites were periodically sampled from 1999 to 2006
to determine the stability of control and to assess variability within plant populations.
Results
The weevil, P. tau, established at all four of the original
ARS release sites. Weevils from these sites were later
collected and released at numerous infestations in the
region. Since the inception of the project, density and
percent cover of Mediterranean sage has declined by
several orders of magnitude at most sites in the Goose
Lake Basin but not in the lower and drier Abert Lake
Basin. The following are some of the early results that
depict the interaction of weevil-plant dynamics through
time at the original release sites and regional release
sites. Plant community types and land use interactions
will be analyzed in future works.
Original release sites
Original plant density at the release sites varied from
3.2 to 12.4 m2 at the time of weevil release (Fig. 1).
Density changes log10 of Mediterranean sage at the four original USDA-ARS study
sites. Triangles: Killer Hill site, a heavily over-grazed site that has changed little since
the release of Phrydiuchus tau. Zero data were entered as 0.0001.
523
XII International Symposium on Biological Control of Weeds
Subsequent monitoring of the Mediterranean sage density showed a slight downward trend during the first
5 years among the four sites (Fig. 1). By 1992, plant
densities had decreased to less than 1 m2. From year
2000 on, plant densities remained less than 0.01 m2 at
all sites, except Killer Hill, which was heavily grazed
each winter by livestock.
We measured plant density, cover and weevil numbers at the Geyser site and found cover and weevils
correlated with plant density (Fig. 2). The weevil population increased to more than 100 per observer hour at
two of the sites from 1976 to 1978. After this peak, weevil numbers dropped to 1.3 per observer hour by 1993.
By 1993, plant density had dropped to 0.02 m2. No
plants or weevils were found at the Geyser site during
the July 1995 visit (Fig. 2). In 1999 and in 2000, several plants with adult feeding were observed but none
in 2001 samples, as the site had been recently mowed
and sprayed with herbicides to control other species of
weeds. The Geyser site was representative of the trends
that we observed in Dicks Creek and Cottonwood sites,
in that weevil numbers increased and later decreased as
plant density decreased. However, the heavily grazed
Killer Hill site changed little in 25 years (Fig. 1).
Regional survey
More than 120 releases of P. tau occurred from 1979
to 1983, which were collected at the original ARS release sites. After these releases, casual inspections by
Figure 2.
various cooperators revealed that the weevil was widespread throughout the Mediterranean sage infested area.
In 1995, we determined that 25 release sites were amenable to quantitative monitoring to document changes
over time (Table 1). Variations in Mediterranean sage
densities appear to be associated with plant community
type and community disturbance history (grazing intensity, fire, road construction and agriculture). The lowest Mediterranean sage densities observed were at sites
that consisted of perennial grass/shrubs with no-to-light
grazing. Conversely, we observed higher densities of
Mediterranean sage at sites characterized by salt desert
scrub and annual grasses. At sites with heavier grazing
pressure (>50% biomass removal of perennial grasses),
control of Mediterranean sage (0.50.05 mature plants/
m2) was better in communities with a strong perennial
grass component. The poor control of Mediterranean
sage in the ungrazed areas was due partly to historic
heavy grazing, having reduced the perennial grasses,
and more to recent fires, which allowed the invasion
of annual grasses. Again, the lack of competition and
resetting the successional clock were likely factors.
In the 1995 survey, 12 sites were not grazed, five
were lightly grazed and nine were heavily grazed
(Table 1). It appears that heavy grazing reduces the
amount of plant competition against Mediterranean
sage plants weakened by the biological control agent.
Caution should be used when comparing these sites,
as our inventory was not designed to cover an equal
number of sites in each category.
Changes in log10 of plant density, percent plant cover and numbers of Phrydiuchus tau
(PHTA) per observer hour over time at the original USDA-ARS Geyser site. Zero data
were entered as 0.0001.
524
Biological control of Mediterranean sage (Salvia aethiopis) in Oregon
Table 1.
omparison of the level of biological control of Mediterranean sage, plant community type and grazing levels at
C
25 Phrydiuchus tau regional release sites in Oregon in 1995.
Control levela
Number of sites
Excellent
Good
Fair
6
4
4
3
2
4
2
25
Poor
Total
Community typeb
Grazing intensityc
Shrub/perennial grass
Shrub/perennial grass
Shrub/annual
Salt scrub/annual grass
Shrub/perennial grass
Shrub/annual
Salt scrub/annual grass
Heavy
Light
None
0
1
3
2
1
2
0
9
4
0
0
1
0
0
0
5
2
3
1
1
1
2
2
12
Control level = average mature plants per square metre; poor, 1; fair, 0.990.1; good, 0.0090.01; excellent, 0.009.
Community type based on dominant species: shrub = Artemisia tridentata, Purshia tridentata; salt scrub = Sarcobatus vermiculatus, Atriplex
confertifolia; perennial grass = Agropyron cristatum, Festuca idahoensis; annual grass = Bromus tectorum, Taeniatherum caput-medusae.
c
Grazing intensity utilization of above ground biomass: heavy, 50%; light, 50%.
Of the regional release sites, 12 were dominated by
perennial grasses and shrubs and 13 had a large component of annual grasses (Table 1). Four of the sites had
no detectable level of Mediterranean sage (three were
shrub/perennial grass communities and one was in agricultural production). In contrast, the highest density
of Mediterranean sage (>1.0/m2) remained in salt desert
scrub and shrub/annual grass communities, irrespective
of grazing. When compared to the original minimum
density estimate 1.0 mature plant per square metre,
68% of the survey sites showed reductions of Mediterranean sage of one to three orders of magnitude. We
are more concerned about the numbers of mature plants
Figure 3.
at sites because they produce seed and contribute to rising generations.
Among the 25 sites visited in 1995, 16 sites were
periodically re-inventoried through 2006. We observed
two distinct patterns in Mediterranean sage density
through time based on plant community type (Fig. 3).
One pattern was a precipitous drop in Mediterranean
sage from the initial estimated plant density (>1 m2;
Fig. 3). The second pattern exhibited a stable trend of
Mediterranean sage density over time in plant communities that had high components of annual grasses,
mostly due to recent disturbances (fire and heavy grazing). From 1995 through 2006, Mediterranean sage con-
Changes in plant density (log10) of ODA regional release sites (n = 25) over time. The
large rectangle represents a conservative estimate of density of Mediterranean sage at
release of Phrydiuchus tau and during the 5 or 6 years after release of P. tau. Circles
represent salt desert scrub sites in the Abert Lake Basin with high annual grass component and triangles represent Goose Lake Basin sites with high perennial shrub/grass
components. Zero data were entered as 0.0001.
525
XII International Symposium on Biological Control of Weeds
tinued to decline to the point that, by 2006, nine sites
exhibited excellent control (<0.009 mature plants/m2).
Most severe infestations typically have two to five mature plants and five to 15 rosettes per square metre. As
original mature plant density estimates were 1.0/m2,
our conservative appraisal of control may underestimate the effectiveness of P. tau. By 2006, 17 of the
25 ODA regional release sites inventoried showed a
reduction in Mediterranean sage density of more than
90%, with nine sites near 99%. One mature plant per
square metre was considered as the economic threshold
level to justify control measures.
In 2005, after a period of heavy late spring rains,
Mediterranean sage populations in the annual grass
dominated salt desert scrub community in the Abert
Lake basin flourished (1.5/m2), and we expected to
see a large increase of Mediterranean sage the following year. However, outbreak-level populations of
P. tau severely defoliated rosettes and flowering plants
(5090% defoliation). Monitoring in 2006 showed an
average reduction of mature plants to 0.1 m2, which we
attributed to the heavy impact by the weevil.
Discussion
Smith and DeBach (1942) identified three criteria essential for evidence of successful biological control. The
Mediterranean sage biological control project meets
two of those criterions. First, the densities of Mediterranean sage declined after the introduction, establishment and increase of P. tau. In fact, two sites used by
ODA personnel to collect P. tau in the late 1980s were
no longer viable in 1994 because the host plants were
nearly eliminated. Second, the levels of Mediterranean
sage for the most part have remained at low levels after the establishment and increase of P. tau. The third
criterion, to show that the target weed would return to
its original density when the weevils are removed, remains to be experimentally tested.
Populations of P. tau established and increased at
the four original study sites, later followed by a decrease in the density of Mediterranean sage and eventually a decrease on P. tau. Following a lag time of 10 to
20 years after redistribution of the weevil, reductions of
Mediterranean sage density occurred throughout much
of the Lakeview area. The majority of landowners,
managers and county extension agents we consulted
agreed that the overall density of Mediterranean sage
in the Lakeview Valley was much lower than before
P. tau was released.
Our studies suggest that larval destruction and weakening of Mediterranean sage rosettes, supplemented
with the stress of competing perennial vegetation, were
the main causes of control. Plant competition appears
to play an important role in enhancing Mediterranean
sage control. Best control occurred in those communities with a strong perennial grass component. Conversely, despite weevil presence, wherever recent fires
and heavy grazing had reduced the vigour of competing grasses, the level of Mediterranean sage remained
troublesome or decreased slightly. Mediterranean sage
sites in some salt desert shrub and annual grass communities (dominated by Bromus tectorum L. and Taeniatherum caput-medusae L.) showed little change from
their original densities. Plant competition and other
environmental stress factors, including limited rainfall that can impact Mediterranean sage (Wilson et al.,
1994), contributed little to control before introduction
of P. tau. Some fluctuation in the density of Mediterranean sage is to be expected, but most sites under
good to excellent control have not returned to initial
pre-P. tau levels. We estimate that 68% of the original
release sites were controlled, averting an annual forage
loss of about $0.8 M over 508,000 ha.
Despite successful biological control at many sites
in south central Oregon, Mediterranean sage continues
to flourish and infest new and isolated areas in northern
Oregon and several western states. Although many of
those smaller infestations have been targeted for intensive control, P. tau has been released and recovered at
some. Impacts of P. tau in those areas should be monitored. The ability for Mediterranean sage to spread and
prosper at disturbed sites suggests that additional natural
enemies should be sought to improve control. Mediterranean sage is seed limited; therefore, seed destroying
biological control agents may prove more effective in
areas where P. tau is insufficient.
Without long-term monitoring, the biological control of Mediterranean sage may have been forgotten and
its efficacy never demonstrated. This can be a difficult
task when old biological control projects span across
the careers of multiple scientists, agency priorities and
funding commitments.
Acknowledgments
We acknowledge the assistance of Willy Riggs, Oregon
State University, Lake County Extension Service, for
access to old file records, and ODA personnel for their
assistance in the redistribution and monitoring efforts.
We also thank Linda Wilson, University of Idaho and
Joe Balciunas, USDA-ARS, Albany, CA, and Erin
McConnell, Lakeview Bureau of Land Management,
Oregon, Roger Sheley, USDA-ARS, Burns, Oregon,
for reviewing early versions of the manuscript. The
authors also wish to acknowledge the dedication and
professionalism of our colleague and friend, Charles E.
Turner, who passed away on April 15, 1997.
References
Andres, L.A. (1966) Host specificity studies of Phrydiuchus
topiarius and Phrydiuchus sp. Journal of Economic Entomology 59, 6976.
Andres, L.A. and Drea, J., Jr. (1963) Exploration for natural enemies of Salvia aethiopis L. and Linaria dalmatica
526
Biological control of Mediterranean sage (Salvia aethiopis) in Oregon
Miller in the Mediterranean and Near East, MayJune,
1962. Special report of the USDA-ARS Entomology Research Division, Insect Identification and Foreign Parasite Introduction Research Branch. US Department of
Agriculture, Beltsville, MD.
Andres, L.A. and Rizza, A. (1965) Life history of Phrydiuchus topiarius (Coleoptera: Curculionidae) on Salvia verbenacea (Labiatae). Annals of the Entomology Society of
America 58, 314319.
Andres, L.A., Coombs, E.M. and McCaffrey, J.P. (1995)
Mediterranean Sage, Salvia aethiopis L. (Lamiaceae). In:
Nechols, J. R., Andres, L.A., Beardsley, J.W., Goeden,
R.D. and Jackson, C.G. (eds) Biological Control in the
Western United States. University of California Division
of Agriculture and Natural Resources Publication 3361,
pp. 296298.
Bailey, L.H. (1935) The Standard Cyclopedia of Horticulture, vol. III. Macmillan, New York, USA.
Coombs, E.M. and Wilson, L. (2004) Mediterranean sage. In:
Coombs, E.M., Clark, J.K., Piper, G. L and Cofrancesco,
A.F. Jr. (eds) Biological Control of Invasive Plants in the
United States. Oregon State University Press, Corvallis,
USA, pp. 263267.
Coombs, E.M., Isaacson, D.L. and Hawkes, R.B. (1996)
The status of biological control of weeds in Oregon.
In: Delfosse, E. S. and Scott, R.R. (eds) Proceedings of
the VIII International Symposium Biological Control of
Weeds. Lincoln University, Canterbury, New Zealand, pp.
463471.
Davis, P.H. (1975) Flora of Turkey and east Aegean Islands,
vol. VII. Edinburgh University Press, Edinburgh, UK.
Dennis, L.R.J. (1980) Gilkeys Weeds of the Pacific Northwest. Oregon State University Press, Corvallis, USA,
pp. 245246.
Howell, J.T. (1942) Plants new to California. Leaflets Western Botany 3, 7980.
McEvoy, P.B. and Coombs, E.M. (2000). Why things bite
back: unintended consequences of biological weed control. In: Follett, P.A. and Duan, J.J. (eds) Nontarget Ef-
fects of Biological Control. Kluwer, Boston, MA, USA,
pp. 167194.
Mijatovic, K. (1973) A contribution to the study of distribution and ecology of Mediterranean sage (Salvia aethiopis
L.). Zastita Bilja 24,131146.
Mowrer K. (1996) Biological pest control section. Annual
Report. Colorado Department of Agriculture, Palisade,
CO, USA.
Radtke, H. and Davis, S. (2000) Economic analysis of containment programs, damage, and production losses from
noxious weeds in Oregon. Technical Report Prepared for
Oregon Department of Agriculture, Salem, USA.
Roche, C. T. and Wilson, L.M. (1999) Mediterranean sage.
In: Sheley, R.L. and Petroff, J.K. (eds) Biology and Management of Noxious Rangeland Weeds. Oregon State University Press, Corvallis, USA, pp. 261270.
Smith, H.S. and DeBach, P. (1942) The measurement of entomophagous insects on population densities of their hosts.
Journal of Economic Entomology 35, 845849.
Turner, C.E. and Coombs, E.M. (1996) Mediterranean sage.
In: Rees, N. E., Quimby, P.C., Piper, G.L., Turner, C.E.,
Coombs, E.M., Spencer, N.R. and Knutson, L.V. (eds) Biological Control of Rangeland Weeds in the West. Western
Society of Weed Science, Las Cruces, NM, USA.
Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M.,
Valentine, D.H., Walters, S.M. and D.Webb, D.A. (eds)
(1972) Flora Europaea, vol. III. Cambridge University
Press, Cambridge, UK.
Warner, R.E. (1969) The genus Phrydiuchus with description
of two new species (Coleoptera: Curculionidae). Annals
of the Entomological Society of America 62, 12931302.
Wilson, L.M. and McCaffrey, J.P. (1993) Bionomics of Phrydiuchus tau (Coleoptera: Curculionidae) associated with
Mediterranean sage in Idaho. Environmental Entomology
22, 704708.
Wilson, L.M., McCaffrey, J.P. and Coombs, E.M. (1994) Biological control of Mediterranean sage. Pacific Northwest
Extension Publication PNW 473. University of Idaho Cooperative Extension, Moscow, USA.
527
Preliminary results of a survey on the
role of arthropod rearing in classical
weed biological control
R. De Clerck-Floate,1 H.L. Hinz,2 T. Heard,3 M. Julien,4
T. Wardill5 and C. Cook6
Summary
The rearing of arthropods is an essential but sometimes neglected and underestimated part of a classical weed biological control programme. Success in rearing is usually a pre-requisite to conducting
host-specificity tests, obtaining enough individuals for initial field release or, later, for large-scale
implementation. Although most biological control researchers can list situations where agent development has been stopped or slowed due to rearing difficulties, failures seldom get reported in the
literature, thus preventing us from gauging the extent and relevance of rearing issues. To rectify this,
a questionnaire was developed to investigate the prevalence of rearing problems in weed biological
control programmes and to classify their occurrence according to a list of variables (e.g. taxonomy,
biological features, genetic issues and researcher/programme attributes). The questionnaire was sent
to 80 researchers from eight countries; 65% responded, generating 79 useful responses. Results confirm that, of the challenges faced in programmes, rearing is the most prevalent (56% out of ten possible general problem categories). The most common rearing problems encountered were conditions
that were not conducive to mating and/or oviposition (30% of reported arthropod cases) or development (22% of reported arthropod cases). Our results identify key areas for rearing improvement, thus
contributing to increased weed biological control project successes.
Keywords: international survey, project challenges, rearing difficulties.
Introduction
The ability to rear arthropods for classical weed biological control programmes is a topic seldom formally
addressed within our scientific discipline, yet it touches
all stages of a programme and has the potential to seriously affect its progress and direction. Rearing can
ensure a reliable source of arthropods for either hostspecificity testing or initial field releases when agents
Agriculture and Agri-Food Canada, Lethbridge Research Centre,
P.O. Box 3000, Lethbridge, AB, Canada T1J 4B1.
2
CABI Europe-Switzerland, Rue des Grillons 1, 2800, Delmont,
Switzerland.
3
CSIRO Entomology, 120 Meiers Rd, Indooroopilly 4068, Australia.
4
CSIRO European Laboratory, Campus International de Baillarguet,
34980 Montferrier sur Lez, France.
5
University of Sheffield, Department of Biomedical Science, Western
Bank, Sheffield S10 2TN, UK.
6
University of Queensland, School of Natural and Rural Systems Management, St. Lucia, QLD 4072, Australia.
Corresponding author: R. De Clerck-Floate <[email protected]>.
CAB International 2008
1
are rare in their place of origin (Blossey et al., 2000).
Furthermore, because host-range testing relies on arthropod behaviours that also are required for rearing
(e.g. mating, oviposition), the testing of these arthropods can be greatly hindered or prevented if they cannot be reared in a laboratory environment (Palmer,
1993; Marohasy, 1994; Klein, 1999). Sole reliance on
field-collected arthropods from the country of origin for
either host-specificity testing or field release is risky.
For instance, events that eliminate a field population of
arthropods used in testing could set back a programme
for years, at great cost. This could be averted if a reared
colony of the organism is available. There also is less
control over factors that may influence host choice (e.g.
prior host experience, readiness for host acceptance;
Marohasy, 1998) when field-collected arthropods are
used directly in host-specificity tests.
Once an arthropod has been given regulatory approval for release, the possession of a reared colony can
greatly aid in achieving successful establishment. This
is especially true for arthropods that may be susceptible
528
Preliminary results of a survey on the role of arthropod rearing in classical weed biological control
to Allee effects (e.g. ability to find mates), demographic
stochasticity or catastrophic environmental events if
released in small numbers (Hopper and Roush, 1993;
Grevstad, 1999). It also is preferable for the sake of risk
management to release agents from a reared colony of
known taxonomic identity, purity and genetic source to
reduce the possibility of unwanted inclusions such as
other species or parasitized or diseased individuals. Although exemptions are possible, a regulatory condition
of the Australian permitting system is that the agent
must undergo a minimum of one generation in containment before release to avoid such hazards. Other countries, or laboratories within countries, may also attempt
to follow this practice for precautionary reasons. Thus,
the inability to rear an agent in the laboratory environment may present a direct impediment to its release.
The ability to rear weed biological control agents
also influences the widespread implementation of successful agents. Typically, arthropods that are easy to
rear and demonstrate rapid population increases under
artificial conditions are the most amenable to mass production (Grossrieder and Keary, 2004). Compared to
the mass production of entomophagous biological control agents for augmentative use (Thompson, 1999), the
development of quality-controlled, mass rearing techniques for classical weed biological control arthropods
is still in its infancy. However, recently, there has been
a growing effort to experimentally develop methods
for laboratory mass rearing of weed biological control
agents on plants (Visalakshy and Krishnan, 2001) or
artificial diet (Blossey et al. 2000; Wheeler and Zahniser, 2001; Goodman et al., 2006; Raina et al., 2006)
or within scaled-up outdoor nurseries (Story et al.,
1994, 1996; Julien et al., 1999; De Clerck-Floate et al.,
2007).
Although examples of failures or difficulties in
arthropod rearing are well-recognized, these are seldom reported in the literature (however, see Palmer,
1993; Klein, 1999). The arthropods involved are either
dropped from or reduced in priority on candidate lists,
Table 1.
even if they show potential as effective agents. To determine the prevalence and types of rearing problems
within the field of classical weed biological control,
an international survey of researchers was conducted.
This paper summarizes key, preliminary results and
provides some perspective on how rearing can be managed to improve the science and implementation of
weed biological control.
Materials and methods
A questionnaire on the role of arthropod rearing in
classical weed biological control programmes was
prepared through collaboration and piloted among six
colleagues to ensure that the questions were clearly
stated for the information requested. The questionnaire
was divided into subsections: (1) general background
on the researcher name, location, years and type of
experience in classical weed biological control; (2)
programme-specific information (i.e. per weed) duration
of programme, number of arthropod species available
on the weed in its place of origin, number of arthropod species tested, released and established, indication
of avoidance of arthropod groups because of rearing
issues, role of release size on arthropod agent establishment, role, prevalence and type of mass rearing in
programmes, prevalence of monitoring for arthropod
quality in reared colonies, evidence of genetic management during rearing (i.e. measures taken to reduce likelihood of inbreeding), a detailed list of all arthropod
species either used or considered for biological control of
the weed, with information on taxonomy, feeding guild
and any problems encountered during development
or consideration of the agent. The arthropod speciesspecific information became a large and key component
of the data set and allowed analyses by family, order,
feeding guild, general problems encountered in the development of each agent (e.g. ability to rear; Table 1),
specific rearing problems, if they occurred (Table 2)
and whether a species was rejected or given lower
General problems encountered in classical weed biological control programmes and their frequency of occurrence.
General problems for weed biocontrol programmes
Ability to rear
Ability to establish agent in the field
Collecting sufficient specimens during exploration stage
Ability to elicit appropriate oviposition or feeding behaviours during
host-specificity testing
Detecting/confirming establishment at release sites
Inadequate taxonomy
Ability to accurately document agent impact on the target weed
Obtaining/collecting sufficient agents for general distribution
Other
None
Count
na
Frequency (%)
246
72
85
50
439
362
444
411
56
20
19
12
35
40
28
27
24
62
322
444
319
321
315
319
11
9
9
8
8
19
n for each problem varied depending on whether the programme stage allowed for encounter of the problem, and thus the inclusion of appropriate cases in the calculation of frequency.
529
XII International Symposium on Biological Control of Weeds
Table 2.
pecific problems encountered when rearing arthropods for classical weed biological control and their frequency
S
of occurrence (n = 246). Note that more than one problem could arise per arthropod case that was listed.
Rearing problems
Experimental conditions not conducive to mating and/or oviposition
Experimental conditions not conducive to normal development
Long life cycle
Obtaining appropriate phenological stage of host plant for rearing
Unknown biology or life cycle
Field collection of enough individuals to start a laboratory colony
Host plant nutrition/quality
Incompatibility with host plant
Mortality during storage of diapausing arthropods
Synchronization between northern and southern hemispheres
Presence of parasitoids
Presence of pathogens
Failure to break diapause
Difficulties with artificial rearing medium
Inadequate labour or facilities
Presence of predators
Inbreeding or genetic adaptation to laboratory conditions reducing quality of colony
Presence and possible effects of endosymbionts
Mutualist needed
Difficulties in distinguishing sexes
Other
Count
Frequency (%)
74
53
52
51
49
44
34
33
28
27
26
20
19
7
6
4
4
4
3
3
4
30
22
21
21
20
18
14
13
11
11
11
8
8
3
2
2
2
2
1
1
2
Table 3. T
he stages in classical weed biological control programmes at which rearing problems occurred
and the number and frequency of occurrence for each stage (n = 85).
Project stage
(A) Exploration
(B) Host-specificity testing
(C) Early release and monitoring for establishment
(D) Mass production
(E) General distribution of agents
(F) Monitoring for impact by agents
priority at any point because of rearing problems. Only
summary results from the survey are presented in this
paper due to the large volume of information that was
obtained. A journal publication presenting the broader
results is in preparation.
A total of 80 researchers from eight countries were
surveyed. Respondents were asked to complete a separate questionnaire for each weed species with which
they had been involved and to only answer questions
pertinent to their experience within programmes, as
identified by stage (Table 3). Programme stage was
taken into account during determination of the prevalence of various general problems or rearing problems
by including these data for summary or analysis only
if the researcher could have conceivably encountered
the problem(s) at their programme stage(s). This controlled for any bias due to stage of programme during
data summary. For example, General Problem, Ability
to accurately document agent impact (Table 1) would
Count
Frequency (%)
16
43
20
4
2
0
19
51
23
5
2
0
only be encountered at stage F of a programme (Table
3); thus, only agents in programmes that had reached
stage F were included in calculations involving this
problem.
Results and discussion
General results
A total of 52 researchers (65%) responded to the survey, generating 84 questionnaires, which represent the
majority of active weed biological control programmes
worldwide. The countries and number of researchers
based in each were Australia (16), South Africa (9),
USA (9), Switzerland (6), New Zealand (5), Canada
(4), France (2) and Argentina (1). The average experience in weed biological control of the researchers was
about 19 years, and the cumulative experience by the
respondents was 987 years. Most of the researchers
530
Preliminary results of a survey on the role of arthropod rearing in classical weed biological control
were based in the introduced range of the plant (54%),
some in the native range (21%), while others worked
in both (25%).
Of the 79 questionnaires that were included in the
analyses (five were not completed properly and were
excluded), the median duration of the weed programmes was 17 years. All programmes had completed
stage A, exploration and were at least in stage B, hostspecificity testing (see Table 3 for a list of stages). The
majority of programmes in the introduced range had
reached the final stage F (monitoring for impact). In
65% of cases, respondents reported work on a unique
target weed, while in 35% of cases, the same species
was reported from another country (i.e. work within
native vs introduced ranges or two or more countries
working on the same weed) or at a different programme
stage. Specifically, the same weed was reported twice
in eight cases and three times in four cases.
The analysed responses covered a total of 64 weed
species located in various climatic zones including;
temperate (32 species), temperate/subtropical (5), subtropical (11), subtropical/tropical (9), tropical (6) or temperate/subtropical/tropical (one aquatic species). The
majority of target weeds were perennials (52 species)
with a few biennials (7) or annuals (5). The number of
Figure 1.
herbivorous arthropod species occurring on a weed in
its area of origin was estimated through the literature,
field surveys or both by respondents for 45 of the target
weeds. The average number of herbivores was 120 but
generally increased from temperate to tropical zones
(Fig. 1), which is an expected trend based on studies
of arthropod diversity along latitudinal gradients (Price,
1997).
Information on a total of 384 arthropod species (nine
eriophyoid mites and 375 insect species) were recorded
in the questionnaires. The insects were from 6 orders
and 66 families. A few of the arthropods were used for
more than one weed target, resulting in a total of 444
cases for data analysis.
Incidence of rearing problems
Respondents reported many general problems in
the development of arthropod agents for weed biological control; by far, the most common was ability to rear with 56% of the arthropod cases listed in
questionnaires (Table 1). Other common problems,
such as collecting sufficient specimens during exploration stage and ability to establish agent in the
field, may also be related to the ability to rear, as it is
Mean number (SE) of herbivore arthropod species estimated (from field surveys, the
literature or both) to occur on a target weed species in its area of origin according to
climatic zone. Te/ST Temperate/subtropical, ST/Tr subtropical/tropical.
531
XII International Symposium on Biological Control of Weeds
suspected that they may not have attained this level of
relevance if rearing were easy. For instance, there is
evidence that releases at more sites and/or with larger
numbers of individuals per release in a geographic area
improves establishment success (Hopper and Roush,
1993; Memmott et al., 1998; Grevstad, 1999). It has
also been acknowledged that difficulties in rearing
may jeopardize agent establishment where there are
known hazards to small, founding, field populations
(Spafford Jacob et al., 2007). In our questionnaire,
73% (174/237) of answered cases indicated that there
were examples within programmes where it was suspected that failure to establish may have been due to
small release size.
Specific rearing problems experienced by researchers can be grouped into several categories. The most
common were conditions that were not conducive to
mating and/or oviposition or normal development of
the immature stages (Table 2). Often, appropriate artificial light and temperature conditions must be experimentally identified to induce arthropod behaviours
required for successful rearing (Baars, 2001). In addition, very common were problems due to long or unknown life cycles or unknown biology. Issues related
to the host plant such as nutrition, quality, incompatibility or providing the correct phenological stage were
very common. Obtaining enough individuals to start a
colony was a common problem as were problems with
synchronization and diapause. Issues associated with
organisms such as parasitoids, pathogens, predators,
endosymbionts and mutualists were reported less frequently. In some situations, arthropods reared under
laboratory conditions simply did not thrive for obscure
reasons, which require further investigation (Hoffmann, 1988). For instance, among respondents to this
survey, only 2% of cases had inbreeding or genetic
adaptation to laboratory conditions reducing quality of
colony listed as a rearing problem (Table 2); yet, when
they were specifically asked whether they monitor for
arthropod quality (e.g., fecundity, fertility) during rearing, only 39% (170/430) of cases were affirmative, and
in 44% (190/430) of cases, agents were not monitored
at all. This lack of quality control clearly indicates a
reduced possibility of identifying genetic problems in
rearing, even though they are known to occur (Torres
et al., 1991; Wardill et al., 2004). Changes in fecundity
or fertility also may indicate the effect of endosymbionts such as Wolbachia spp., which are seldom considered in biocontrol programmes (Floate et al., 2006).
The stage in a project in which an arthropod was
rejected or given lower priority because of rearing
problems was recorded for only 85 cases out of 444
(Table 3). Despite the poor response to this question,
the results still show that when decisions are clearly
made due to rearing difficulties, most rejections or priority adjustments occur during the host-specificity testing stage. Of note is that 20 species were rejected at
the release stage and four at the mass-production stage.
These 24 cases may represent very costly situations, as
large programme expenses were likely already incurred
at the time of rejection.
Rejection or priority adjustment also occurred during exploration when rearing is required for identification, biology and life-history studies and sometimes
preliminary host-range tests (Table 3). After the survey,
it occurred to us that there may have been a blurred
boundary between exploration and host testing, thus inflating the relevance of the host-specificity testing stage
as a time when rearing issues surfaced. For instance, if
exploration were conducted through quick, infrequent
visits to the place of weed origin by remotely situated
researchers, rather than from a more permanent base
located in the place of origin, then all work conducted
in the recipient country may have been grouped under
host testing. This ability to clearly distinguish the
boundary between these two stages may be a limitation
of the survey.
Importantly, identification of rearing difficulties in
the early stages of a programme allows for either the
deployment of resources to develop needed rearing
techniques or the rapid adjustment of programme goals,
so that resources can instead be allocated to candidates
that are easier to rear (Hoffmann, 1988). Although details on the taxonomic affiliations of difficult-to-rear
arthropods will be forthcoming in a future publication,
27% (104/379) of answered cases in our survey indicated that certain taxonomic groups or feeding guilds
were initially either avoided or given lower priority due
to perceived potential rearing problems. Interestingly,
there was a positive linear relationship between the
estimated number of species feeding on a weed in its
place of origin and the number of agents tested by researchers (R2 = 0.605, P = <0.001, n = 364). However,
there was no correlation between the estimated number of species and the incidence of rearing difficulties
(R2 = 0.003, P = 0.287, n = 381). Thus, there was no indication that, if researchers have more insects to choose
from as potential agents, they tend to test those that are
easier to rear, or it could quite simply mean that we are
poor at predicting at the start of a programme which
arthropods will be easiest to rear.
Conclusions
The preliminary results of our survey indicate that
arthropod rearing issues are perceived to be important
among classical weed biological control researchers.
They have a very important impact in the development
of successful agents and therefore on the outcome of
programmes. Despite this, there is ample opportunity
to improve rearing efforts and methodology.
Where possible, we recommend that rearing be incorporated into the exploration stage of programmes,
as suggested by Hoffmann (1988), or as very early
studies in the recipient country in cases where exploration is practiced as quick visits for arthropod collec-
532
Preliminary results of a survey on the role of arthropod rearing in classical weed biological control
tion by remote researchers. Should rearing difficulties
occur, emphasis could then be placed on developing
laboratory-based rearing techniques. Only then should
otherwise promising species with apparently insoluble
rearing problems be given lower priority or abandoned
as candidate agents before major costs are incurred. The
fact that 27% of respondents avoid taxonomic groups
that have perceived potential rearing problems may be
good for immediate resource management. However,
unless species in the troublesome groups are studied,
this avoidance will not lead to improved rearing capability, and potentially useful agents may be rejected
prematurely.
Acknowledgements
The following researchers are acknowledged for their
useful responses to the questionnaire: Adair, R., Batchelor, K., Cuda, J., Day, M., R, Ding, J., Flores, D., Forno,
W., Fowler, S., Gassmann, A., Gerber, E., Gordon, A.,
Grosskopf, G., Hansen, R., Harris, P., Hill, M., Hill, R.,
Hoffman, J., Impson, F., Ireson, J., Klein, H., Kok, L.,
Littlefield, J., Lockett, C., McClay, A., McFadyen, R.,
McKay, F., Nod, K., Olckers, T., Palmer, B., Peschken,
D., Pitcairn, M., Purcell, M., Ragsdale, D., Sagliocco,
J.L., Schaffner, U., Sheppard, A., Smith, L., Smyth, M.,
Sobhian, R., Story, J., Urban, A., van Klinken, R., Williams, H., Winks, C., Wright, T., Yeoh, P., Zachariades,
C. and Zwelfer H. We also acknowledge data entry
by C. Clech-Goods and the helpful statistical advice of
T. Entz.
References
Baars, J.R. (2001) Biology and laboratory culturing of the
root-feeding flea beetle, Longitarsus columbicus columbicus Harold, 1876 (Chrysomelidae: Alticinae): a potential
natural enemy of Lantana camara L. (Verbenaceae) in
South Africa. Entomotropica 16, 149155.
Blossey, B., Eberts, D., Morrison, E. and Hunt, T. (2000)
Mass rearing the weevil Hylobius transversovittatus
(Coleoptera: Curculionidae), biological control agent of
Lythrum salicaria, on semiartificial diet. Journal of Economic Entomology 93, 16441656.
De Clerck-Floate, R.A., Moyer, J.R., Van Hezewijk, B.H.
and Smith, E.G. (2007) Farming weed biocontrol agents:
a Canadian test case in insect mass-production. In: Clements, D.R. and Darbyshire, S.J. (eds) Invasive Plants:
Inventories, Strategies and Action. Topics in Canadian
Weed Science, vol. 5. Canadian Weed Science Society
Socit canadienne de malherbologie, Sainte Anne de
Bellevue, Qubec, pp. 111130.
Floate, K.D., Kyei-Poku, G.K. and Coghlin, P.C. (2006)
Overview and relevance of Wolbachia bacteria in biocontrol research. Biocontrol Science and Technology 16,
767788.
Goodman, C.L., Phipps, S.J., Wagner, R.M., Peters, P., Wright,
M.K., Nabli, H., Saathoff, S., Vickers, B., Grasela, J.J.
and McIntosh, A.H. (2006) Growth and development of
the knapweed root weevil, Cyphocleonus achates, on a
meridic larval diet. Biological Control 36, 238246.
Grevstad, F.S. (1999) Factors influencing the chance of population establishment: Implications for release strategies in
biocontrol. Ecological Applications 9, 14391447.
Grossrieder, M. and Keary, I.P. (2004) The potential for the
biological control of Rumex obtusifolius and Rumex crispus using insects in organic farming, with particular ref
erence to Switzerland. Biocontrol News and Information
25, 65N79N.
Hoffmann, J.H. (1988) Pre-release assessment of Nanaia sp.
(Lepidoptera: Phycitidae) from Opuntia aurantiaca for
biological control of Opuntia aurantiaca (Cactaceae).
Entomophaga 33, 8186.
Hopper, K.R. and Roush, R.T. (1993) Mate finding, dispersal,
number released, and the success of biological control introductions. Ecological Entomology 18, 321331.
Julien, M.H., Griffiths, M.W. and Wright, A.D. (1999) Biological Control of Water Hyacinth. The Weevils Neochetina bruchi and N. eichhorniae: Biologies, Host Ranges,
and Rearing, Releasing and Monitoring Techniques for
Biological Control of Eichhornia crassipes. ACIAR
Monograph No. 60. Centre for International Agricultural
Research, Canberra, 87 pp.
Klein, H. (1999) Biological control of three cactaceous
weeds, Pereskia aculeate Miller, Harrisia martini (Labouret) Britton and Cereus jamacaru de Candolle, in South
Africa. African Entomology Memoir 1, 314.
Marohasy, J. (1994) Biology and host specificity of Weiseana barkeri (Col.: Chrysomelidae): A biological control
agent for Acacia nilotica (Mimosaceae). Entomophaga
39, 335340.
Marohasy, J. (1998) The design and interpretation of hostspecificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News and
Information 19, 13N20N.
Memmott, J., Fowler, S.V. and Hill, R.L. (1998) The effect of
release size on the probability of establishment of biological control agents: gorse thrips (Sericothrips straphylinus)
released against gorse (Ulex europaeus) in New Zealand.
Biocontrol Science and Technology 8, 103115.
Palmer, W.A. (1993) A note on the host specificity of the mirid Slaterocoris pallipes (Knight). Proceedings of the Entomological Society of Washington 95, 640641.
Price, P.W. (1997) Insect Ecology, 3rd edn. Wiley, New York,
USA, 874 pp.
Raina, A., Gelman, D., Huber, C. and Spencer, N. (2006)
Laboratory rearing procedures for two lepidopteran weed
biocontrol agents. Florida Entomologist 89, 9597.
Spafford Jacob, H., Rielly, T.E. and Batchelor, K.L. (2007)
The presence of Zygina sp. and Puccinia myrsiphylli reduces survival and influences oviposition of Crioceris sp.
Biocontrol 52, 113127.
Story, J.M., Good, W.R. and White, L.J. (1994) Propagation
of Agapeta zoegana L. (Lepidoptera, Cochylidae) for biological control of spotted knapweed: Procedures and cost.
Biological Control 4, 145148.
Story, J.M., White, L.J. and Good, W.R. (1996) Propagation
of Cyphocleonus achates (Fahraeus) (Coleoptera: Curculionidae) for biological control of spotted knapweed: procedures and cost. Biological Control 7, 167171.
Thompson, S.N. (1999) Nutrition and culture of entomophagous insects. Annual Review of Entomology 44, 561592.
533
XII International Symposium on Biological Control of Weeds
Torres, D.O., Vargas, D.G., Ritual, S.M. and Alforja, E.M.
(1991) Studies on causes of infertility of Pareuchaetes
pseudoinsulata Rego Barros. In: Muniappan, R. and
Ferrar, P. (eds) Proceedings of the Second International
Workshop on Biological Control of Chromolaena odorata, 48 February. BIOTROP Special Publication 44,
Bogor, Indonesia, pp. 113119.
Visalakshy, P.N.G. and Krishnan, S. (2001) Propagation
methods of Ceutorhynchus portulacae, a potential biocontrol agent of Portulace oleracea L. Procedures and
cost. Entomon 26, 7985.
Wardill, T.J., Graham, G.C., Manners, A., Playford, J., Zalucki, M., Palmer, W.A. and Scott, K.D. (2004) Investigating genetic diversity to improve the biological control
process. In: Sindel, B.M. and Johnson, S.B. (eds) Proceedings of the 14th Australian Weeds Conference. Weeds
Society of New South Wales, Sydney, pp 364367.
Wheeler, G.S. and Zahniser, J. (2001) Artificial diet and
rearing methods for the Melaleuca quinquenervia (Myrtales: Myrtaceae) biological control agent Oxyops vitiosa
(Coleoptera: Curculionidae). Florida Entomologist 84,
439441.
534
Beginning success of biological control
of saltcedars (Tamarix spp.) in the
southwestern USA
C.J. DeLoach,1 P.J. Moran,2 A.E. Knutson,3
D.C. Thompson,4 R.I. Carruthers,5 J. Michels,6
J.C. Herr,5 M. Muegge,7 D. Eberts,8 C. Randal,9
J. Everitt,10 S. OMeara8 and J. Sanabria11
Summary
Saltcedars, Tamarix spp., exotic, invading deciduous shrubs or small trees from Asia and the Mediterranean area, have become the most damaging weeds of riparian areas in the western USA. We and
our cooperators have obtained highly successful initial control of saltcedar by introducing the north
Asian leaf beetle (Diorhabda sp., China/Kazakhstan ecotype) at five sites north of the 38th parallel,
but they failed to establish farther south. In 2001, we discovered southern-adapted Diorhabda beetles
on saltcedar and began testing them. In 2004, we released 2408 Greek beetles at Big Spring, TX; by
September 2004, they had defoliated two trees, by 2005, 210 trees (0.8 ha) and by 2006, 7.3 ha of the
saltcedar stand and 1.4 ha at Cache Creek, California, and had begun defoliating saltcedar at Pecos
and Imperial, TX. The Uzbek beetles are increasing rapidly at Lake Meredith, TX and Fukang, China
beetles at Artesia, NM, but the Greek and Tunisian beetles have not established near Kingsville in
south Texas. We have revised the taxonomy of the five Tamarix-feeding Diorhabda ecotype/sibling
species and predicted their climatic affinities in North America, correlated depletion of stored carbohydrates by beetle defoliation with plant death, developed pheromone attractants, remote sensing, improved release methods and a model of beetle dispersal and estimated possible damage to beneficial
T. aphylla (Linnaeus) Karsten (athel) in the open field.
Keywords: saltcedar, athel, Tamarix, biological control weeds, Diorhabda.
Introduction
Western North American riparian ecosystems, beginning in the mid-1800s, have been invaded by exotic
saltcedars (Tamaricales: Tamaricaceae: Tamarix spp.),
shrubs or small trees native in Asia and the Mediterranean area (Baum, 1978). Three species have become
serious weeds, Tamarix ramosissima Ledebour, Tamarix chinensis Loureiro and hybrids between them and
US Department of Agriculture, Agricultural Research Service (USDAARS), Grassland, Soil and Water Research Laboratory, 808 E. Blackland Road, Temple, TX 76502, USA.
2
USDA-ARS, Beneficial Insects Research Unit, 2413 E. Highway 83,
Weslaco, TX 78596, USA.
3
Texas A&M University (TAMU) and Texas Agricultural Experiment
Station (TAES), 17360 Coit Road, Dallas, TX 75252, USA.
4
New Mexico State University (NMSU), Department of Entomology,
Plant Pathology and Weed Science, Las Cruces, NM 88003, USA.
5
USDA-ARS, Exotic and Invasive Weeds Research Unit, 800 Buchanan
St., Albany, CA 94710, USA.
CAB International 2008
1
other species in the western USA and northern Mexico and Tamarix parviflora de Candolle in California
(Gaskin and Schaal, 2002, 2003). In addition, Tamarix
canariensis Willdenow and Tamarix gallica Linnaeus
(identification uncertain) and their hybrids with T. ramosissima and/or T. chinensis occur along the Gulf of
Mexico coast from Louisiana to Port Isabel, TX but appear less invasive. Tamarix aphylla (Linnaeus) Karsten
(athel) is a large evergreen, cold-intolerant tree that
TAMU and TAES, P.O. Drawer 10, Bushland, TX 79012, USA.
TAMU and TAES, Airport Drive, P.O. Box 1298, Fort Stockton, TX
79735, USA.
8
US Department of Interior (USDI), Bureau of Reclamation, Ecological
Services Center, Denver Federal Center, P.O. Box 25007 (86-68220),
Denver, CO 80225, USA.
9
USDA-APHIS-PPQ, Olney, TX 76374, USA.
10
USDA-ARS, Integrated Farming and Natural Resources Research Unit,
2413 E. Highway 83, Weslaco, TX 78596, USA.
11
TAES, 720 E. Blackland Road, Temple, TX 76502, USA.
Corresponding author: C.J. DeLoach <[email protected].
gov>.
6
7
535
XII International Symposium on Biological Control of Weeds
is less invasive and is used for shade and windbreaks
mostly in desert areas of northern Mexico and the southern third of the adjoining states in the United States; it is
not presently a target for biological control. Saltcedars
occupy approximately 800,000 ha of prime bottomlands from Montana into northern Mexico where they
severely damage native plant and animal communities
and rangeland ecosystems, including that of many endangered or threatened species (DeLoach et al., 2000).
Investigations into biological control began in 1986
by the US Department of Agriculture, Agricultural Research Service (USDA-ARS) at Temple, TX (DeLoach),
and from 1991 with our overseas cooperators, R. Jashenko and I.D. Mityaev (Kazakhstan), R. Wang, Q.G.
Lu, B.P. Li and H.Y. Chen (China), S. Myartseva (Turkmenistan), D. Gerling (Israel), A. Kirk, R. Sobhian and
L. Fornasari (ARS, European Biological Control Laboratory (EBCL, France) and J. Kashefi (EBCL, Greece;
DeLoach et al. 2004; Carruthers et al., 2008). Carruthers
joined the project in 1998, followed by the other cooperators, especially after field releases began. This collaborative project includes scientists and their co-workers at
each of the release sites and for specific research.
generations with little fertility decline and Karshi beetles hybridized with both Crete and Fukang beetles, albeit with reduced fitness in both crosses. DNA analyses
supported species status for all five ecotypes examined
(D. Kazmer, unpublished data). Pheromone analyses
revealed that an aggregation pheromone produced by
the males consisted of two complex alcohols, which
varied in ratio between the four ecotypes from 1:1 to
1:30 (Coss et al., 2005; A. Coss, unpublished data).
Differences in pheromones between species may partly
explain why Tracy did not find hybrids among the
many specimens examined from the Old World: The
species with different pheromones were not attracted to
each other and did not mate but may mate and produce
hybrids when confined in cages. Both adults and larvae
of Diorhabda feed and females oviposit on the foliage
of Tamarix. The adults overwinter and the larvae pupate, under litter on the soil surface. The larvae have
three instars and the life cycle requires approximately
34 days during summer. The beetles have two generations a year in areas north of the northern 38th parallel
(Lewis et al., 2003b; Bean et al., 2007a,b) and three to
five further south (Milbrath et al., 2007).
Release, impact and establishment
of the China/Kazakhstan ecotype
of Diorhabda beetles
Taxonomy and biology of
Diorhabda spp.
The first natural enemy chosen for introduction and
testing in 1992 was the leaf beetle, Diorhabda sp. (Coleoptera: Chrysomelidae), from Fukang in northwestern
China and Chilik in eastern Kazakhstan. In 2002, other
more southern-occurring Diorhabda ecotypes were
collected. All were identified by leading taxonomic experts as Diorhabda elongata (Brull). Male and female
genitalia were dissected from these collections and from
some 700 museum specimens from over 370 locations in
37 countries of Asia and the Mediterranean area. J. Tracy
(personal observation) found that the specimens fell
into the five following ecotypes, corresponding to five
previously named species or subspecies (later synonymyzed but which Tracy reassigned as five species; with
probable areas of best adaptation in North America):
(1) Fukang and Turpan, China and Chilik, Kazakhstan
ecotype (deserts north of the 33rd northern parallel); (2)
Crete and Posidi, Greece ecotype (Mediterranean climate CA, central TX); (3) Sfax, Tunisia ecotype (southern TX, southern coastal CA, Sonoran Desert, northern
Mexico); (4) Karshi and Bukhara, Uzbekistan ecotype
(plains grasslands of Oklahoma, New Mexico, TransPecos, TX; Mojave Desert) and (5) coastal Iran ecotype
(southern TX, Sonoran Desert), not yet imported.
Tracys species designations also are supported by
other evidence. Cross-mating experiments that measured egg fertility [Thompson (with K. Gardner), New
Mexico State University, Las Cruces (NMSU)], agreed
with most of Tracys designations except that the beetles from Crete and Tunisia mated freely for several
The biological control program incurred a 6-year
delay from the time we were prepared to release the
Diorhabda beetles in June 1995 until actual release in
May 2001. This was caused by the listing of the southwestern willow flycatcher, Empidonax trailii Audubon ssp. extimus Phillips (which had begun nesting in
saltcedar in recent years, mostly in Arizona), as federally endangered (USDI-FWS, 1995). This required the
submission of a biological assessment (DeLoach and
Tracy, 1997) and consultation with the US Department
of Interior-Fish and Wildlife Service (USDI-FWS). The
resulting Letter of Concurrence allowed releases only
at the ten specified sites in six states, only farther than
160 km from where the flycatcher nested in saltcedar.
The first releases were to be in field cages for 1 year and
extensively monitor beetle populations and dispersal,
damage to saltcedar and any non-target plants and effects on native vegetation and wildlife communities.
Populations of this flycatcher have increased greatly,
especially at Elephant Butte Reservoir in central New
Mexico and at Roosevelt Lake in central AZ after the
mid-late 1990s, as willows increased after water level
changes, to become among the largest breeding flycatcher populations known, nesting mostly in native
willows (Sferra et al., 1997; Moore and Ahlers, 2005).
Overseas observations and quarantine testing at
Temple, TX since 1992 and at Albany, CA since 1998
had demonstrated that these beetles were attracted to,
fed and completed their life cycle only on Tamarix and
536
Beginning success of biological control of saltcedars (Tamarix spp.) in the southwestern USA
rarely on related species of Frankenia (order Tamaricales, family Frankeniacae) on which the beetles probably cannot sustain a population in nature (Lewis et
al., 2003a; Dudley and Kazmer, 2005). These results
led to agreements to release from FWS and the six state
Departments of Agriculture and release permits from
APHIS-PPQ.
These beetles were released in cages in 1999 and in
the open field in May 2001 at ten sites in six western US
states: California, Nevada, Utah, Colorado, Wyoming
and Texas (DeLoach et al., 2004). They established easily at five of the six sites north of the northern 38th parallel and, by 2006, had defoliated 30,000 ha of saltcedar
at Lovelock, NV, 2000 ha at Schurz, NV Delta, UT, and
Lovell, WY, and 150 ha (limited by surrounding herbicide and insecticide applications) at Pueblo, CO. However, these beetles did not establish at any of the four
sites farther south in California or Texas or farther north
in Montana or Oregon. At Lovelock, NV, 40% of the
trees died where defoliation occurred twice annually
for 4 years and 65% died where defoliated for 5 years
(Carruthers et al., 2008; T. Dudley, personal communication; J. Knight, personalcommunication), caused by
gradual reduction in carbohydrate reserves (Hudgeons
et al., 2007). Tests by Lewis et al. (2003b) and Bean et
al. (2007a) revealed that the lack of establishment in
the southern areas was caused by a requirement of the
beetles for 14.5-h summer-day length to avoid premature diapause. This day length occurs in the northern
areas but is not reached south of the 38th parallel.
Discovery, testing, release and
impact of southern-adapted beetles
Carruthers, with J. Kashefi (ARS-EBCL, Greece),
discovered Diorhabda beetles on Tamarix in Crete in
2001. They were also found in Tunisia (A. Kirk, personal communication) and Uzbekistan (R. Sobhian,
personal communication) in 2002. Testing by Bean et
al. (2007b) revealed that all of these beetle populations
were adapted to the southern areas of California, New
Mexico and Texas where day lengths are shorter. Quarantine tests at Temple (Milbrath and DeLoach, 2006a,b)
and at Albany (Herr, unpublished data) revealed that
all are host-specific to Tamarix as is the Fukang/Chilik
ecotype. Approvals were obtained from FWS, Texas,
New Mexico and California Departments of Agriculture and APHIS-PPQ, and releases began in 2004.
Releases, establishment and impact
through 2007
The first southern-adapted beetles to establish (from
Crete) were released by DeLoach (with J. Tracy and
T. Robbins) near Big Spring, TX. In April 2004, they
released 37 adults and another 171 in July, that with
their offspring, defoliated two small trees near the nursery cage in mid-July. Then, they released another 2200
adults in August that together defoliated a large tree near
the cage by October 2004. By October 2005, the beetles
had defoliated 210 trees covering 0.8 ha, by July 2006,
2.4 ha, and by October 2006, 10 ha.By October 2007,
the beetles had dispersed 8.5 km and defoliated trees for
7 km along nearby Beals Creek. Remote sensing by lowlevel, aerial color photography by Everitt et al. (2007)
clearly shows beetle-defoliated areas, permitting estimates of the areas of canopy that have been defoliated.
A mathematical model is under development to describe
the advancing wave-form of beetle dispersal during each
generation (J. Sanabria, personal communication).
The Crete ecotype also was released at 20 other locations in the Colorado River watershed in west Texas
during 2005 to 2007 as part of a study by Knutson
and Muegge (with E. Bynum) to develop optimum release methods. The beetles appear to be established at
some sites. Along the Pecos River, they released approximately 500 adults from a nursery cage near each
of Pecos and Imperial, TX, in early August 2006; by
fall 2007, the beetles had defoliated approximately
500 trees along 1 km of the river near Pecos and had
dispersed along 400 m of the river near Imperial. The
Crete beetles failed at Seymour and Kingsville, TX, and
increased and then declined at Lake Meredith, TX, and
Artesia, NM. Crete beetles released in 2004 at Cache
Creek (near Rumsey, CA) by Carruthers and Herr (unpublished data) established on T. parviflora in that area,
increased rapidly during 2006, and by October 2007 had
defoliated nearly all saltcedar for 32 km along the creek.
In the meantime, Michels (with V. Carney and E.
Jones) released the Diorhabda ecotype from Posidi
Beach (near Thessaloniki), Greece at Lake Meredith,
TX, where they increased initially but then declined.
The Uzbek ecotype, also released at this site, is increasing. Releases of Posidi beetles near Carlsbad, NM by
OMeara, and the Turpan China ecotype in southeastern Colorado by Eberts failed, probably because the
sites flooded. The Fukang beetles released at Artesia,
NM by Thompson (with K. Gardner) have adapted to
shorter day lengths and appear to have established. Tunisian beetles were released in separate areas from the
Crete beetles by Moran near Kingsville; both Crete and
Tunisia beetles reproduced well in field nursery cages
and in sleeve bags but failed to establish in the open
field at this site.
Possible causes of failure to establish
The strong influence of short summer day length
on induced premature diapause and failure to overwinter in southern areas was demonstrated by Bean et al.
(2007a,b) with the Fukang/Chilik ecotype. In addition,
distributional patterns of Diorhabda appear to be influenced by biome differences (e.g. desert versus maritime
and continental versus maritime climates), latitude and
elevation (J. Tracy, personal communication). Also important is the selection of sites unlikely to flood when
537
XII International Symposium on Biological Control of Weeds
pupae or overwintering adults are on the soil surface
and, of course, that will not receive herbicide or insecticide applications, mechanical controls or be burned or
harmed by human disturbance. Attack on the beetles by
arthropod predators in the trees or by ground beetles,
mice, etc. on the ground may have caused some failures,
which might be improved by selecting sites with simple
vegetation communities that produce fewer insects that
attract predators. At some locations, the rapid rate of
increase of the Diorhabda beetles such as at Cache
Creek, California, has allowed them to overcome predator attack. In addition, numbers released (which may
influence production of sufficient concentrations of
the beetle aggregation pheromone), cage type, time of
year and the match of Diorhabda ecotype/species and
Tamarix species/hybrids present all seem important to
successful establishment. Differences in pheromone
concentrations could explain differences in reproduction in cages vs field. At some sites, beetles confined
in cages mated and larval populations developed well,
but after being released into the open field, the adults
dispersed widely and the pheromone concentration may
have been too low to allow mate finding. If populations
are high, the beetles often migrate in swarms, which
may maintain the necessary pheromone levels.
Releases: Rio Grande, western Texas/
Mexico and non-target effects on athel
The largest and most damaging stands of saltcedars are along the Colorado River of Colorado, Utah,
Arizona, California, and Mexico and the Rio Grande of
New Mexico, Texas, and Mexico. We refrained from
releasing Diorhabda sp. along the USMexican border
because of Mexican concerns, including the potential
for beetle attack on non-target athel (T. aphylla), which
the Diorhabda beetles had attacked to some extent in
our earlier laboratory and field-cage tests (Lewis et al.,
2003a; Milbrath and DeLoach, 2006a,b; Herr et al., unpublished data). However, in preparation for possible
agreement with Mexico for the release of the beetles
along the Rio Grande, DeLoach (unpublished data) conducted overwintering cage tests along the Rio Grande
between Presidio and Candelaria in western Texas with
the Crete, Tunisia and Uzbek Diorhabda ecotypes, beginning in November 2006. All populations overwintered, although survival of the Crete beetles was slightly
higher. Unfortunately, we then discovered that the Tunisian and Uzbek ecotypes had hybridized with the Crete
ecotype in the outdoor cage cultures maintained at Temple, TX, and we destroyed these beetles before release.
Uncaged, open-field tests were conducted at two locations in Texas to compare Diorhabda host selection
for and damage to saltcedar, athel and Frankenia. At
Big Spring in 2005, approximately 1- to 1.5-m-tall potted athel and local saltcedar and 10-cm-tall Frankenia
plants were transplanted together in ten 1-m-diameter
plots into an old saltcedar stand (46 m tall) being de-
foliated by Crete beetles released the year before. At
Kingsville, Moran released over 6000 Crete and Tunisia beetles from 2004 to 2006 into an isolated stand
of saltcedar previously inter-planted with 25 athel trees
(both 34 m tall) and also onto large (1012 m tall)
roadside athel trees located 40 to 60 km away with no
saltcedar present (no beetles were previously present
at either site). The saltcedar at Kingsville is a hybrid
between T. canariensis, T. gallica, T. ramosissima or
T. chinensis, to which the beetles were not strongly attracted in outdoor-cage tests at Temple (Milbrath and
DeLoach, 2006 a,b).
At Big Spring, the beetles attacked only the saltcedar transplants from June to late August, but then,
when populations increased to high levels, they attacked both saltcedar and athel. In 2005, DeLoach et al.
(unpublished data) counted a total of 1711 adults on
the saltcedar test plants, 588 on athel (34.5% of the total adults counted) and four adults on Frankenia. They
also counted 170 egg masses on saltcedar, 30 on athel
(15.0% of the total) and none on Frankenia. At Kingsville, Moran (unpublished data) found that Crete and
Tunisia beetles developed and reproduced in sleeve
bags on both saltcedar and athel but laid two- to fourfold more eggs on saltcedar. When adults were released
from the sleeve bags, they laid two- to tenfold more
eggs and produced many more larvae on saltcedar than
on athel within the first 2 weeks of release and never
established on athel. However, the beetles did not establish on saltcedar at Kingsville either, probably because the T. canariensis hybrid dominant there was less
attractive to them. When several hundred adults were
placed in sleeve bags on the roadside athel plants, they
produced larvae that defoliated branches inside sleeve
bags, but when released, they simply flew away, apparently in search of a more attractive saltcedar species or
hybrid, and never established a population on athel. In
similar open-field tests at Cache Creek, California, in
2007 (Herr and Carruthers, unpublished data), the Crete
beetles defoliated potted F. salina plants placed among
the resident T. parviflora stand during August when
the beetles reached high populations and defoliated the
T. parviflora, but the F. salina plants later recovered.
These open-field tests confirmed the cage-test results
of Lewis et al. (2003a), Milbrath and DeLoach (2006a,b)
and Herr et al. (unpublished data) that the female beetle
searching for an ovipositional host plant is the most highly selective life stage. Adult plant selection for alighting/
feeding or larval development is less specific. These results suggest that, if athel grows within a saltcedar stand,
adults will alight on athel, but females will deposit fewer
eggs than on saltcedar, leading to smaller larval populations and less damage to athel than to saltcedar. In the
absence of a choice, i.e. if athel is spatially isolated from
saltcedar or occurs in the midst of a defoliated saltcedar
stand, the beetles would be unlikely to establish populations on athel or to cause damage sufficient to interfere
with its use as a shade or windbreak tree.
538
Beginning success of biological control of saltcedars (Tamarix spp.) in the southwestern USA
We met three times with Mexican scientists and officials, and in June 2007, they agreed not to oppose our
releases along the Rio Grande of west Texas. Therefore,
on 29 June 2007, we released the Crete beetles from
seven overwintering cages on five private ranches. So
far, increases in beetle populations have been slow and
most sites have flooded, but the beetles have survived at
most sites, and at three sites populations have increased
to a few hundred or a few thousand near the cages, suggestive of establishment. In general, the establishment of
the southern-adapted beetles in the field has had a lower
rate of success and lower rates of increase and dispersal than that of the Fukang/Chilik ecotype released in
the northern areas. However, damage to saltcedar still is
sufficient to promise successful biological control.
References
Baum, B.R. (1978) The Genus Tamarix. Israel Academy of
Sciences and Humanities, Jerusalem, 209 pp.
Bean, D.W., Dudley, T.L. and Keller, J.C. (2007a) Seasonal
timing of diapause induction limits the effective range of
Diorhabda elongata deserticola (Coleoptera: Chrysomelidae) as a biological control agent for tamarisk (Tamarix
spp.). Environmental Entomology 36, 1525.
Bean, D.W., Wang, T., Bartelt, R.J. and Zilkowski, B.W.
(2007b) Diapause in the leaf beetle Diorhabda elongata
(Coleoptera: Chrysomelidae), a biological control agent
for tamarisk (Tamarix spp.). Environmental Entomology
36, 531540.
Carruthers, R.I., DeLoach, C.J., Herr, J.C., Anderson, G.L.
and Knutson, A.E. (2008) Salt cedar areawide pest management in the western USA. In: Koul, O., Cuperus, G.
and Elliot, N. (eds) Areawide Pest Management: Theory
and Implementation. CABI, Wallingford, pp. 271299.
Coss, A.A., Bartelt, R.J., Zilkowski, B.W., Bean, D.W. and
Petroski, R.J. (2005) The aggregation pheromone of Diorhabda elongata, a biological control agent of saltcedar
(Tamarix spp.): identification of two behaviorally active
components. Journal of Chemical Ecololgy 31, 657670.
DeLoach, C.J. and Tracy, J.L. (1997) Biological Assessment:
Effects of biological control of saltcedar (Tamarix ramosissima) on endangered species. US Fish and Wildlife Service for Consultation under Section 7, Endangered Species
Act, 17 October 1997 (submitted to USDI Fish and Wildlife Service as C.J. DeLoach and Juli Gould, 1997).
DeLoach, C.J., Carruthers, R.I., Lovich, J.E., Dudley, T.L.
and Smith, S.D. (2000) Ecological interactions in the biological control of saltcedar (Tamarix spp.) in the United
States: toward a new understanding. In: Spencer, N.R.
(ed) Proceedings of the X International Symposium on
Biological Control Weeds. Montana State University, MT,
USA, pp. 819873.
DeLoach, C.J., Carruthers, R.I., Dudley, T.L., Eberts, D.,
Kazmer, D.J., Knutson, A.E., Bean, D.W., Knight, J., Lewis, P.A., Milbrath, L.R., Tracy, J.L., Tomic-Carruthers, N.,
Herr, J.H., Abbott, G., Prestwich, S., Harruff, G., Everitt,
J.H., Thompson, D.C., Mityaev, I., Jashenko, R., Li, B.,
Sobhian, R., Kirk, A., Robbins, T.O. and Delfosse, E.S.
(2004) First results for control of saltcedar (Tamarix spp.)
in the open field in the western United States. In: Cullen,
J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin,
L. and Scott, J.K. (eds) Proceedings of the XI International
Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 505513.
Dudley, T.L. and Kazmer, D. J. (2005) Field assessment of the
risk posed by Diorhabda elongata, a biocontrol agent for
control of saltcedar (Tamarix spp.), to a nontarget plant,
Frankenia salina. Biological Control 35, 265275.
Everitt, J.H., Yang, C., Fletcher, R.S., DeLoach, C.J. and Davis,
M.R. 2007. Using remote sensing to asses biological control of saltcedar. Southwestern Entomologist 32, 93103.
Gaskin, J.F. and Schaal, B.A. (2003) Molecular phylogenetic
investigation of U.S. invasive Tamarix. Systematic Botany
28, 8695.
Gaskin, J.F. and Schaal, B.A. (2002) Hybrid Tamarix widespread in U.S. invasion and undetected in native Asian
range. Proceedings of the National Academy of Sciences
USA 99, 1125611259.
Hudgeons, J.L., Knutson, A.E., Heinz, K.M., DeLoach, C.J.,
Dudley, T.L., Pattison, R.R. and Kiniry, J.R. (2007) Defoliation by introduced Diorhabda elongata leaf beetles
(Coleoptera: Chrysomelidae) reduces carbohydrate reserves and regrowth of Tamarix (Tamaricaceae). Biological Control 43, 213221.
Lewis, P.A., DeLoach, C.J., Herr, J.C., Dudley, T.L. and
Carruthers, R.I. (2003a) Assessment of risk to native Frankenia shrubs from an Asian leaf beetle, Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), introduced
for biological control of saltcedars (Tamarix spp.) in the
western United States. Biological Control 27, 148166.
Lewis, P.A., DeLoach, C.J., Knutson, A.E., Tracy, J.L. and
Robbins, T.O. (2003b) Biology of Diorhabda elongata
deserticola (Coleoptera: Chrysomelidae), an Asian leafbeetle for biological control of saltcedars (Tamarix spp.)
in the United States. Biological Control 27, 101116.
Milbrath, L.R. and DeLoach, C.J. (2006a) Host specificity of
different populations of the leaf beetle Diorhabda elongata
(Coleoptera: Chrysomelidae), a biological control agent of
saltcedar (Tamarix spp.). Biological Control 36, 3248.
Milbrath, L.R. and DeLoach, C.J. (2006b) Acceptability and
suitability of athel, Tamarix aphylla, to the leaf beetle,
Diorhabda elongata (Coleoptera: Chrysomelidae), a biological control agent of saltcedar (Tamarix spp.). Environmental Entomology 35, 13791389.
Milbrath, L.R., DeLoach, C.J. and Tracy, J.L. (2007) Overwintering survival, phenology, voltinism, and reproduction among different populations of the leaf beetle
Diorhabda elongata (Coleoptera: Chrysomelidea). Environmental Entomology 36, 13561364.
Moore, D. and Ahlers, D. (2005) 2004 Southwestern Willow
Flycatcher Study Results Selected Sites Along the Rio
Grande from Velarde to Elephant Butte Reservoir, New
Mexico. US Department of the Interior, Bureau of Reclamation, Denver, CO.
Sferra, S.J., Corman, T.E., Paradzick, C.E., Rourke J.W.,
Spencer, J.A. and Sumner, M.W. (1997) Arizona Partners in Flight Southwestern Willow Flycatcher Survey:
19931996 Summary Report. Nongame and Endangered
Wildlife Program, Arizona Game and Fish Department,
Technical Report 113, Phoenix, AZ, USA, 104 pp.
USDI-FWS (1995) Endangered and threatened wildlife and
plants; final rule determining endangered status for the southwestern willow flycatcher. Federal Register 60(38), 10964.
539
Monitoring the rust fungus, Puccinia jaceae
var. solstitialis, for biological control of
yellow starthistle (Centaurea solstitialis)
A.J. Fisher,1 D.M. Woods,2 L. Smith1 and W.L. Bruckart3
Summary
Yellow starthistle, Centaurea solstitialis L., is a noxious weed that infests more than 7 million ha of
rangeland in California. The rust fungus, Puccinia jaceae Otth var. solstitialis, was first released as a
classical biological control for yellow starthistle in California in 2003. In 2005, a research program
was initiated to monitor the life cycle and spread of P. jaceae solstitialis. The rust was released at two
sites representing different climatic zones, the coastal hills and Central Valley, in January 2005 and
2006. Releases resulted in infected plants at both sites in both years. Natural urediniospore (infective
spore) reinfection occurred throughout yellow starthistles growing season at the Central Valley site,
but infection did not persist at the coastal hills site. P. jaceae solstitialis spread at least 100 m in 2005
at the Central Valley site but did not spread at the coastal hills site. The results of this study show
that the spread of the rust is most concentrated in areas closest to release sites. Teliospores (dormant
spores) were produced during plant senescence at both sites in 2005. Our results suggest that P. jaceae
solstitialis is likely to establish and spread to new yellow starthistle populations in the Central Valley
but not in the coastal hills, near Napa California.
Keywords: reinfection, establishment, spread.
Introduction
Yellow starthistle (Centaurea solstitialis L., Asteraceae) is an invasive alien weed that infests more than
7 million ha of rangeland in California (Pitcairn et al.,
2006). Yellow starthistle displaces desirable plants in
both natural and agricultural areas. Its spiny flowers
deter feeding by grazing animals and lower the value
of recreational lands (Sheley et al., 1999). Yellow
starthistle is a winter annual adapted to the mild wet
winters and dry summers of the Mediterranean (Maddox, 1981). Seeds usually germinate soon after the beginning of fall rains, rosettes develop slowly during the
winter and plants bolt in late spring and flower continuously until the plant senesces from lack of mois-
USDA, ARS, Exotic and Invasive Weeds Research Unit, Albany, CA
94710, USA.
2
California Department of Food and Agriculture, Sacramento, CA
95832, USA.
3
USDA, ARS, Foreign Disease-Weed Science Research Unit, Fort Detrick, MD 21702, USA.
Corresponding author: A.J. Fisher < [email protected] >.
CAB International 2008
1
ture. Scientists have introduced six insect biological
control agents, all of which attack flower heads and
destroy developing seeds (Turner et al., 1995; Pitcairn
et al., 2004). However, these have not reduced yellow
starthistle populations to acceptable levels (Balciunas
and Villegas, 1999; Woods et al., 2004).
In 2003, the rust fungus, Puccinia jaceae Otth var.
solstitialis, was introduced to California for the biological control of yellow starthistle (Woods et al., 2003).
This is the first exotic plant pathogen to be approved for
release for the classical biological control of a weed in
the continental USA using the modern permitting process required by the Animal and Plant Health Inspection
Service (APHIS; Bruckart et al., 1999).
P. jaceae var. solstitialis is macrocyclic (its life cycle includes five spore stages), and it conducts its entire
life cycle on yellow starthistle (Savile, 1970). Plants
infected during the growing season produce pustules
that release urediniospores. Urediniospores disperse
aerially to infect other leaves and plants, resulting in
multiple generations within a growing season that may
progressively increase the incidence, intensity and spatial spread of infection. The latent period (time from
infection to spore formation) ranges from 10 to 15 days
540
Monitoring the rust fungus, Puccinia jaceae var. solstitialis, for biological control of yellow starthistle
at 25C to 15C, respectively (Bennet et al., 1991).
Many rust fungi produce teliospores, dormant spores,
during plant senescence. Teliospores germinate in the
presence of a host plant to produce basidiospores during
the wet season. To complete the life cycle, basidiospores are thought to infect yellow starthistle seedlings to produce pycnia. Pycnia have been observed
on yellow starthistle seedlings in California at a site
where P. jaceae solstitialis was released the previous
year (Fisher et al., 2006). Pycnia produce aecia, which
produce urediniospores to continue the infection cycle
(Savile, 1970).
In 2005, a research program was initiated to monitor
the life cycle and spread of P. jaceae solstitialis in California. The goals of this study were to: (1) monitor urediniospore natural reinfection over the yellow starthistle
growing season, (2) monitor teliospore emergence and
(3) monitor the spread of P. jaceae solstitialis at two
release sites in California.
Materials and methods
Fungal isolate and site characteristics
P. jaceae var. solstitialis field isolate FDWSRU 8471 was collected by S. S. Rosenthal in 1984 east of
Yarhisar and Hafik (near ivas), Turkey. This isolate
was used for host-specificity testing and was the isolate
released from quarantine (Bruckart, 1989; Bruckart
et al., 1999). Urediniospores for the present study were
propagated at the California Department of Food and
Agriculture, Sacramento, CA, using the methods described by Woods and Popescu (2004).
Permanent experimental plots were established in
January 2005 and 2006 at sites near Napa and Woodland, CA. The Napa site is an ungrazed rangeland in the
coastal hills near Napa, Napa County, 427-m elevation,
dominated by yellow starthistle and exotic European
annual grasses, and surrounded by oak and manzanita forest. The Woodland site is an ungrazed rangeland, near Woodland, Yolo County, 5-m elevation. It
is dominated by exotic European annual grasses, surrounded by actively grazed rangeland and agriculture.
The Woodland site is in the Central Valley, which has
a hot summer, and the Napa site is in the coastal hills,
which is cooler.
Urediniospore natural reinfection over
a single growing season
In 2005, six permanent 1 0.5 m plots marked by
wooden stakes were installed at each site. In late January 2005, each plot was uniformly sprayed with 200 ml
of deionized water containing 50 mg urediniospores
and 0.15% Tween 20 (=100 mg spores/m2) using plastic 250-ml finger pump spray bottles. This was sufficient to wet all plants in the plot to runoff. A portable
rectangular wall of plastic sheeting was placed around
each plot to prevent drift during application. Plots were
evaluated 30, 60, 90 and 120 days after inoculations
and scored either as positive or negative for disease
symptoms.
In 2006, the experiment was repeated on a smaller
scale, and six additional 0.5 0.5 m plots were installed. Plots were inoculated on the same dates, using
the same methods, as in 2005.
Teliospore emergence
In addition to the 12 plots above, a separate 1-m2 plot
was inoculated at both sites, with 200 ml of deionized
water containing 50-mg spores and 0.15% Tween 20, to
monitor teliospore emergence in 2005. Plots were reinoculated once a month from February to June, if needed, to maintain infection. Once a month, from March
to September, five infected leaves were harvested from
each site. In the lab, spores were scraped off leaves into
0.1% water agar, and the number of urediniospores and
teliospores in a 200-spore sample from each leaf was
counted using a compound microscope. The Proc GLM
procedure was used to evaluate the effect of location
(Napa or Woodland) on teliospore production. Analyses were carried out using SAS Institute software version 9.1 (SAS Institute, 2000).
Monitoring P. jaceae solstitialis spread
In a separate study to develop an optimal strategy
to release P. jaceae solstitialis (Fisher et al., 2007),
9 g of urediniospores were used to inoculate yellow
starthistle at the two sites described above. Plots of
yellow starthistle were inoculated up to five times
from January to June. At the first inoculation, plants
were at the rosette stage, and during the last inoculation, plants were beginning to flower. To document
spread, permanent 0.5 2 m plots were installed 20,
40, 60, 80 and 100 m from inoculated plots at both
sites in May 2005, positioned along four compass
directions. Because of yellow starthistle patchiness
and site boundaries, there were a total of seven monitoring plots 20 m from release plots, five monitoring
plots 40 and 60 m from release plots, four monitoring
plots 80 m from release plots and six monitoring plots
100 m from release plots. In May 2005, we visually
inspected plants within a metre on all four sides of
each of the six plots inoculated in January to observe
natural urediniospore spread. In June 2005, we visually inspected all monitoring plots 20, 40, 60, 80 and
100 m from inoculated plots and recorded the number
of infected plants out of a maximum of 50 plants per
plot. To determine if the rust fungus had spread beyond 100 m in July 2005, roadside yellow starthistle
populations were inspected for P. jaceae solstitialis
infection at the Central Valley site up to a distance of
approximately 5 km.
541
XII International Symposium on Biological Control of Weeds
Results and discussion
P. jaceae var. solstitialis pustules emerged on yellow
starthistle at both sites in 2005 and 2006 within 30 days
after inoculations (Fig. 1). In the coastal hills (Napa
site) in 2005, rust pustules were present in all plots
after 60 days but declined to three of six plots after 90
days and then to two of six plots after 120 days. In 2006
at the Napa site, there was no natural reinfection at
plots inoculated in 2005, and plots inoculated in 2006
had pustules only at 30 days after inoculation. All six
plots in the Central Valley (Woodland site) contained
infected plants at each sampling date in both years. It is
not clear why P. jaceae solstitialis did not consistently
persist in release plots at the Napa site. To identify factors that limit the success of the pathogen, it will be
necessary to monitor local climate conditions. Bennett
et al. (1991) estimated that optimal P. jaceae solstitialis
infection occurs at 1520C with dew periods between
8 and 16 h. Data on dew periods at these locations
Figure 1.
would allow us to determine if and when these criteria
are met in the field.
As the season progressed, temperatures rose and yellow starthistle senesced; P. jaceae solstitialis produced
more teliospores (dormant spores that germinate the
next winter) as a proportion of all spores. Production of
teliospores gradually increased at both sites in August
and September 2005 (Fig. 2). A higher percentage of
spores sampled in the Central Valley were teliospores
(P < 0.001) compared to the coastal hills.
The rust spread up to 100 m at the Woodland site in
2005 (Fig. 3). In May of 2005, all of the plots inoculated in January had pustules at least 1 m outside of
plots at the Woodland site. By July, the rust had spread
to one of the five monitoring plots 100 m from the nearest release plot. Of the monitoring plots, the highest
percentage with rust, aside from the metre surrounding
plots, occurred at 60 m (Fig. 3). It is not clear why a
greater percentage of plots had rust at 60 m compared
to 20 or 40 m. When we compare the percentage of
Occurance of Puccinia jaceae var. solstitialis pustules indicating recent
infection of yellow starthistle plants in permanent plots that were inoculated in late January in 2005 and 2006.
542
Monitoring the rust fungus, Puccinia jaceae var. solstitialis, for biological control of yellow starthistle
Figure 2.
Natural production of Puccinia jaceae var. solstitialis teliospores (dormant spores), as a proportion of all
spores (means SE), at the Napa and Woodland, CA, sites in 2005 after inoculations in release plots starting in January 2005.
plants in each plot with at least one pustule, 26% of the
plants in plots 20 m from release plots were infected,
compared with 2% in 40-m plots, 6% in 60-m plots, 2%
in 80-m plots and 2% in 100-m plots. These data show
that the spread of the rust is most concentrated in areas
closest to the release plots. No rust fungus was found
during visual inspection of roadside yellow starthistle
populations surrounding the release site (up to 5 km
away). Spread was modest compared to other successful biological control pathosystems, for example the
white smut fungus, Entyloma ageratinae Barreto and
Figure 3.
Evans, released for mist flower, Ageratina riparia (Regel) King and Robinson, spread 80 km over water in 2
years (Barton et al., 2007).
The year after plots were inoculated, P. jaceae solstitialis naturally reinfected yellow starthistle at the site
in the Central Valley but not in the coastal hills (Fisher
et al., 2007). For P. jaceae solstitialis to reinfect yellow starthistle after a dormant season, it must produce
teliospores that persist and germinate in the winter to
complete the life cycle. Teliospores were observed in
the summer, and pycnia were observed in release plots
Spatial spread of Puccinia jaceae var. solstitialis, as represented by pustules on yellow
starthistle plants during one growing season in Woodland, CA, in 2005.
543
XII International Symposium on Biological Control of Weeds
the following February (Fisher et al., 2006). Soon after
pycnia were observed, in March, urediniospores were
identified in the 2005 release plots.
In general, the rust fungus was more productive in
the Central Valley (Woodland site) than the coastal hills
(Napa site), with higher rates of reinfection, both within
a single growing season and after a dormant season.
In addition, P. jaceae solstitialis spread at least 100 m
during the first year at the Central Valley site. In the
coastal hills, yellow starthistle exhibited disease symptoms soon after inoculation, but infection decreased
over time and did not return the following year. These
results suggest that this isolate of P. jaceae solstitialis should establish well in Californias Central Valley
but not at sites with climate conditions similar to the
coastal hills.
Acknowledgements
Thanks to V. Popescu and M. Pitcairn at the California
Department of Food and Agriculture and M. Plemons
at USDA, ARS. This study was financially supported
by the USDA Research Associate Program and the
University of California Integrated Pest Management,
Exotic Pests and Disease Research Program.
References
Balciunas, J. and Villegas, B. (1999) Two new seed head flies
attack yellow starthistle. California Agriculture 53, 811.
Barton, J., Fowler, S.V., Gianotti, A.F., Winks, C.J., de Beurs,
M., Arnold, G.C. and Forrester, G. (2007) Successful biological control of mist flower (Ageratina riparia) in New
Zealand: agent establishment, impact and benefits to the
native flora. Biological Control 40, 370385.
Bennett, A.R., Bruckart, W.L. and Shishkoff, N. (1991) Effects of dew, plant age, leaf position on the susceptibility
of yellow starthistle to Puccinia jaceae. Plant Disease 75,
499501.
Bruckart, W.L. (1989) Host range determination of Puccinia
jaceae from yellow starthistle. Plant Disease 73, 155160.
Bruckart, W.L., Woods, D.M. and Pitcairn, M.J. (1999)
Proposed field release of a rust fungus, Puccinia jaceae
Otth var. solstitialis (Pucciniaceae, Uredinales, Basidiomycotina) from Europe for biological control of yellow
starthistle, Centaurea solstitialis L. (Asteraceae). Petition
to Technical Advisory Group (TAG). TAG Petition 00-07.
Foreign Disease Weed Science Research Unit, Fort Detrick, MD, USA, 42 pp.
Fisher, A.J., Bruckart, W.L., McMahon, M.B., Luster, D.G.
and Smith, L. (2006) First report of Puccinia jaceae var.
solstitialis pycnia on yellow starthistle in the United
States. Plant Disease 90, 1362.
Fisher, A.J., Woods, D.M., Smith, L. and Bruckart, W.L.
(2007) Developing an optimal release strategy for the rust
fungus Puccinia jaceae var. solstitialis for biological control of Centaurea solstitialis (yellow starthistle). Biological Control 42, 161171.
Maddox, D. M. (1981) Introduction, phenology, and density
of yellow starthistle in coastal, intercoastal, and central
valley situations in California. US Department of Agriculture, Agricultural Research Results ARR-W-20. Agricultural Research Service, Oakland, CA, USA, 33 pp.
Pitcairn, M.J., Piper, G.L. and Coombs, E.M. (2004) Yellow
starthistle. In: Coombs, E.M., Clark, J.K., Piper, G.L. and
Cofrancesco, A.F., Jr. (eds) Biological Control of Invasive Plants in the United States. Oregon State University
Press, Corvallis, USA, pp. 421435.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, D.
(2006) Yellow starthistle continues to spread in California. California Agriculture 60, 8390.
SAS Institute (2000) SAS Statistics Users Guide. SAS Institute, Cary, NC, USA.
Savile, D.B.O. (1970) Some Eurasian Puccinia species attacking Cardueae. Canadian Journal of Botany 48, 1553
1566.
Sheley, R.L., Larson, L.L. and Jacobs, J.J. (1999). Yellow
starthistle. In: Sheley, R.L. and Petroff, J.K. (eds) Biology and Management of Noxious Rangeland Weeds.
Oregon State University Press, Corvallis, OR, USA, pp.
408416.
Turner, C.E., Johnson, J.B. and McCaffrey, J.P. (1995) Yellow starthistle. In: Nechols, J.R., Andres, L.A., Beardsley,
J.W., Goeden, R.D. and Jackson, C.G. (eds) Biological
Control in the Western United States: Accomplishments
and Benefits of Regional Research Project W-84, 1964
1989. University of California DANR, Publ. No. 3361,
Oakland, CA, USA, pp. 270275.
Woods, D.M., Bruckart, W.L., Popescu, V. and Pitcairn, M.J.
(2003) First field release of Puccinia jaceae var. solstitialis, a natural enemy of yellow starthistle. In: Woods,
D.M. (ed) Biological Control Program Annual Summary
2003. California Department of Food and Agriculture,
Plant Health and Pest Prevention Service, Sacramento,
CA, USA, p. 31.
Woods, D.M. and Popescu, V. (2004) Large scale production
of the rust fungus, Puccinia jaceae var. solstitialis, for
biological control of yellow starthistle, Centaurea solstitialis., In: Woods, D.M. (ed) Biological Control Program
Annual Summary 2004. California Department of Food
and Agriculture, Plant Health and Pest Prevention Service, Sacramento, CA, USA, pp. 2122.
Woods, D.M., Joley, D.B., Pitcarin, M.J. and Popescu, V.
(2004) Impact of biological control insects on yellow
starthistle at one site in Yolo County. In: Woods, D.M.
(ed) Biological Control Program Annual Summary 2003.
California Department of Food and Agriculture, Plant
Health and Pest Prevention Service, Sacramento, CA,
USA, pp. 3537.
544
Is ragwort flea beetle (Longitarsus
jacobeae) performance reduced by
high rainfall on the West Coast,
South Island, New Zealand?
A.H. Gourlay, S.V. Fowler and G. Rattray
Summary
The New Zealand biological control programme against ragwort, Senecio jacobeae L., began in the
1920s, and four biological control insects have been released. The ragwort flea beetle, Longitarsus
jacobaeae (Waterhouse) (Coleoptera, Chrysomelidae), was introduced from Oregon into the field in
1983. The beetle has established and has had a major impact on ragwort plant populations in many areas of New Zealand, but not on the West Coast of the South Island. Field studies looked at sites where
the beetle had established to determine why it had not reduced ragwort populations. The abundance
of ragwort plants (13/m2) and of flea beetles (three per rosette) was recorded at five sites. The biology
of flea beetles and ragwort plants at each site were also recorded and rainfall data were collected.
The inability of the ragwort flea beetle to reduce ragwort populations in some areas in New Zealand
may be due to the high density and rapid growth of ragwort plants, possibly due to higher rainfall,
and this may also have a negative effect on flea beetle populations. Numbers of flea beetles per plant
was lower on the West Coast (four per rosette) than East Coast sites (ten per rosette). The density of
ragwort plants was higher on the West Coast (13/m2) than the East Coast (4/m2).
Keywords: ragwort flea beatle, Longitarsus jacobaeae, rainfall.
Introduction
The biological control programme against ragwort, Senecio jacobaeae L. (Asteraceae), began in the 1920s
and was one of the first such programmes implemented
in New Zealand. Four biological control insects have
been released: two seed-feeding flies, Botanophila
seneciella (Meade) and Botanophila jacobaeae (Hardy),
both from England, one foliage-feeding moth, Tyria
jacobaeae (L.), from England and a root-feeding flea
beetle, Longitarsus jacobaeae (Waterhouse), from Oregon. Only one of the seed-feeding flies, B. jacobaeae,
has established and only in the central North Island.
The foliage-feeding moth has established sporadically
throughout both islands, reducing plant seeding and
height, but it has had little impact on plant populations.
However, the flea beetle has established in many areas
Landcare Research, PO Box 40, Lincoln 7640, New Zealand.
Corresponding author: A.H. Gourlay <gourlayh@landcareresearch.
co.nz>.
CAB International 2008
of New Zealand and has had a major impact on ragwort
plant populations. Reductions of 90100% in plant
density have been recorded at many sites in 2 to 10
years after initial release and establishment of the flea
beetle (Hayes, 1996). There are two main exceptions
to the successful establishment of the flea beetle and
the reduction in ragwort populations: on the West Coast
and southern regions of the South Island and in western
parts of the North Island (Hayes, 1998).
Ragwort is a biennial or perennial herb. Early in
winter, plants develop into a leafy rosette 25 cm high
and up to 15 cm in diameter. Daisy-like bright yellow
flowers bloom from November to January. Stems are
reddish purple and can grow up to 60 cm. Each plant
can produce up to 250,000 seeds a year; these can lie
dormant in the soil for up to 16 years. Ragwort seeds
are dispersed by wind and water, on vehicles, machinery, clothing and in hay and chaff. The plant reproduces
through seed and vegetatively from cut root fragments
(Meander Valley Weed Strategy, 1999).
Field experiments were conducted at 11 sites in
New Zealand, where, in paired plots, ragwort plants
545
XII International Symposium on Biological Control of Weeds
were sprayed with Halmark to remove flea beetles
and compared with ragwort on unsprayed plots. The
density of ragwort plants and flea beetles in each
plot was measured over 2 years at each site. In some
cases, these plots have been monitored for a further
2 years, although spraying was discontinued in 2003.
The data showed that the flea beetle significantly reduced ragwort plant density within 1 or 2 years. Once
established, the flea beetle continued to keep ragwort
plant densities at low levels (P. McGregor, 2001, unpublished results).
This study was initiated because it was not clear why
this success is not reflected at release sites in western
areas of New Zealand. This paper describes the study
results.
Methods
Five sites (Table 1) were chosen, where the flea beetle was present and because they represented diverse
geographical and climatic areas throughout the West
Coast region.
At each site, 20 juvenile (rosette), 20 mature (bolting) and 20 multi-crown ragwort plants were selected,
and adult flea beetles were collected from them using a
leaf sucker. Beetle numbers were recorded per plant. At
Table 1.
20 randomly selected locations at each site, we counted
the total number of juvenile (rosette), mature (bolting)
and multi-crown ragwort plants per 0.25 m2 quadrat.
At each site, we measured and recorded the rosette
diameters of 50 juvenile plants, the basal diameters of
50 mature plants and the total basal diameters of 50
multi-crown plants to determine the total dry biomass
of ragwort per size class of plant per quadrat. Where
multi-crown plants occurred, 30 plants, including
roots, were collected to calculate the total dry biomass
of multi-crown plants.
Monthly rainfall data were recorded for each site,
and a maximum daily rainfall event was also retrieved
from NIWAs National Climate Database (via the CliFlo Web Access Service). Data for daily temperatures
were not available for these sites.
Twenty beetles from each site were placed in vials
of 70% alcohol and later dissected in the laboratory to
determine the sex ratio and age by looking at the state
of the wing muscles: Newly emerged adults have wing
muscles, but older beetles lose their wing muscles after
summer aestivation and mating.
Landowners were asked if any weed control or other
relevant treatments had taken place on the pasture.
During sampling, we noted the weather and recorded
an estimate of soil moisture content. Other features
aximum ragwort density and maximum beetle numbers per rosette at other New Zealand sites (P. McGregor,
M
2001, unpublished data).
Site
Max. ragwort density
Max. beetle numbers
Orere Farms (Auckland)
2.3
50
Southland
1.8
25
H/I BoP
1.3
15
Wellington
1.6
11.2
Manawatu-Wanganui
0.4
10.6
Sisam, BoP
3.5
McCann BoP
1.1
Alcove Properties
2.4
Nett. BoP
1.4
Taranaki
1.3
West Coast
13
546
Comments
Dramatic drop in plant numbers in watersprayed plots compared with plots treated
with Hallmark insecticide
Fluctuating numbers of ragwort plants but no
effects of Hallmark insecticide despite large
beetle numbers
Large drop in plant numbers before Hallmark
insecticide used
Large drop in plant numbers before Hallmark
insecticide used
Ragwort numbers very low at start, but
significant increase in plant density when
Hallmark insecticide applied
Large drop in plant numbers before Hallmark
insecticide used
No significant changes in plant numbers,
over time or in presence/absence of Hallmark
insecticide
Complex patterns perhaps due to Kikuyu
grass competition
No significant changes in plant numbers, over
time or in the presence/absence of Hallmark
insecticide
No significant changes in plant numbers,
over time or in presence/absence of Hallmark
insecticide
No significant changes in plant numbers,
over time
Is ragwort flea beetle (Longitarsus jacobeae) performance reduced by high rainfall on the West Coast
of the pasture were also recorded, e.g. pasture length,
how heavily grazed the pasture was, snow fall, extreme
weather events and whether or not it had been sprayed.
To determine the range where impacts on individual
plants might be expected, we compared the West Coast
with sites elsewhere in New Zealand. Information was
extracted from impact experiments carried out by regional council staff in collaboration with Landcare Research (P. McGregor, 2001, unpublished data) on the
density of ragwort plants and the mean number of flea
beetles recorded per rosette (Table 1). With the exception of the Southland site, all sites above the dotted
line in Table 1 (which had maximum beetle numbers
of eight or more per rosette) showed changes in plant
density consistent with successful suppression by the
flea beetle. Conversely, all the sites with maximum
beetle numbers per rosette of four or less showed no
indication of any effective suppression of plant density
by ragwort flea beetle.
Results and discussion
Some variations in ragwort densities could have been
caused by weather. For example, the monthly low in
the density of juvenile plants in July 2004 at Landsborough Valley (Fig. 1) might have been caused by recent heavy snow. Some frost damage to ragwort plants
was also noted in July 2004 at Cook River Flat (Fig.
1). Otherwise, there was no apparent effect of abiotic
factors on the abundance of ragwort plants or our ability
a) Cook River Flat
8
7
6
Pasture
long
5
4
Moderately
grazed
Frost damage
to plants
2
1
0
17-Aug-03
6-Oct-03
25-Nov-03
14-Jan-04
4-Mar-04
23-Apr-04
12-Jun-04
1-Aug-04
20-Sep-04
4-Mar-04
23-Apr-04
12-Jun-04
1-Aug-04
20-Sep-04
4-M ar-04
23-Apr-04
12-Jun-04
1-Aug-04
20-Sep-04
b) Howard Valley
8
7
6
5
Cinnabar
moth attack
Pasture
long
3
2
1
0
17-Aug-03
6-Oct-03
25-Nov-03
14-Jan-04
c) Inchbonnie
Pasture
very long
8
7
Herbicide
applied
6
5
4
3
2
1
0
17-Aug-03
6-Oct -03
25-Nov-03
14-Jan-04
547
XII International Symposium on Biological Control of Weeds
d) Landsborough
8
7
6
Heavily grazed
Previous
heavy snow
4
3
2
1
0
17-Aug-03
6-Oct-03
25-Nov-03
14-Jan-04
4-Mar-04
23-Apr-04
12-Jun-04
1-Aug-04
20-Sep-04
e) Tauranga Bay
8
Very heavily grazed with
hoof damage
7
6
5
4
3
2
1
0
17-Aug-03
Figure 1.
6-Oct -03
25-Nov-03
14-Jan-04
4-M ar-04
12-Jun-04
1-Aug-04
20-Sep-04
Number of ragwort plants per 0.25-m2 quadrat during the study period. Plants were recorded
as juvenile (rosette stage) shown as solid lines or as adults with flowering stem(s) shown as
dotted lines. Each point is a mean of 20 quadrats per date per site; bars are 1 SEM.
to sample them (except when sampling was prevented
by snow in July 2004 at Howard Valley). None of the
rainfall measurements (daily peak per month, monthly
total and total for year) were correlated with ragwort
density. Monthly rainfall data are summarized in Table 2.
Flea beetles were found at three sites only. They were
Landsborough Valley, Tauranga Bay and Howard Valley.
Table 2.
23-Apr-04
Plant density and flea beetle density were not related
at any of the sites sampled (Fig. 2). Table 1 suggests
that the impacts of the flea beetle on ragwort plant
density are only likely when beetle numbers per rosette
exceed four (P. McGregor, 2001, unpublished data).
The highest numbers of beetles recorded were from
multi-crowned plants (mean peak of four or more per
Monthly rainfall data (mm) for the five study sites.
Site/month
September 2003
October 2003
November 2003
December 2003
January 2004
February 2004
March 2004
April 2004
May 2004
June 2004
July 2004
August 2004
Yearly total
Tauranga Bay
Landsborough
Pleasant Flat
Cook River Flat
Porika Hills Howard
Valley
Red View Farm
199
181
144
164
265
234
86
105
230
323
133
201
2265
630
200
410
318
511
493
382
76
460
555
158
430
4623
588
422
484
476
728
783
491
122
606
618
277
446
6041
22
124
127
108
118
260
38
28
158
(no record)
86
178
1247
(no record)
310
412
374
585
746
484
136
400
677
190
563
4877
548
Is ragwort flea beetle (Longitarsus jacobeae) performance reduced by high rainfall on the West Coast
a) Howard Valley
6.00
5.00
4.00
3.00
2.00
1.00
0.00
17-Aug-03
6 -Oct-03
2 5-Nov-03
14-Jan-04
4-M ar-04
23-Ap r-0 4
12-Jun-04
1-Aug-04
20-Sep-0 4
b) Landsborough
6.00
5.00
4.00
3.00
2.00
1 .00
0.00
1 7-Aug-03
6-Oct-03
25 -Nov-03
14-Jan-0 4
4-Mar-04
23-Ap r-04
12-Jun-04
1 -Aug-04
2 0-Sep-04
c) Tauranga Bay
6.0 0
5.0 0
4.0 0
3.0 0
2.0 0
1.0 0
0.0 0
17-Aug-0 3
Figure 2.
6 -Oct-03
2 5-Nov-03
14-Jan-0 4
4 -M ar-04
23-Apr-04
12-Jun-04
1-Aug-04
20-Sep-04
Mean number (SEM) of beetles per plant at the three sites where they were found.
Samples were from 20 plants except on a few occasions when fewer than 20 could be
located. Three plant types were sampled: juveniles (rosettes), solid lines; adults (one
flowering stem), dashed lines; multi-stemmed adults dotted lines (these only occurred at
Tauranga Bay). In 20032004, no ragwort plants matured at the Howard Valley site.
plant, Tauranga Bay, January 2004). The highest number of beetles recorded from a single rosette was ten
beetles at Tauranga Bay on 25 Sept. 2003, which was
still relatively low compared with the numbers in Table
1 (P. McGregor, 2001, unpublished data).
In addition, numbers of beetles on single-stemmed
adult plants peaked on this date, coinciding with a
marked decrease in the number of beetles collected from
juvenile plants. This suggests that multi-crowned plants
attracted beetles from the smaller single-stemmed and
rosette plants. However, previous studies showed that,
although adult beetles can be abundant on large, multicrowned plants, the number of larvae inside the roots is
low per gram dry weight compared with rosette plants
(James et al., 1992).
In general, beetle numbers at the three sites were
lowest (less than one beetle per plant) in winter/spring
and increased over summer/autumn to four per plant
(Fig. 2). The numbers of beetles collected were not related to monthly rainfall, peak daily rainfall (per month)
or weather conditions. However, there was a trend suggesting that beetle densities were lower at sites with
higher annual rainfall (Fig. 3).
Dissections showed that new adults were present
from September to March, peaking in December (when
no sexually mature adults were collected at any site).
549
XII International Symposium on Biological Control of Weeds
Mean beetles cm -2 (log n+1)
0.0025
0.002
0.0015
0.001
0.0005
0
1500
2500
3500
4500
5500
6500
Total rainfall
Figure 3.
Mean number of ragwort flea beetles collected per unit area (cm2) of juvenile plants
plotted against the total recorded rainfall for the study year at each of the five sites. The
relationship approaches statistical significance (P = 0.07, r2 = 0.61, y = (3.7 10)
7x + 0.0022, F(1,3) = 7.20).
four at one site, Tauranga Bay, and this threshold was
exceeded in only seven of the 12 monthly sampling
dates, although only in 15 rosettes out of the total of
240 sampled during the study year. When rainfall for
each site, where beetles were found, was overlaid onto
the graphs of beetle and plant density, these events appeared not related (Fig. 3). Sites where beetles were not
detected were not used in the analysis. Thus, although
the high rainfall on the West Coast may affect beetle
density, this remains unproven.
Ragwort, typically a biennial plant, will behave as a
perennial if the flowering stalk is cut, grazed, mowed,
trampled or mechanically injured while flowering
1
0.8
0.6
0.4
Aug-04
Jul-04
Jun-04
May-04
Apr-04
Mar-04
Feb-04
Jan-04
Dec-03
Oct-03
Nov-03
0.2
Sep-03
Proportion immature/mature
The pattern of wing muscle appearance in adult beetles
on the West Coast shows that the flea beetle has a single
generation per year, that the females live for about a
year and that they are capable of laying eggs throughout the rest of the winter and after spring. The ratio
of males to females in each sample was approximately
1:1 at the Howard Valley site and did not differ significantly over the year. However, at Tauranga Bay and
Pleasant Flat Valley twice as many males as females
were collected during the year (Fig. 4).
The mean beetle numbers per rosette in our West
Coast study never exceeded three per rosette (Table 1).
The numbers per individual rosette plant only exceeded
Date
Figure 4.
Proportion of dissected adult female beetles that were mature (dark areas) and immature (lighter areas) during the study period.
550
Is ragwort flea beetle (Longitarsus jacobeae) performance reduced by high rainfall on the West Coast
(McEvoy, 1984). After such damage, the plant can
regenerate from crown buds, root fragments or intact
roots. When flowers are removed before they set seed,
the plant can re-flower later the same season. Defoliated rosettes will continue to grow for several years as
vegetative perennials. The heavy-stocking regime at
the Tauranga Bay site appears to have allowed perennial plants to survive and set much seed. It may also
have created gaps in the pasture and allowed substantial seedling recruitment throughout the year (Fig. 1).
The number of beetles per plant was lower at the
West Coast sites (a maximum of ten but an average of
one beetle per plant) than some East Coast sites where
ragwort has been controlled by the flea beetle (an average of 40 beetles per plant at one site but an average
range of eight to 50 beetles per plant; P. McGregor,
2001, unpublished data).
References
Hayes. L. (ed.) (1996) The Biological Control of Weeds Book.
A New Zealand Guide. Landcare Research, Lincoln, New
Zealand.
Hayes, L. (1998) Weed Clippings. Landcare Research, Lincoln, New Zealand, 16 pp.
James, R.R., McEvoy, P.B. and Cox, C.S. (1992) Combining the Cinnabar moth (Tyria jacobaeae) and the ragwort
flea beetle (Longitarsus jacobaeae) for control of ragwort
(Senecio jacobaea): an experimental analysis. Journal of
Applied Ecology 29, 589596.
McEvoy, P.B. (1984) Dormancy and dispersal in dimorphic
achenes of tansy ragwort, Senecio jacobeae L. (Compositae). Oecologica 61, 160168.
Meander Valley Weed Strategy, Tasmania (1999) Available
at: http://www.hotkey.net.au/~d.elliott (accessed 26th November 2007).
551
Host-range investigations of potential
biological control agents of alien invasive
hawkweeds (Hieracium spp.) in the
USA and Canada: an overview
G. Grosskopf,1 L.M. Wilson2 and J.L. Littlefield3
Summary
Several European Hieracium species, e.g. Hieracium caespitosum Dumort. and Hieracium aurantiacum L., are noxious weeds in North America. A project for the biological control of alien invasive
hawkweeds has therefore been initiated in 2000. Five European insect species investigated before
their release in New Zealand and two additional gall wasps have been tested on North American test
plants. The stolon-tip galling cynipid, Aulacidea subterminalis Niblett (Hym., Cynipidae) proved to
be the most specific candidate attacking four Hieracium spp. in the subgenus Pilosella. Aulacidea
hieracii (L.), a gall wasp reared from Hieracium procerum Fries (subgenus Pilosella) and Hieracium
robustum Fries (subgenus Hieracium), which severely galls the flower stalks, did not attack any of the
target weeds. Another gall wasp, Aulacidea pilosellae (Kieffer), galling the midrib of leaves, stolons
and flower stalks, attacked two native North American hawkweed species under no-choice conditions
but none of the natives exposed in open-field tests. As a negative effect on the target weeds has not
yet been shown, host-range investigations are postponed. Preliminary tests with Oxyptilus pilosellae
Zeller (Lep., Pterophoridae) were stopped due to attack of non-target species. Macrolabis pilosellae
(Binnie) (Dipt., Cecidomyiidae), a multivoltine gall midge galling the rosette centre, flower heads and
stolon tips, can develop on most native North American Hieracium spp. As attack occurred also in
field cages and in the field, this agent was removed from the list of potential agents. The root-feeding
hoverfly, Cheilosia urbana (Meigen) (Dipt., Syrphidae), and the rosette-feeding hoverfly, Cheilosia
psilophthalma (Becker), develop on seven and at least two native hawkweed species, respectively,
in no-choice larval transfer tests. However, under open-field conditions, attack rates of C. urbana
on native Hieracium spp. are much lower than on H. caespitosum. Further experiments are planned
to explore the level of C. urbana attack in the field. Neither test nor control plants were attacked by
C. psilophthalma in open-field tests in 2005 and 2006. Therefore, host-range tests with C. psilophthalma are progressing slowly.
Keywords:host specificity, non-target feeding, Cheilosia urbana, Cheilosia psilophthalma, Aulacidea pilosellae, Aulacidea hieracii, Aulacidea subterminalis,
Macrolabis pilosellae.
Introduction
Several Hieracium spp. (Asteraceae, Lactuceae) of
Eurasian origin have become troublesome weeds in
CABI Europe-Switzerland, Rue des Grillons 1, 2800 Delmont, Switzerland.
2
University of Idaho, Department of Plant, Soil and Entomological Sciences, Invasive Plant Ecology and Management, P.O. Box 442339,
Moscow, ID 83944-2339, USA.
3
Montana State University, Department of Land Resources and Environmental Sciences, Bozeman, MT 59717-3120, USA.
Corresponding author: G. Grosskopf <[email protected]>.
CAB International 2008
1
pastures, nature reserves, roadsides, and in deforested
areas in New Zealand (Syrett and Smith, 1998) and
North America (Wilson et al., 1997). The most invasive hawkweed species in North America are species
within the subgenus Pilosella, which have a high rate
of reproduction and dispersal due to high seed output
and asexual propagation, e.g. Hieracium caespitosum
Dumort. and Hieracium aurantiacum L. (Wilson and
Callihan, 1999). Traditional management efforts, e.g.
fertilizer and herbicide application, are not only costly
but are also problematic in remote areas and nature
reserves or other sensitive sites. A biological control
project was therefore initiated in New Zealand in 1993
552
Host-range investigations of potential biological control agents of alien invasive hawkweeds (Hieracium spp.)
(Syrett and Smith, 1998). Based on a literature review
and initial field surveys, five European insect species
attacking different plant parts of mouse-ear hawkweed,
Hieracium pilosella L., were chosen for further investigation for New Zealand: the plume moth, Oxyptilus
pilosellae Zeller, the gall midge, Macrolabis pilosellae (Binnie), the gall wasp, Aulacidea subterminalis
Niblett, and the hoverfly species, Cheilosia urbana
(Meigen) and Cheilosia psilophthalma (Becker) (Syrett
et al., 1999; Grosskopf, 2006). Host-specificity tests indicated that these five insect species are at least genusspecific and therefore sufficiently host-specific for
release in New Zealand (Syrett et al., 1999; Grosskopf,
2006) where all Hieracium spp. are naturalized (Webb
et al., 1988). In contrast, approximately 29 Hieracium
spp. in the subgenera Chionoracium and Hieracium are
indigenous to North America (Strother, 2006). Therefore, to predict the potential host range of these biological control agents, native North American Hieracium
spp. were tested with all of the above-mentioned insect
species. O. pilosellae was removed from the list of po
tential biological control agents at an early stage due to
difficulties in rearing the moth and its lack of specificity, and the results are therefore not presented. A second
gall wasp, Aulacidea pilosellae (Kieffer), was collected
from several target weeds in Germany, Poland and the
Czech Republic, e.g. H. aurantiacum and H. caespitosum, and thus included in the list of potential agents
(Grosskopf et al., 2004a). A third cynipid, Aulacidea
hieracii (L.), was collected from Hieracium procerum
Fries (subgenus Pilosella) and Hieracium robustum
Fries plants (subgenus Hieracium) in the Ukraine but
repeatedly failed to produce galls on any of the target
weeds (Grosskopf et al., 2004a,b). This paper provides an overview of the host specificity and the current status of five potential biological control agents of
invasive hawkweeds in North America, i.e. A. subterminalis, A. pilosellae, M. pilosellae, C. urbana and C.
psilophthalma.
Potential biological control agents
C. urbana (Diptera, Syrphidae)
Females of C. urbana oviposit into the leaf axils of
Hieracium spp., and the neonate larvae move into the
soil to feed externally on the roots, creating small holes
resulting in reduced above-ground biomass (Grosskopf,
2005). In Europe, mature larvae pupate in late September/October and overwinter within the soil, very close
to the surface.
C. psilophthalma (Diptera, Syrphidae)
Adults of this univoltine hoverfly emerge in April
and May. Females oviposit into the leaf axils of hawkweed plants and the larvae feed on the above-ground
plant parts, i.e. rosette centre, leaf axils and stolon tips
(Grosskopf, 2005). The larvae are mobile and can migrate among plants as host quality decreases. Mature
larvae pupate on the soil surface in late September and
October. C. psilophthalma and the root-feeding hoverfly, C. urbana, often co-occur and have a very similar
phenology (Grosskopf, 2005).
M. pilosellae (Diptera, Cecidomyiidae)
This multivoltine gall midge deforms stolon tips,
flower heads and rosette centres. Females oviposit in
leaf axils close to the meristematic tissue. Larval feeding prevents the unfolding of the leaves, and M. pilosellae larvae live gregariously in-between them. Mature larvae move into the soil where they spin a cocoon
in which they pupate. The gall midge has three generations at Delmont, Switzerland, and overwinters in
the larval stage. Galled plants have shorter stolons and
fewer flower heads than uninfested plants (Grosskopf,
2006). According to the literature, the host range of this
gall midge is restricted to Hieracium spp. in the subgenus Pilosella (Buhr, 1964).
A. subterminalis
(Hymenoptera, Cynipidae)
This univoltine gall wasp induces multi-chambered
galls in the tips of elongating stolons. The larvae overwinter within the galls, pupate in spring and adults
emerge in May and June. Adults exhibit thelytoky. In
host-range tests carried out for New Zealand, only two
Hieracium spp. out of nine, i.e. H. aurantiacum and H.
pilosella, were attacked (Syrett et al., 1999), indicating
a narrow host range.
A. pilosellae (Hymenoptera, Cynipidae)
A. pilosellae is a small, uni- to bivoltine gall wasp,
which induces galls on stolons, midrib of leaves and
flower stalks (Buhr, 1964). Thus far, we have not been
able to demonstrate significant impact of this insect
on plant growth in garden studies (Grosskopf et al.,
2007).
Materials and methods
Test plant list
A test plant list was compiled by L. Wilson and
J. Birdsall (2001, unpublished results) based on the
phylogenetic approach proposed by Wapshere (1974,
1989). Emphasis was placed on native North American Hieracium species in the subgenera Hieracium
and Chionoracium and target weeds in the subgenus
Pilosella. Other plant species belonging to the closely
related subtribes Crepidinae, Lactucinae, Microseridinae and Stephanomeriinae were also tested but were
not attacked.
553
XII International Symposium on Biological Control of Weeds
Host-range tests
All tests were carried out at the CABI EuropeSwitzerland Centre at Delmont, with the exception of
A. subterminalis testing, which was conducted in quarantine facilities at Montana State University, Bozeman,
MT, USA. As all insects are of Central European origin
and known to have a narrow host range, open-field and
field-cage tests could be carried out without restrictions
at Delmont.
No-choice tests
No-choice larval transfer tests were carried out with
C. urbana and C. psilophthalma. Seven neonate larvae
were transferred into the leaf axils of potted test plants.
All plants were individually covered with gauze bags
and embedded in sawdust in a garden bed. All pots
were checked for immature stages from the middle of
September onwards. In the case of C. urbana, the soil
was sieved and checked for larvae and puparia. In tests
carried out with C. psilophthalma, only the upper 5 cm
of the soil and the above-ground plant parts were
checked for immature stages.
As M. pilosellae adults are short-lived and hard to
relocate, three males and three females were released
onto potted test plants covered with gauze bags and
were left on the plants during their entire life span. A.
subterminalis was tested in sequential no-choice tests.
Three females were transferred onto a caged test plant
for 3 days, and transferred onto a different test plant
afterwards. In contrast, due to their small size, A. pilosellae adults were not retrieved from the plants. Two
females and two males of A. pilosellae were placed
onto potted test plants covered with gauze bags. The
number of galls was recorded on each test plant, at the
earliest 4 to 6 weeks after exposure to the gall midge
or gall wasps.
Multiple-choice tests
Potted Hieracium plants, i.e. target weeds and test
plants, were exposed to naturally occurring Cheilosia
females in garden beds at Delmont in 2005 and 2006.
In 2005, control plants (H. caespitosum) and test plants
were exposed simultaneously. All plants were checked
at the end of the summer for mature hoverfly larvae
of C. urbana and C. psilophthalma. However, as numerous test plants died before evaluation of the tests
in 2005, a two-phase open-field oviposition test was
carried out in 2006. In the first phase, test and control
plants (H. caespitosum and H. aurantiacum) were exposed simultaneously, while in the second phase, only
native North American hawkweed species were exposed. The number of eggs on the different plants was
recorded. As C. urbana and C. psilophthalma eggs cannot be distinguished, eggs were kept in separate Petri
dishes for hatch, and freshly hatched larvae were transferred onto H. aurantiacum plants to determine the syr-
phid species at the pupal stage (Grosskopf et al., 2007).
Larvae of C. urbana and C. psilophthalma have high
survival rates on H. aurantiacum. The gall midge M.
pilosellae was tested in three field cages measuring 2
2 1.6 m. In two cages, H. caespitosum and test plants
were exposed simultaneously to the gall midges and,
in the third one, test plants only. M. pilosellae adults
emerged from rearing pots placed into the field cages.
A. pilosellae adults were released onto open-field plots
previously used for host-range testing of Cheilosia (see
above). Plants were checked for galls 6 to 8 weeks after
release of the adults.
Results and discussion
As none of the insects tested developed on any plant
species outside the genus Hieracium, we only focus
on the presentation of the hawkweeds tested. The testing programme was impeded by the difficulty in rearing Hieracium spp. in the subgenus Chionoracium
resulting in numerous invalid replicates. Therefore,
not all Hieracium spp. could be tested in sufficient
replicates.
Of the five insect species presented in this paper, the
cynipid, A. subterminalis, is by far the most specific
potential biological control agent attacking four Hieracium spp. in the subgenus Pilosella, i.e. H. pilosella,
Hieracium flagellare Willd., H. aurantiacum and Hieracium floribundum Wimmer & Grab. (Table 1). Its
narrow host range is due to the fact that A. subterminalis females oviposit into the stolon tips, which are exclusively produced by Hieracium spp. in the subgenus
Pilosella, whereas native North American hawkweed
species never produce stolons for vegetative propagation. Insects that exclusively or mainly attack stolons
are therefore given priority and should result in a lower
risk of non-target effects. Although H. caespitosum
plants from Idaho also produce stolons, they were not
galled by A. subterminalis. This gall wasp appears to
be sufficiently host specific for introduction into North
America.
Hieracium scouleri and Hieracium bolanderi,
both native North American hawkweed species, were
utilized as hosts in no-choice tests carried out with
A. pilosellae and adults emerged from H. scouleri galls.
However, the gall wasp has not galled any of the indigenous Hieracium spp. exposed in open-field tests
(Table 2) conducted to date. Although A. pilosellae
seems to have a narrow field host range, screening tests
are postponed since a negative impact on the target
weeds has not yet been shown (Grosskopf et al., 2007).
Further impact experiments with this cynipid are being
carried out in 2007.
Although gall-inducing insects are generally
known to have a restricted host range (Shorthouse and
Watson, 1976), the gall midge M. pilosellae does not
show a sufficient level of specificity for field release
in North America. M. pilosellae galls 12 indigenous
554
Host-range investigations of potential biological control agents of alien invasive hawkweeds (Hieracium spp.)
Table 1.
No-choice tests with five potential biological control candidates.
Insect species
Cheilosia urbana
Cheilosia
psilophthalma
Macrolabis
pilosellae
% larvae retrieved
Subgenus Pilosella
Hieracium
aurantiacum L.
Hieracium caespitosum Dumort.
Hieracium
flagellare Willd.
Hieracium floribundum Wimm. et Grab.
Hieracium
glomeratum Froel.
Hieracium
pilosella L.
Hieracium piloselloides Vill.
Subgenus Hieracium
Hieracium
canadense Michx.a
Hieracium
umbellatum L.a
Subgenus Chionoracium
Hieracium
albiflorum Hook.a
Hieracium
argutum Nutt.a
Hieracium
bolanderi Graya
Hieracium
carneum Greenea
Hieracium
fendleri Schultz-Bip.a
Hieracium
gracile Hook.a
Hieracium greenei
Porter et Britt.a
Hieracium
gronovii L.a
Hieracium
horridum Friesa
Hieracium longiberbe
T. J. Howella
Hieracium
longipilum Torr.a
Hieracium
parryi Zahna
Hieracium
scabrum Michx.a
Hieracium
scouleri Hook.a
Hieracium
venosum L.a
Aulacidea
subterminalis
Aulacidea
pilosellae
% plants galled
60.7 (n = 8)
52.4 (n = 15)
50.0 (n = 6)
25.6 (n = 73)
24.9 (n = 59)
83.9 (n = 62)
38.6 (n = 44)
0 (n = 3)
55.6 (n = 9)
100 (n = 21)
50.0 (n = 16)
25.4 (n = 9)
10.0 (n = 10)
60.0 (n = 15)
52.9 (n = 17)
58.7 (n = 9)
52.9 (n = 10)
44.4 (n = 9)
27.4 (n = 12)
15.9 (n = 9)
83.3 (n = 12)
54.3 (n = 46)
44.4 (n = 9)
3.2 (n = 9)
53.3 (n = 15)
0 (n = 11)
0 (n = 7)
100 (n = 4)
83.3 (n = 6)
16.7 (n = 12)
0 (n = 12)
14.3 (n = 21)
0 (n = 15)
0 (n = 14)
4.8 (n = 12)
0 (n = 9)
20.0 (n = 30)
0 (n = 13)
0 (n = 16)
0 (n = 3)
9.5 (n = 3)
5.9 (n = 17)
0 (n = 18)
0 (n = 5)
0 (n = 3)
0 (n = 1)
0 (n = 1)
0 (n = 18)
66.7 (n = 6)
0 (n = 15)
33.3 (n = 3)
0 (n = 15)
47.1 (n = 17)
0 (n = 15)
0 (n = 9)
0 (n = 1)
14.3 (n = 7)
0 (n = 10)
0 (n = 1)
25.0 (n = 12)
0 (n = 7)
0 (n = 7)
10.0 (n = 10)
0 (n = 14)
0 (n = 17)
0 (n = 17)
0 (n = 5)
0.8 (n = 18)
0 (n = 12)
3.9 (n = 11)
0 (n = 5)
0 (n = 4)
3.6 (n = 4)
0 (n = 3)
0 (n = 10)
0 (n = 3)
0 (n = 2)
16.7 (n = 6)
6.1 (n = 7)
0 (n = 5)
0 (n = 6)
0 (n = 2)
27.3 (n = 11)
0 (n = 11)
6.1 (n = 7)
0 (n = 9)
37.5 (n=16)
0 (n = 15)
0 (n = 2)
0 (n = 8)
30.8 (n = 13)
0 (n = 11)
80.0 (n = 5)
0 (n = 14)
0 (n = 3)
0 (n = 8)
3.6 (n = 4)
0 (n = 3)
Values in brackets indicate the number of replicates
a
Hawkweed species indigenous to North America.
555
0 (n = 15)
0 (n = 5)
0 (n = 2)
XII International Symposium on Biological Control of Weeds
Table 2.
Multiple-choice host-range tests
Cheilosia urbana
Macrolabis pilosellae
Hieracium
Hieracium
caespitosum present caespitosum
absent
Subgenus Pilosella
Hieracium aurantiacum
Hieracium caespitosum
Subgenus Hieracium
Hieracium canadensea
Hieracium umbellatuma
Subgenus Chionoracium
Hieracium albifloruma
Hieracium argutuma
Hieracium bolanderia
Hieracium carneuma
Hieracium fendleria
Hieracium gracilea
Hieracium gronoviia
Hieracium longipiluma
Hieracium scabruma
Hieracium scouleria
Hieracium venosuma
Hieracium
caespitosum
present
+
+
Hieracium
caespitosum
absent
+
+
+
+
Aulacidea
pilosellae
+
+
+
+
C. urbana and A. pilosellae were tested in the field, whereas the gall midge M. pilosellae was tested in 2 2 1.6 m field cages. H. caespitosum
present: test plants and H. caespitosum plants were exposed simultaneously, H. caespitosum absent: test plants were exposed in the absence
of H. caespitosum.
a
Hawkweed species indigenous to North America; +, attack; , no attack
North American Hieracium species, including Hieracium carneum and Hieracium gronovii (Table 1),
which were also attacked in field cage tests in the
presence of the target weed H. caespitosum. Adult
midges were reared from eight native North American hawkweed species. In addition, in open-field tests
carried out in 2004, the indigenous North American
hawkweed species H. carneum and H. scouleri were
attacked in the presence of the target weed H. caespitosum (Grosskopf et al., 2004b). However, alien invasive hawkweeds are of increasing concern in North
America (Wilson and Callihan, 1999). Recently, the
European species, Hieracium glomeratum Froel., was
recorded to also become invasive in North America
(Wilson et al., 2006). If M. pilosellae proves to be a
successful biological control agent of North American
target Hieracium species in New Zealand, a reconsideration of the riskbenefit assessment for this agent
might be worthwhile.
In no-choice tests, C. urbana and C. psilophthalma
develop on seven and at least two native hawkweed
species, respectively. However, under open-field conditions, attack rates of C. urbana are much lower on
the native North American Hieracium spp. than on the
target weeds, i.e. in open-field tests, 5.6 larvae were
recorded on average on H. caespitosum in comparison
to 0.1 on H. scouleri and 0.3 on H. gronovii and H.
venosum, respectively (Grosskopf et al., 2006). Further
experiments are planned to explore the level of C. urbana attack in the field. The low number of native Hieracium spp. attacked by C. psilophthalma is probably
due to the low number of valid replicates and due to the
fact that not all species have been tested. No immature
stages of C. psilophthalma were retrieved from plants
exposed in open-field tests. Due to repeated failure of
open-field tests with C. psilophthalma, host-range investigations with this potential agent will require more
time than needed for C. urbana.
Conclusions
The selection of potential biological control agents
for use in North America against alien invasive hawkweeds has proven difficult. We contend with a complex
of target weeds and must consider potential non-target
impacts on numerous native species. The most host specific of the agents tested, A. subterminalis, only attacks
a portion of the target species, whereas less specific
agents, such as M. pilosellae, attack most or all of the
target weeds but will also infest several native species.
Therefore, for a majority of our agents, we may have
to balance possible non-target impacts with benefits
obtained by effectively controlling the invasive species. Additional field surveys in Russia, Romania and
Ukraine for stolon-attacking and gall-inducing insects
are planned in the near future.
556
Host-range investigations of potential biological control agents of alien invasive hawkweeds (Hieracium spp.)
Acknowledgements
Field and laboratory assistance of I. Vaisman, K. Senhadji Navarro, L. Harris, V. Chevillat, C. Lucas, S. Butler, M. Brockington, H. Schneider and A. de Meij are
greatly acknowledged. Funding was provided by the Invasive Hawkweed Consortium, including the Montana
Noxious Weed Trust Fund, Idaho State Department of
Agriculture, British Columbia Ministry of Forests and
Range, MSU AG. Experiment Station, Bureau of Land
Management, Washington State Noxious Weed Control
Board, Stevens County Weed Board, US Forest Service, and the Inland Empire and Selkirk Cooperative
Weed Management Areas.
References
Buhr, H. (1964) Bestimmungstabellen der Gallen (Zoo- und
Phytocecidien) an Pflanzen Mittel- und Nordeuropas.
Band 1: Pflanzengattungen A-M, Gallennummern 1-3488.
VEB Gustav Fischer Verlag, Jena, 761 pp.
Grosskopf, G. (2005) Biology and life history of Cheilosia
urbana (Meigen) and Cheilosia psilophthalma (Becker),
two sympatric hoverflies approved for the biological control of hawkweeds (Hieracium spp.) in New Zealand. Biological Control 35, 142154.
Grosskopf, G. (2006) Investigations on three species of
Diptera associated with hawkweeds in Europe and their
potential for biological control of alien invasive Hieracium spp. in New Zealand and North America. PhD
thesis, Christian-Albrechts-Universitt, Kiel, Germany,
138 pp.
Grosskopf, G., Senhadji Navarro, K., Ferguson, L., Maia, G.
and Poll, M. (2004a) Biological control of hawkweeds,
Hieracium spp. Annual Report 2003. Unpublished Report. CABI Bioscience Switzerland Centre, Delmont,
Switzerland, 39 pp.
Grosskopf, G. and Senhadji Navarro, K. (2004b) Biological
control of hawkweeds, Hieracium spp. Annual Report
2004. Unpublished Report, CABI Bioscience Switzerland
Centre, Delmont, Switzerland, 33 pp.
Grosskopf, G., Harris, L. and Grecu, M. (2006) Biological
control of hawkweeds, Hieracium spp. Annual Report
2005. Unpublished Report, CABI Bioscience Switzerland
Centre, Delmont, Switzerland, 21 pp.
Grosskopf, G., Moscaliuc, L., Schneider, H. and Vaisman, I.
(2007) Biological control of hawkweeds, Hieracium spp.
Annual Report 2006. Unpublished Report, CABI EuropeSwitzerland, Delmont, Switzerland, 40 pp.
Shorthouse, J.D. and Watson, A.K. (1976) Plant galls and
the biological control of weeds. Insect World Digest 3,
811.
Strother, J.L. (2006) Hieracium. In: Editorial Committee
(eds) Flora of North America, vol. 19. Flora of North
America North of Mexico, New York, pp. 278294.
Syrett, P. and Smith, L. (1998) The insect fauna of four
weedy Hieracium (Asteraceae) species in New Zealand.
New Zealand Journal of Zoology 25, 7383.
Syrett, P., Grosskopf, G., Meurk, C. and Smith, L. (1999)
Predicted contributions of a plume moth and a gall wasp
to biological control of hawkweeds in New Zealand. In:
Matthiesen, J.N. (ed.) Proceedings of the 7th Australasian
Conference on Grassland Invertebrate Ecology. CSIRO
Entomology, Perth, Australia, pp. 219226.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 201211.
Wapshere, A.J. (1989) A testing sequence for reducing rejection of potential biological control agents for weeds. Annals of Applied Biology 114, 515526.
Webb, C.J., Sykes, W.R. and Garnock-Jones, P.J. (1988)
Flora of New Zealand, vol. IV. Naturalized Pteridophytes,
Dicotyledons. Botany Division, DSIR, Christchurch, New
Zealand, 1365 pp.
Wilson, L.M., McCaffrey, J.P., Quimby, P.C., Jr. and Birdsall, J.L. (1997) Hawkweeds in the Northwestern United
States. Rangelands 19, 1823.
Wilson, L.M. and Callihan, R.H. (1999) Meadow and orange
hawkweed. In: Sheley, R.L. and Petroff, J.K. (eds) Biology and Management of Noxius Rangeland Weeds. Oregon
State University Press, Corvallis, USA, pp. 238248.
Wilson, L.M., J. Fehrer, J., Brutigam, S. and Grosskopf,
G. (2006) A new invasive hawkweed, Hieracium glomeratum (Lactuceae, Asteraceae), in the Pacific Northwest. Canadian Journal of Botany 84, 133142.
557
Azolla filiculoides Lamarck (Pteridophyta:
Azollaceae) control in South Africa:
a 10-year review
M.P. Hill,1 A.J. McConnachie2 and M.J. Byrne3
Summary
Azolla filiculoides Lamarck (red water fern) is a floating aquatic fern that was introduced to South Africa in 1948 and, by 1990, had infested a large number of water bodies and impacted water utilization
and aquatic biodiversity. The frond-feeding weevil, Stenoplemus rufinasus Gyllenhal, was released
against this weed in 1997. Pre-release studies showed this agent to be host-specific, damaging, capable of a high rate of population increase and had a wide thermal tolerance. The weevil was released
at 112 sites throughout South Africa and rapidly dispersed to all sites of A. filiculoides. Quantitative
post-release evaluations revealed that the weevil caused a dramatic reduction in the populations of
the weed, with local extinctions occurring at the majority of the sites within the space of a year. In
the last 10 years, the weed has reoccurred at a number of sites. These re-infestations did not reach the
levels recorded before 1997 and were brought under control by the weevil. The weevil has shown the
predicted wide thermal tolerance in the field and an ability to disperse unaided, up to 300 km. Despite
local extinctions of the host plant, the weevil has been able to persist by moving between infestations
of the weed. A. filiculoides no longer poses a threat to aquatic ecosystems in South Africa and is considered to be under complete control.
Keywords: Stenoplemus rufinasus, post-release evaluation.
Introduction
The frond-feeding weevil, Stenopelmus rufinasus Gyllenhal, was imported from Florida, USA, in 1995 and
was released as a biological control agent for the floating aquatic fern, Azolla filiculoides Lamarck (red water fern), in South Africa in 1997. Pre-release studies
showed that the weevil was host-specific (Hill, 1998)
and was subsequently shown to be very damaging (McConnachie et al., 2004), with a broad thermal tolerance
(McConnachie, 2004), and should therefore contribute
to the control of the weed throughout its range in South
Africa. Some 10 years after the initial releases of the
weevil, we test these pre-release predictions. The review presented in this paper is based on country-wide
Rhodes University, P.O. Box 94, Department of Zoology and Entomology, Grahamstown 6140, South Africa.
2
ARC-Plant Protection Research Institute, Private Bag X 6006, Hilton
3245, South Africa.
3
University of the Witwatersrand, Animal, Plant and Environmental Sciences, Private Bag X 3, Witwatersrand 2050, South Africa.
Corresponding author: M.P. Hill <[email protected]>.
CAB International 2008
1
field surveys by the authors during 1999, two in 2000,
2001 and 2006, ad hoc opportunistic surveys while on
field trips for other aquatic weeds or from material collected by other biological control researchers, conservation bodies or land owners. In all cases, identity of
the weed and the weevil were confirmed by one of the
authors.
Host specificity
S. rufinasus was collected on Azolla caroliniana Willdenow in Florida, USA, and imported into quarantine in South Africa in late 1995. As this weevil was
targeted for release on A. filiculoides in South Africa, it
was initially considered a new association (Hokkanen
and Pimentel, 1984) on A. filiculoides, and this was used
as an initial explanation for the weevils dramatic impact on the weed. However, a recent revision of species
of the American Azolla species based on leaf trichomes
and glochidia has synonymized A. filiculoides and A.
caroliniana (Evrard and van Hove, 2004). Accordingly, A. caroliniana now refers to A. caroliniana sensu
Willednow (=A. filiculoides) and A. caroliniana sensu
558
Azolla filiculoides Lamarck (Pteridophyta: Azollaceae) control in South Africa: a 10-year review
Mettenius (=A. cristata). Thus, the weevil is no loner
considered to be a new association on A. filiculoides.
The laboratory host range of S. rufinasus was determined through adult no-choice oviposition and larval
starvation trials on 31 plant species in 19 families (Hill,
1998). These trials showed that, while A. filiculoides
was significantly the preferred host, some development
also occurred on Azolla pinnata ssp. africana (Desv.)
R.K.M. Saunders and K. Fowler, Azolla nilotica De
Caisne Ex. Mett. and A. pinnata ssp. asiatica R.K.M.
Saunders and K. Fowler. The first two species are considered to be indigenous to southern Africa, while the
last species is introduced. Hill (1998) concluded that
performance on these species was so poor in comparison to that on the target species that they would be
unlikely to support field populations, and thus, the weevils should be cleared for release.
In the last 10 years, ad hoc surveys have been carried out on all three non-target species. Weevils have
been recovered from Azolla pinnata ssp. africana at
two sites in the Kruger National Park but in very low
numbers in comparison to those found on A. filiculoides. The resultant impact on this non-target species is
still to be quantified.
Distribution
Temperature is one of the major factors influencing insect development (Stewart et al., 1996) and has been
implicated as a major contributing factor in the success
or failure of biological control programmes. McClay
and Hughes (1995), Stewart et al. (1996) and Good
et al. (1997) have shown that failure of establishment of
several biological control agents could be directly attributed to climate incompatibility of the agent to its area of
introduction. A. filiculoides has a temperate distribution
in South Africa and was especially problematic in the
high-lying areas of the country where air temperature
ranges between 11C and 32C in summer, and -9C
and 12C in winter (Schulze, 1997). Therefore, McConnachie (2004) investigated the thermal physiology of
the S. rufinasus to predict its potential distribution in
South Africa. In the laboratory, the weevil revealed an
unusually wide thermal tolerance with its lethal limits
ranging from -12C at the lowest level and 36.5C
at the upper limit: a lower developmental threshold
of 9.18C, above which it requires only 256.4 days
to complete development (McConnachie, 2004). This
suggested that there would be very few localities in
South Africa, where the agent would fail to establish.
This prediction was upheld in that the weevil established at 91 of the original 112 (the remaining 21 sites
either washed away or were not revisited) release sites,
of which 53 sites were located in the cooler, high-lying
above 1200 m), central regions of the country characterized by frosts in winter where the winter temperatures drop below -5C on at least 20 nights of the year
(Schulze, 1997). S. rufinasus has even managed to es-
tablish in the UK (Gassman et al., 2006), suggesting
that the agent is limited by the presence of its host plant
and not by climate.
Dispersal
McConnachie (2004) undertook a series of semi-field
experiments on the dispersal ability of S. rufinasus
and predicted that the weevil would be a moderate disperser, capable of short-distance flights among patches
of A. filiculoides. Inference from field evidence, however, revealed that the weevil is capable of three different dispersal patterns: within site, short dispersal
between close sites and long-distance dispersal. It appears as though often a single female will find a mat
of the weed, oviposit and the resultant population then
move out in a concentric wave destroying the mat.
Once food quality declines, the adults disperse to other
red water fern infestations in the vicinity (up to 20 km
away) and the original mat rapidly rots and sinks. This
results in mass mortality of the immature weevil stages.
This behaviour has also been recorded from Florida in
the United States (T. Center, personal communication).
Short-distance dispersal flights of up to 20 km allow
S. rufinasus to find most of the mats in a limited area;
however, dispersal distances of up to 350 km have now
been recorded (McConnachie et al., 2004).
The weevil was originally released at 112 sites
throughout South Africa (McConnachie et al., 2004), and
it has now been recorded from an additional 42 sites. The
weed reoccurred at 22 of the original 112 release sites up
to 2 years after the initial clearing. The weevil has since
been able to relocate and clear the weed at all of these
sites. Average clearance time was within 10 months.
S. rufinasus is now widely established throughout South
Africa, and there is no need to artificially distribute the
agent. We underestimated the dispersal capabilities of
this agent in our original predictions.
Impact on the weed
The ultimate aim of biological control is to reduce the
weed population to below an economic or ecological
threshold (De Bach, 1964), but this is often not achieved
using a single agent, and additional agents or intervention from other control methods are required (Stiling
and Cornelissen, 2005). S. rufinasus was first released
as 300 adults on a 1-ha pond of red water fern in Pretoria in December 1997. In January 1998, the weevils
were still present, and by February, the red water fern
mat had sunk and more than 30,000 adult weevils were
reared from 2 m2 of rotting red water fern. This dramatic population explosion occurred at all of the subsequent release sites of the agent. We were able to
monitor 91 of the 112 original release sites around
South Africa, all of which were completely controlled,
with an average time to control being 6.9 months (McConnachie et al., 2004). Where the weed has re-infested
559
XII International Symposium on Biological Control of Weeds
sites from germination of spores, the weevil has relocated and controlled these infestations within a
year. The weevil has not failed to control a single site
where it has been released and monitored. Hill (1998)
expected that S. rufinasus had the potential to contribute to the control of the weed. Indeed, an additional
agent, the flea beetle, Pseudolampsis guttata (LeConte),
was imported from Florida, USA, into quarantine in
South Africa and screened for release but rejected due
to a lack of host specificity (Hill and Oberholzer, 2002).
We clearly underestimated the impact of this weevil,
and this is a good example of where a thorough preand post-release evaluation should be undertaken on an
agent before additional agents are considered.
A full cost/benefit analysis was performed on the
A. filiculoides biological control programme (McConnachie et al., 2003), which showed that the economic
returns would increase from 2.5:1 in 2000 to 13:1 in
2005, and are predicted to increase to 15:1 by 2010.
Based on the field evidence, there is no reason to suggest that this prediction is an underestimate.
Conclusion
A. filiculoides is considered to be under complete biological control in South Africa in that it no longer poses
a threat to the biodiversity and utilization of waterways
in that country. The factors that have contributed to the
success of this programme include an agent that has
an exceptionally high rate of increase (Hill, 1998),
is a voracious feeder on the weed both as larvae and
adults (McConnachie et al., 2004), is a good disperser,
is thermally tolerant to the South African climate
(McConnachie, 2004) and does not appear to be subject
to significant predation or any parasitism.
Acknowledgements
The Water Research Commission of South Africa supported the biological control programme on red water
fern in South Africa.
References
De Bach, P. (1964) The scope of biological control. In: De
Bach, P. (ed) Biological Control of Insect Pests and
Weeds. Chapman and Hall, London, pp. 320.
Evrard, C. and van Hove, C. (2004) Taxonomy of American
Azolla species (Azollaceae): a critical review. Systematics
and Geography of Plants 74, 301318.
Gassmann, A., Cock, M.J., Shaw, R. and Evans, H. (2006)
The potential for biological control of invasive alien
aquatic weeds in Europe: a review. Hydrobiologia 570,
217222.
Good, W.R., Story, J.M. and Callan, N.W. (1997) Winter cold
hardiness and supercooling of Metzneria paucipunctella
(Lepidoptera: Gelechiidae), a moth introduced for biological control of spotted knapweed. Environmental Entomology 26, 11311135.
Hill, M.P. (1998) Life history and laboratory host range of
Stenopelmus rufinasus, a natural enemy for Azolla filiculoides in South Africa. BioControl 43, 215224.
Hill, M.P. and Oberholzer, I.G. (2002) Laboratory host
range testing of the flea beetle, Pseudolampsis guttata
(LeConte) (Coleoptera: Chrysomelidae), a potential natural enemy for red water fern, Azolla filiculoides Lamarck
(Pteridophyta: Azollaceae) in South Africa. The Coleopterists Bulletin 56, 7983.
Hokkanen, H.M.T. and Pimentel, D. (1984) A new approach
for selecting biological control agents. Canadian Entomologist 121, 829840.
McClay, A.S. and Hughes, R.B. (1995) Effects of temperature on development rate, distribution, and establishment
of Calophasia lunula (Lepidoptera: Notuidae), a biocontrol agent for toadflax (Linaria spp.). Biological Control
5, 368377.
McConnachie, A.J. (2004) Post release evaluation of Stenopelmus rufinasus Gyllenhal (Coleoptera: Curculionidae)
a natural enemy released against the red water fern,
Azolla filiculoides Lamarck (Pteridophyta: Azollaceae) in
South Africa. Unpublished PhD Thesis. University of the
Witwatersrand, South Africa, 212 pp.
McConnachie, A.J., Hill, M.P. and Byrne, M.J. (2004)
Field assessment of a frond- feeding weevil, a successful biological control agent of red water fern, Azolla
filiculoides, in southern Africa. Biological Control 29,
326331.
McConnachie, A.J., de Wit, M.P., Hill, M.P. and Byrne, M.J.
(2003) Economic evaluation of the successful biological
control of Azolla filiculoides in South Africa. Biological
Control 28, 2532.
Schulze, R.E. (1997) South African Atlas of Agrohydrology
and Climatology. Water Research Commission, Pretoria,
Report TT82/96.
Stewart, C.A., Emberson, R.M. and Syrett, P. (1996) Temperature effects on the alligator weed flea-beetle, Agasicles
hygrophila (Coleoptera: Chrysomelidae): implications for
biological control in New Zealand. In: Moran, V.C. and
Hoffmann, J.H. (eds) Proceedings of the IX International
Symposium on Biological Control of Weeds. University of
Cape Town, South Africa, pp. 393398.
Stiling, P. and Cornelissen, T. (2005) What makes a successful biocontrol agent? A meta-analysis of biological control
agent performance. Biological Control 34, 236246.
560
Species pairs for the biological control of
weeds: advantageous or unnecessary?
C.A.R. Jackson and J.H. Myers
Summary
Two or more species of insects in the same genera have been introduced in some weed biological
control programs, either by accident or on purpose, to increase the climatic range distribution of the
agents. If the species are ecological equivalents, the introduction of both species will be unnecessary
and may be detrimental if competition reduces their overall impacts. If however the species vary in
the type of attack or their distributions, the introduction of congeners can be advantageous. In this
paper, we review the following cases of species pair releases in the biological control of weeds in
North America: the beetles Chrysolina quadrigemina (Suffrian) and Chrysolina hyperici (Forster)
for St. Johnswort (Hypericum perforatum L.); the gallflies Urophora affinis Frfld. and Urophora
quadrifasciata (Meig.) for knapweed (Centaurea) species; the weevils Neochetina bruchi Hustache
and Neochetina eichhorniae Warner for water hyacinth [Eichhornia crassipes (Mart.) Solms] and
the beetles Galerucella pusilla Duftschmidt and Galerucella calmariensis L. for purple loosestrife
(Lythrum salicaria L.). In the cases of Chrysolina and Galerucella, the species pairs appear to be
complementary for control of the target weeds, with each species providing good control in slightly
different habitats and coexisting at some sites. With the Urophora gallflies, one species is more effective than the other, but overall, seed reduction is greater when both agents are present in combination (although this seed reduction is insufficient for a reduction of the target weed). In the case of
the Neochetina beetles, a difference in the ability of the two species to kill plants was apparent, but
whether they differed in habitat preferences was less clear. Given the increasing focus on the possible
non-target effects of weed biological control introductions, we recommend that greater care be taken
to avoid mixed species introductions and that judicious use be made of controlled field experimentation to determine impacts of the species. Molecular studies of species before introduction could help
prevent the accidental introduction of multiple species.
Keywords: congeneric, risk, non-target effects, competition, climate, single agent,
multiple agents.
Introduction
The issue of single vs multiple agent releases in the
biological control of weeds has been a subject of much
debate. Some have asserted that choosing the most effective agents will reduce the potential for undesirable
non-target effects (Louda et al., 2003). Denoth et al.
(2002) reviewed cases of successful biological control
of weeds and found that a majority of successes are attributed to a single agent. They suggest that additional
agents should be introduced only if reductions in plant
University of British Columbia, Department of Zoology, 2360-6270
University Boulevard, Vancouver, BC, Canada V6T 1Z4.
Corresponding author: C.A.R. Jackson <[email protected]>.
CAB International 2008
density are not being achieved by the initially introduced agent(s). Where two or more agents are released
simultaneously, they should be released in geographically distinct areas where the differential success of the
agents can be assessed.
The success of biological control agents in reducing
target plant density depends on factors such as climate,
target weed phenology, nutrient conditions, dispersal
ability, fecundity, and type and level of damage to key
life stages of the plant. In some cases, two species in
the same genus have been introduced even though they
are very similar in their interactions with their host
plants. This has sometimes been by accident, such as
the case of the two species of Urophora and the mixture of the two species of Galerucella and sometimes
on purpose from different habitats. When the species
are difficult to distinguish morphologically, consider-
561
XII International Symposium on Biological Control of Weeds
able confusion can occur over the impacts, distributions
and coexistence of each. This is currently the situation
with Larinus minutus Gyllenhall. and Larinus obtusus
Gyllenhall. introduced on knapweeds in North America
(Harris, 2005).
The competitive exclusion principle predicts that
two very similar sympatric species cannot occupy the
same niche, and therefore, for coexistence, even very
similar species must have some differences in niche
requirements (Hardin, 1960). It is of interest therefore
to know if differences between congeneric biological
control agents contribute to their successful control of
target weeds. For example, if congeneric species act
in a cooperative and complementary fashion to reduce
plant densities, their introductions would be warranted.
The greatest advantage of congeneric biological control agents occurs if differential climatic tolerances
result in effective control of the host plant across different biomes.
In this paper, we address two questions in regard
to congeneric releases: (1) Is there evidence that both
species are necessary for successful biological control?
(2) Do congeneric species exploit different climatic regions and thus broaden the geographic effectiveness of
biological control?
Methods and materials
We restricted our analysis to agents released to control
weeds in North America, based on studies reported by
Coombs et al. (2004) and Mason and Huber (2002). Of
the 25 weed control programmes reported in Coombs
et al. (2004), 19 (76%) involved the release of more
than one species of biological control agent insect,
and 10 (40%) involved the release of congeneric species of agents.
For each case of congeneric biological control agent
releases, we scoured the literature, searching Web of
Science using agent species names as the search terms,
paying particular attention to papers where the two species had been compared. We also searched for papers
in the literature cited sections of each paper to find additional information.
We excluded from further analysis species pairs or
combinations for which clear differences existed between the species in terms of target weeds. We also
excluded species that had not yet been established or
for which insufficient post-release information was
available for appropriate comparisons. In addition,
we excluded the Aphthona flea beetle species released
for control of leafy spurge, Euphorbia esula L. Five
species, Aphthona czwalinae (Weise), Aphthona la
certosa Rosenhauer, Aphthona cyparissiae (Koch),
Aphthona flava Guillebeau and Aphthona nigriscutis
Foudras, have been released in the United States and
Canada, and a sixth species, Aphthona abdominalis
Duftschmidt, has been released in the United States but
failed to establish (Hansen et al., 2004). The five species were introduced because of differences in habitat
preferences, but control of E. esula in shrubby riparian areas remains a challenge (Bourchier et al., 2002).
Due to the large number of different species released,
their recent releases and the lack of comparative data
among all combinations, we did not consider this
group of congeneric species. We also did not consider
agents currently under consideration for release. The
four species pairs discussed in detail below represent
a range of habitat (grassland, riparian and aquatic) and
target weed types and have (1) established, (2) reached
substantial populations, (3) been studied in their nonnative range and (4) been compared in the literature.
Results
St. Johnswort, Hypericum perforatum L.:
Chrysolina hyperici (Forster) and
Chrysolina quadrigemina (Suffrian)
The defoliating beetles, C. hyperici and C. quadri
gemina, comprise two of the five species of biological
control insects established in North America for control of the rangeland weed St. Johnswort (H. perfora
tum). These two beetles represent some of the earliest
attempts at biological control, having been introduced
in 1945 and 1946, respectively, in the United States
(Piper, 2004) and in 1951 in Canada (Harris, 1962).
C. hyperici originates from northern and central
Europe and western Asia. Eggs are laid in the fall (or
in the spring in the colder continental interior; Piper,
2004). Larvae hatch and feed on leaf buds and leaves
and pupate in the soil. Adults emerge in the spring,
feed and then enter the soil for summer diapause before
emerging in the fall to mate and lay eggs.
The native range of C. quadrigemina extends further
south from Denmark to North Africa, and this species
prefers warmer, drier areas. Both Chrysolina species
are univoltine and have similar life cycles but slightly
different phenologies.
Field and laboratory studies in New Zealand show
that, for both Chrysolina species, the termination
of summer diapause is triggered by shortening day
length. However, C. quadrigemina terminates summer diapause approximately 3 to 4 weeks earlier, at a
day length of approximately 13.5 h compared to 12.5 h
for C. hyperici. C. quadrigemina females reach sexual
maturity more quickly and therefore oviposit earlier. In
areas with mild winters C. quadrigemina is considered
to be the superior agent since the larvae feed on plants
for a longer time than C. hyperici larvae (Schops et al.,
1996).
Early studies in British Columbia by Harris (1962)
suggested that C. hyperici could be the superior agent
in areas with early frosts. Since oviposition occurs on
562
Species pairs for the biological control of weeds: advantageous or unnecessary?
average a month later than that of C. quadrigemina,
many of the eggs do not hatch until the following
spring. Thus, a greater portion of the next generation
overwinters in the more resistant egg stage, thus escaping winter larval mortality.
In a subsequent assessment, Harris et al. (1969)
found C. quadrigemina to dominate in dry subhumid
sites, particularly those with Ponderosa pine (Pinus
ponderosa Douglas ex Lawson et. C. Lawson), while
C. hyperici was most effective in moist subhumid sites
where Douglas fir [Pseudotsuga menziesii (Mirbel)
Franco] was present. Both species of beetle were successful in reducing H. perforatum populations.
Peschken (1972) compared C. quadrigemina from
British Columbia (BC) with C. quadrigemina from
California (the original source of the BC introductions two decades previously) to determine if postcolonization adaptation had occurred. The BC beetles
laid a larger number of eggs per female, and this could
increase the number of beetles surviving the harsher
winters to the next generation. Furthermore, the BC
beetles demonstrated a greater tendency to seek shelter
under cold temperature conditions. Although C. hyper
ici appeared initially to be the more effective agent in
northern latitudes, its intolerance of dry conditions may
limit its overall effectiveness.
Campbell and McCaffrey (1991) similarly con
cluded that, in Northern Idaho, C. hyperici beetlesat
tack plants in more mesic forested areas, while C.
quadrigemina dominates at grassland sites where the
target weed H. perforatum most commonly occurs.
Overall, it is clear that, worldwide, C. quadrigemina
is responsible for the majority of the reduction in H.
perforatum populations; however, in areas with very
cold winters or more mesic sites at higher elevation, C.
hyperici offers good complementary control (Schops et
al. 1996; Jensen et al. 2002).
Such prior knowledge of differences in phenology
and cold hardiness could inform future agent introduction and redistribution efforts in other biological control programmes.
Diffuse knapweed, Centaurea diffusa
Lamark, spotted knapweed, Centaurea
stoebe ssp. micranthos (S.G. Gmelin ex
Gugler) Hayek and meadow knapweed,
Centaurea pratensis Thuill.: Urophora affinis
Frfld. and Urophora quadrifasciata (Meig.)
The gallfly, U. affinis, was released in Canada in
1970 and in the United States in 1973 for the control of
diffuse and spotted knapweed in the genus Centaurea
(Harris, 1980a; Story et al., 1987). U. affinis oviposits
in immature flower heads, inducing the formation of
a woody gall in the receptacle, and reduces knapweed
seed production (Harris, 1980b; Shorthouse, 1989).
Another gall fly, U. quadrifasciata, was accidentally
released in 1973. U. quadrifasciata oviposits in slightly
larger, more mature flower heads, induces the plant to
form a thin, papery gall in the ovary and also reduces
seed production (Harris, 1980b).
U. affinis is primarily univoltine, whereas U. quadri
fasciata is partially bivoltine. In Western Montana, the
peaks of first-generation emergence of the two species
occur approximately 1 week apart: 25 June for U. affi
nis and 2 July for U. quadrifasciata (Story et al., 1992).
At the time of fly emergence, the majority of knapweed flower heads are at the most suitable stage for U.
affinis. Attack by U. affinis stunts the growth of the remaining heads so that many do not reach the size acceptable to U. quadrifasciata (Berube, 1980), and thus
they must disperse to find sites with capitula suitable
for oviposition for the second generation (Harris and
Myers, 1984; Story et al., 1992; Mays and Kok, 2003).
This could account for the broader distribution of
U. quadrifasciata in Canada and the United States (Story
and Coombs, 2004a,b) despite U. affinis being less
active fliers (Roitberg, 1988). In a survey in Montana,
U. quadrifasciata occurred at almost all sites examined
as compared to U. affinis that was found at approximately half of the sites (Story et al., 1987).
Surveys of original release sites in British Columbia
have revealed that U. affinis is most often the dominant species (Harris, 1980a; Myers unpublished data),
in agreement with predictions made by Story et al.
(1992). However, because of geographic and annual
variation in flower head sizes at emergence times, it
is unlikely that U. affinis will eventually displace U.
quadrifasciata (Berube, 1980). In addition, the supercooling capacity of U. affinis is superior to that of U.
quadrifasciata (Story et al., 1993), and Nowierski et al.
(2000) conclude that U. affinis is more tolerant to cold
winter temperatures. This could also help to explain
the predominance of U. affinis, particularly at the more
northerly release sites.
The exception to U. affinis dominance is in southwest
Virginia, where surveys have shown that U. quadrifas
ciata, which was not introduced but is thought to have
dispersed from releases in Maryland, now outnumbers
U. affinis. The longer growing season in this area may
favour U. quadrifasciata, although it does not appear to
have displaced previously established U. affinis populations (Mays and Kok, 2003).
U. affinis, although smaller than U. quadrifasciata
(Roitberg, 1988), reduces seed production by 2.4 seeds/
head in comparison to 1.9 seeds/head for U. quadri
fasciata and thus could be considered to be a more effective agent (Harris, 1980a). Myers and Harris (1980)
found that overall seed reduction was slightly greater
when both agents were present in combination. As Cen
taurea is not seed-limited, however, these gallflies have
not been successful in reducing overall weed density
(Myers and Risley, 2000).
563
XII International Symposium on Biological Control of Weeds
Water hyacinth, Eichhornia crassipes
(Mart.) Solms: Neochetina eichhorniae
Warner and Neochetina bruchi Hustache
The weevil N. eichhorniae was released in the
southern USA in 1972. The adults and larvae feed on
water hyacinth (E. crassipes) leaves, often damaging
meristematic tissues, and leaving distinct scars which
limit plant growth. N. bruchi feeds in a similar manner.
Both species are multivoltine (DeLoach and Cordo,
1976; Center, 2004).
Both Neochetina species are capable of reversibly
generating and degenerating flight muscles to correspond with dispersal and reproductive phases. N. bruchi
appears to be more sensitive to plant quality and require
high nitrogen content to sustain their high fecundity
(Spencer and Ksander, 2004). On stressed plants, Center and Dray (1992) found more N. bruchi with developed flight muscles than N. eichhorniae. Differences in
preferences for oviposition sites may decrease interspecific competition and lead to complementary effects,
as N. eichhorniae prefers to oviposit on younger more
central leaves, while N. bruchi prefers to oviposit on
older, more outer leaves (DeLoach and Cordo, 1976).
N. bruchi appears to be the superior agent, mainly
because it has a faster population growth rate, the females lay more eggs and the larvae develop faster. Thus,
these beetles can kill plants faster than N. eichhorniae
(DeLoach et al., 1976). The two species occur together
throughout their native range, although N. eichhorniae
predominates at warmer latitudes in northern Argentina, Paraguay and Brazil. Also, N. eichhorniae is more
tolerant to extremely high temperatures for oviposition,
adult feeding and adult and egg survival. N. bruchi
adults, in contrast, survive better at lower temperatures
(DeLoach et al., 1976). In Florida, N. eichhorniae is
the more widespread agent (Center et al., 1992; Center
et al., 1999), and it is this agent that is most often credited with the control of Eichhorniae crassipes in North
America (Goyer and Stark, 1984; Center and Durden,
1986; Center, 1987).
It is possible that N. eichhorniae may be more suitable at warmer latitudes, while N. bruchi may be more
effective at cooler temperatures and under higher nutrient levels (Heard and Winterton, 2000).
Purple loosestrife, Lythrum salicaria L.:
Galerucella pusilla Duftschmidt and
G. calmariensis L.
The defoliating beetles, G. pusilla and Galerucel
la calmariensis were introduced in North America as
mixed releases in 1992 (Hight et al., 1995). Adults and
larvae feed on the buds and leaves of purple loosestrife
(L. salicaria) killing plants or reducing their vigour in
subsequent years. The two species of beetles are very
similar in life history strategies and appearance, and
they can only be distinguished by dissection of the
males. In a study of the beetles in their native range,
Blossey (1995) concluded that these very similar species, which make identical use of their shared food resource, are able to coexist. He proposed that coexistence
could be due to differences in the competitive abilities
of individuals, as proposed by Begon and Wall (1987):
Individual variation, rather than niche differentiation,
promotes the coexistence of competing species.
In a survey of Central New York completed in 2004,
10 years after the initial introduction of the beetles,
Grevstad (2006) found that G. pusilla was generally more abundant than G. calmariensis but that the
abundance of G. calmariensis was increasing. Grevstad concluded, however, that the two species did not
occupy identical niches and that coexistence could be
due to the greater dispersal abilities of G. calmariensis,
which allows the beetle to coexist with an almost identical competitor, G. pusilla.
McAvoy and Kok (2004) found that the phenologies
of both species are almost identical and are well synchronized with the host plant. One species completes
egg development more slowly, while the other species
has faster larval development. In contrast to Neochetina
beetles, neither beetle species exploits the lower quality food source of older leaves.
McAvoy and Kok (2004) concluded that, at higher,
colder latitudes in North America, the overall better
cold tolerance of G. calmariensis makes it a superior
competitor. The situation is complex, however, as
G. calmariensis larvae feed more at lower temperatures, but G. pusilla larvae may survive better when
food is limiting, as they require less food for development (McAvoy and Kok, 2007).
In western Oregon and Minnesota, G. pusilla is the
dominant species (Schooler, 1998; L. Skinner, personal
communication), although G. calamariensis is more
common at northern locations in Minnesota (L. Skinner, personal communication). In Michigan, where
mixtures of the two species were originally introduced,
based on morphological identification, only C. calma
riensis currently occur (Landis et al., 2003; D. Landis,
personal commnunication). In Canada, beetles initially
came from the mixed species stock in the United States.
In 2005, it was found that in, Ontario, G. calmariensis
dominated sites of mixed releases that had been made
in the mid-1990s (J. Corrigan, personal communication). This is similar to observations in Michigan and
Minnesota but conflicts with the continued dominance
of C. pusilla in New York. In western Canada, species
identifications were not confirmed but are thought to
be G. calmariensis (Lindgren et al., 2002; Denoth and
Myers, 2005).
The difficulty of distinguishing the two Galeru
cella species complicates analysis, but G. calmarien
sis seems to be the superior agent in terms of dispersal
and persistence particularly in the North. Successful
biological control apparently occurs with either species
alone or together.
564
Species pairs for the biological control of weeds: advantageous or unnecessary?
Discussion
References
Biological control is highly context-dependent. Agent
persistence and plant damage depend on target plant
nutritional status and phenology. Furthermore, insect
agents are highly sensitive to local climatic conditions of temperature, precipitation and humidity. With
increasingly variable trends in climate due to global
change, having more than one agent could serve as an
insurance policy against fluctuations in survival due to
environmental conditions.
The two Chrysolina beetle species seem to offer
complementary control of St. Johnswort on a continentwide scale. Overall, C. quadrigemina is probably
responsible for most of the control of the weed, with
C. hyperici offering complimentary control in more
mesic, higher elevation and forested sites. This pattern
is similar to the latitudinal gradient in Galerucella spp.,
with control of purple loosestrife by G. pusilla being
dominant at more southerly locations, and G. calmar
iensis dominant at more northerly locations. In the case
of the Urophora gallflies, although U. affinis is superior in terms of seed reduction, overall seed reduction is
slightly greater with the two species, but the seed reduction is insufficient for successful biological control.
Caution, however, is advised. In some cases, competition from an inferior agent can result in reaching
suboptimal levels of control (Crowe and Bourchier,
2006). Furthermore, under certain environmental conditions, it may be best to introduce only one or other
of the agents. For example, under high nutrient conditions, the release of N. bruchi could result in greater
control than the release of N. bruchi in combination
with N. eichhorniae.
Along with the risk of achieving sub-optimal control due to competition from similar species, with each
new species introduction comes the risk of non-target
effects on ecosystems, a subject which has received
much attention of late (Cory and Myers, 2000; Strong
and Pemberton, 2000; Louda et al., 2003). We conclude
that the best strategy is careful field and laboratory prerelease experimentation in the native habitat, followed
by the evaluation of replicated releases of individual
species into geographically distinct areas. Finally, we
recommend that, in addition to maintaining voucher
specimens of insect releases, molecular tools for species identification be developed so that mixed stocks of
species and strains can be identified.
Begon, M. and Wall, R. (1987) Individual variation and
competitor coexistence: a model. Functional Ecology 1,
237241.
Berube, D.E. (1980) Interspecific competition between Uro
phora affinis and U. quadrifasciata (Diptera: Tephritidae)
for ovipositional sites on diffuse knapweed (Centaurea
diffusa: Compositae). Zeitschrift fr Angewandte Ento
mologie 90, 299306.
Blossey, B. (1995) Coexistence of two leaf-beetles in the
same fundamental niche. Distribution, adult phenology
and oviposition. Oikos 74, 225234.
Bourchier, R., Erb, S., McClay, A.S. and Gassmann, A.
(2002) Euphorbia esula (L.), leafy spurge, and Euphorbia
cyparissias (L.), cypress spurge. In: Mason, P.G. and Huber, J.T. (eds) Biological Control Programmes in Canada,
19812000. CABI, Wallingford, UK, pp. 346358.
Campbell, C.L. and McCaffrey, J.P. (1991) Population trends,
seasonal phenology, and impact of Chrysolina quadrige
mina, C. hyperici (Coleoptera, Chrysomelidae), and Agri
lus hyperici (Coleoptera, Buprestidae) associated with
Hypericum perforatum in Northern Idaho. Environmental
Entomology 20, 303315.
Center, T.D. (1987) Insects, mites, and plant pathogens as
agents of waterhyacinth (Eichhornia crassipes (Mart.)
Solms) leaf and ramet mortality. Journal of Lake and Res
ervoir Management 3, 285293.
Center, T.D. (2004). Waterhyacinth Eichhornia crassipes. In:
Coombs, E. (eds) Biological Control of Invasive Plants in
the United States. Oregon State University Press, Corvalis, OR, USA, pp. 402413.
Center, T.D. and Durden, W.C. (1986) Variation in water hyacinth/weevil interactions resulting from temporal differences in weed control efforts. Journal of Aquatic Plant
Management 24, 2838.
Center, T.D. and Dray, F.A. (1992) Associations between water hyacinth weevils (Neochetina eichhorniae and Neo
chetina bruchi) and phenological stages of Eichhornia
crassipes in Southern Florida. Florida Entomologist 75,
196211.
Center, T.D., Dray, F.A., Jubinsky, G.P. and Grodowitz, M.J.
(1999) Biological control of water hyacinth under conditions of maintenance management: can herbicides and
insects be integrated? Environmental Management 23,
241256.
Coombs, E., Clark, J.K., Piper, G.L. and Cofrancesco, A.F.
(2004) Biological Control of Invasive Plants in the United
States. Oregon State University Press, Corvalis, OR, USA.
Cory, J.S. and Myers, J.H. (2000) Direct and indirect ecological effects of biological control. Trends in Ecology &
Evolution 15, 137139.
Crowe, M.L. and Bourchier, R.S. (2006) Interspecific interactions between the gall-fly Urophora affinis Frfld. (Diptera:
Tephritidae) and the weevil Larinus minutus Gyll. (Coleoptera: Curculionidae), two biological control agents
released against spotted knapweed, Centaurea stobe L.
ssp micranthus. Biocontrol Science & Technology 16,
417430.
DeLoach, C.J. and Cordo, H.A. (1976) Life cycle and biology
of Neochetina bruchi, a weevil attacking waterhyacinth in
Argentina, with notes on N. eichhorniae. Annals of the
Entomological Society of America 69, 643652.
Acknowledgements
J. Cory, A. Janmaat and M. Tseng provided helpful
comments on the manuscript. C.A.R. Jackson was
funded through a Canada Graduate Scholarship from
the National Science and Engineering Research Council (NSERC) of Canada. Research was funded by an
NSERC Discovery grant to J.H. Myers.
565
XII International Symposium on Biological Control of Weeds
Denoth, M. and Myers, J.H. (2005) Variable success of biological control of Lythrum salicaria in British Columbia.
Biological Control 32, 269279.
Denoth, M., Frid, L. and Myers, J.H. (2002) Multiple agents
in biological control: improving the odds. Biological Con
trol 24, 2030.
Goyer, R.A. and Stark, J.D. (1984) The impact of Neochet
ina eichhorniae on waterhyacinth in southern Louisiana.
Journal of Aquatic Plant Management 22, 5761.
Grevstad, F.S. (2006) Ten-year impacts of the biological control agents Galerucella pusilla and G. calmariensis (Coleoptera: Chrysomelidae) on purple loosestrife (Lythrum
salicaria) in Central New York State. Biological Control
39, 18.
Hansen, R.W., Spencer, D.E., Fornaseri, L., Quimby, P.C.,
Pemberton, R.W. and Nowierski, R.M. (2004) Leafy
spurge Euphorbia esula (complex). In: Coombs, E, Clark,
J.K., Piper, G.L. and Cofrancesco, A.F. (eds) Biological
Control of Invasive Plants in the United States. Oregon
State University, Corvallis, OR, USA, pp. 233262.
Hardin, G. (1960) The competitive exclusion principle. Sci
ence 131, 12921297.
Harris, P. (1962) Effect of temperature on fecundity and survival of Chrysolina quadrigemina (Suffr.) and C. hyperici
(Forst.) (Coleoptera: Chrysomelidae). The Canadian En
tomologist 94, 774780.
Harris, P. (1980a) Effects of Urophora affinis Frfld and
Urophora quadrifasciata (Meig) (Diptera, Tephritidae)
on Centaurea diffusa Lam and Centaurea maculosa Lam
(Compositae). Journal of Applied Entomology 90, 190201.
Harris, P. (1980b) Establishment of Urophora affinis Frfld. and U. quadrifasciata (Meig.) (Diptera: Tephritidae) in Canada for the biological control of diffuse and
spotted knapweed. Journal of Applied Entomology 89,
504514.
Harris, P. (2005) Larinus obtusus (Gyll.) Soft-achene feeding weevil. In: Harris, P. Classical Biological Control
of Weeds. Agriculture and Agri-Food Canada. Available
at: http://res2.agr.ca/lethbridge/weedbio/agents/alarobt_e.
htm (accessed 2 August, 2007).
Harris, P., Peschken, D. and Milroy, J. (1969) The status of
biological control of the weed Hypericum perforatum in
British Columbia. Canadian Entomologist 101, 115.
Harris, P. and Myers, J.H. (1984). Centaurea diffusa Lam.,
and C. maculosa Lam., diffuse and spotted knapweed
(Compositae). In: Kelleher, J.S. and Hulme, M.A. (eds) Bio
logical Control Programmes Against Insects and Weeds in
Canada 19691980. CAB, Slough, UK, pp. 127137.
Heard, T.A. and Winterton, S.L. (2000) Interactions between
nutrient status and weevil herbivory in the biological control of water hyacinth. Journal of Applied Ecology 37,
117127.
Hight, S.D., Blossey, B., Laing, J. and Declerck-Floate, R.
(1995) Establishment of insect biological control agents
from Europe against Lythrum salicaria in North America.
Environmental Entomology 24, 967977.
Jensen, K.I.N, Harris, P. and Sampson, M.G. (2002) Hy
pericum perforatum L., St Johns Wort (Clusiaceae). In:
Mason, P.G. and Huber, J.T. (eds) Biological Control
Programmes in Canada, 19812000. CABI, Wallingford.
UK, pp. 361368
Landis, D.A., Sebolt, D.C., Haas, M.J. and Klepinger, M.
(2003) Establishment and impact of Galerucella calmar
iensis L. (Coleoptera: Chrysomelidae) on Lythrum sali
caria L. and associated plant communities in Michigan.
Biological Control 28, 7891.
Lindgren, C.J., Corrigan, J. and De Clerck-Floate, R.A.
(2002) Lythrum salicaria L., purple loosestrife (Lythraceae). In: Mason, P.G. and Huber, J.H. (eds) Biological
Control Programmes in Canada, 19812000. CABI,
Wallingford, Oxon, UK, pp. 383390.
Louda, S.M., Pemberton, R.W., Johnson, M.T. and Follett,
P.A. (2003) Nontarget effects the Achilles heel of biological control? Retrospective analyses to reduce risk associated with biocontrol introductions. Annual Review of
Entomology 48, 365.
Mason, P.G. and Huber, J.T. (eds) (2002) Biological Control
Programmes in Canada 19812000. CABI Publishing,
Wallingford, Oxon, UK, 583 pp.
Mays, W.T. and Kok, L.T. (2003) Population dynamics and
dispersal of two exotic biological control agents of spotted knapweed, Urophora affinis and U. quadrifasciata
(Diptera : Tephritidae), in Southwestern Virginia from
1986 to 2000. Biological Control 27, 4352.
McAvoy, T.J. and Kok, L.T. (2004) Temperature dependent
development and survival of two sympatric species, Gale
rucella calmariensis and G. pusilla, on purple loosestrife.
BioControl 49, 467480.
McAvoy, T. and Kok, L. (2007) Fecundity and feeding of Ga
lerucella calamariensis and G. pusilla on Lythrum sali
caria. BioControl 52, 351363.
Myers, J.H. and Harris, P. (1980) Distribution of Urophora
galls in flower heads of diffuse and spotted knapweed
in British Columbia. Journal of Applied Ecology 17,
359367.
Myers, J.H. and Risley, C. (2000) Why reduced seed production is not necessarily translated into successful biological weed control. In: Spencer, N.R. (ed) Proceedings of
the X International Symposium on Biological Control of
Weeds. Montana State University, Bozeman, MT, USA,
pp. 569581.
Nowierski, R.M., Fitzgerald, B.C., McDermott, G.J. and
Story, J.M. (2000) Overwintering mortality of Urophora
affinis and U. quadrifasciata (Diptera: Tephritidae) on
spotted knapweed: effects of larval competition versus
exposure to subzero temperatures. Environmental Ento
mology 29, 403412.
Peschken, D.P. (1972) Chrysolina quadrigemina (Coleoptera:
Chrysomelidae) introduced from California to British Columbia against the weed Hypericum perforatum: comparison of behaviour, physiology and colour in association
with post-colonization adaption. Canadian Entomologist
104, 16891698.
Piper, G.L. (2004). St. Johnswort Hypericum perforatum. In:
Coombs, E, Clark, J.K. Piper, G.L. and Cofrancesco, A.F.
(eds) Biological Control of Invasive Plants in the United
States. Oregon State University Press, Corvalis, OR,
USA, pp. 322334.
Roitberg, B.D. (1988) Comparative flight dynamics of knapweed gall flies Urophora quadrifasciata and U. affinis
(Diptera: Tephritidae). Journal of the Entomological So
ciety of British Columbia 85, 5864.
Schooler, S.S. (1998) Biological control of purple loosestrife Lythrum salicaria by two Chrysomelid beetles
Galerucella pusilla and G. calmariensis. MSc thesis. Oregon State University, Corvallis, OR, USA, 128 pp.
566
Species pairs for the biological control of weeds: advantageous or unnecessary?
Schops, K., Syrett, P., and Emberson, R.M. (1996) Summer
diapause in Chrysolina hyperici and C. quadrigemina (Coleoptera: Chrysomelidae) in relation to biological control
of St Johns wort, Hypericum perforatum (Clusiaceae).
Bulletin of Entomological Research 86, 591597.
Shorthouse, J.D. (1989) Modification of flowerheads of diffuse knapweed by the gall-inducers Urophora affinis
and Urophora quadrifasciata (Diptera: Tephritidae). In:
Delfosse, E.S. (ed) Proceedings of the VII International
Symposium on the Biological Control of Weeds. Istituto
Sperimentale per la Patologia Vegetale (MAF), Rome,
Italy, pp. 221228.
Spencer, D.E. and Ksander, G.G. (2004) Do tissue carbon and
nitrogen limit population growth of weevils introduced to
control waterhyacinth at a site in the Sacramento-San Joaquin Delta, California? Journal of Aquatic Plant Manage
ment 42, 4548.
Story, J.M., Boggs, K.W. and Good, W.R. (1992) Voltinism
and phenological synchrony of Urophora affinis and U
quadrifasciata (Diptera, Tephritidae), two seed head flies
introduced against spotted knapweed in Montana. Envi
ronmental Entomology 21, 10521059.
Story, J.M. and Coombs, E. (2004a). Urophora affinis. In:
Coombs, E., Clark, J.K., Piper, G.L. and Cofrancesco,
A.F. (eds) Biological Control of Invasive Plants in the
United States. Oregon State University Press, Corvalis,
OR, USA, pp. 228229.
Story, J.M. and Coombs, E. (2004b). Urophora quadrifas
ciata. In: Coombs, E., Clark, J.K., Piper, G.L. and Cofrancesco, A.F. (eds) Biological Control of Invasive Plants in
the United States. Oregon State University Press, Corvalis, OR, USA, pp. 229230.
Story, J.M., Nowierski, R.M. and Boggs, K.W. (1987) Distribution of Urophora affinis and U. quadrifasciata, two
flies introduced for biological control of spotted knapweed (Centaurea maculosa) in Montana. Weed Science
35, 145148.
Story, J.M., Good, W.R. and Callan, N.W. (1993) Supercooling capacity of Urophora affinis and U. quadrifas
ciata (Diptera: Tephritidae), two flies released on spotted
knapweed in Montana. Environmental Entomology 22,
831836.
Strong, D.R. and Pemberton, R.W. (2000) Biological control of
invading species risk and reform. Science 288, 19691970.
567
Field studies of the biology of the moth
Bradyrrhoa gilveolella (Treitschke)
(Lepidoptera: Pyralidae) as a potential
biocontrol agent for Chondrilla juncea
J. Kashefi,1 G.P. Markin2 and J.L. Littlefield3
Summary
The root-attacking moth, Bradyrrhoa gilveolella (Lepidoptera: Pyralidae), has been released as a biological control agent for Chondrilla juncea L. (Asteraceae) in Argentina and Australia; however, both
efforts failed. As part of our effort to establish this insect in North America, we conducted a field study
of its biology in northern Greece with the goal of making an informed release according to the most
suitable environmental conditions for the larvae and to synchronize the phenologies of this insect with
its host. Our study population was at 950 m in the mountains of northern Greece, an area climatically
matching the interior mountains of the state of Idaho (USA), our intended target for colonization.
Besides obtaining a much more complete picture of its basic field biology, the most significant finding
of this study was that despite living in an area with a long, cold, snow-covered and extensive winter,
and very hot summers, this insect has a single generation a year but no distinct over-wintering stage
or highly synchronized period of emergence of adults during the summer.
Keywords: Chondrilla juncea, rush skeletonweed, Bradyrrhoa gilveolella, field biology.
Introduction
Chondrilla juncea L. (Asteraceae) is a long-lived perennial of European origin with a deep taproot. A rosette during most of the year, in early summer it bolts
to produce a tall, woody forb with slender, leafless
branches, hence the common name skeletonweed. Its
natural range in Europe extends from Volga River in
Russia to the Atlantic coast, and surrounds the Mediterranean Sea. It is a major weed in Australia, Argentina,
and the northwest United States and adjacent Canadian
provinces (Holm et al., 1997). The most recent and
rapidly spreading population in North America and the
one being targeted for control is in the southwest corner
of the state of Idaho. In Argentina and Australia, the
plant is a major problem in wheat fields, but in Idaho,
it is still primarily restricted to native ecosystems being
managed by the United States Forest Service particularly in drier pine forests and on open hillsides.
USDA, EBCL Substation, Tsimiski 43, 54623 Thessaloniki, Greece.
US Forest Service, RMRS, Bozeman, MT, USA.
3
Montana State University, Bozeman, MT, USA.
Corresponding author: J. Kashefi <[email protected]>.
CAB International 2008
1
2
Biological control was initiated by the Commonwealth Scientific and Industrial Research Organisation
(CSIRO) in Australia, which released and established
three agents: a blister-forming Cecidomyiid midge,
Cystiphora schmidti (Rbsaamen), a gall-forming
Eriophyid mite, Eriophyes chondrilla (Canestrini) and
the rust fungus Puccinia Chondrilla Bubak & Sydenham (Julien and Griffiths, 1998). The programme was
initially very successful, due to the action of the rust
(Burdon et al., 1981; Cullen, 1981). In North America,
the same three agents were introduced and are presently established throughout the range of C. juncea
(Piper et al., 2004). Unfortunately, in southern Idaho,
their combined effect is having little or no discernible
effect in reducing the spread or impact of this invasive
weed.
The Australians also released a fourth agent, the
root-attacking moth Bradyrrhoa gilveolella (Careshe
and Wapshere, 1975); despite numerous attempts,
however, it failed to establish (Cullen, 1981). With the
apparent successful control of C. juncea by the rust,
Australian authorities eventually discontinued work on
this agent. Subsequently, the Argentinians attempted
to establish B. gilveolella but also failed (Julien and
Griffiths, 1998). In Idaho in the mid 1990s, due to the
568
Field studies of the biology of the moth Bradyrrhoa gilveolella
Methods and materials
Targeted release sites in Idaho
Within the south-west Idaho infestations, Garden
Valley, 50 km north of Boise, was selected as the target
area. The introduction of C. juncea, originally carried
out in the mid 1960s, in this valley represents the oldest
and densest stand of C. juncea in southern Idaho.
Collection site in Europe
To find a B. gilveolella population in Europe living
under a similar habitat and climatic zone to Garden
Valley area, we examined populations of C. juncea
around the Black Sea and throughout the Balkans and
found B. gilveolella at many sites. We then attempted
to match habitats in Europe with our target areas in
Idaho by comparing soil temperatures. At five locations within the south-western Idaho C. juncea population, soil temperature probes (Optic Stow-Away
Temp Probes, Onset Computer Corporation, Pocasset, MA, USA) were buried 23 cm deep (where the B.
gilveolella spends its larval stage) at the base of mature C. juncea plants. Probes were similarly buried at
seven sites in Bulgaria and northern Greece. Based on
a comparison of habitats (Littlefield et al., 2008) and
soil temperatures, we selected an area at Lake Prespa in
north-western Greece for collections. This high (950 m)
mountain valley has a long, cold winter, usually with
a prolonged snowpack, and has the closest soil temperature match to Garden Valley in Idaho (Figure 1).
The large population of B. gilveolella found in the
sand beach around this lake has become the source of
the populations being tested and introduced to Idaho.
We have also concentrated on studying the insects biology in this area.
35
30
25
Temperature C
rapid spread of C. juncea and its impact on native ecosystems, a new effort at biological control was initiated
and the first agent selected was B. gilveolella, based
on its known specificity as shown in the work conducted by Australians (Careshe and Wapshere, 1975).
A petition to release B. gilveolella was submitted for
release in North America (Littlefield et al., 2000) and
approved in 2002. Attempts are presently underway
to establish this insect in Idaho. In view of its failure
to establish in both Argentina and Australian, we realized that this may be a difficult insect to colonize.
Accordingly, concurrent to our attempt to establish it
in Idaho, we have been studying both its habitat preference (Littlefield et al., 2008) and its biology under
field conditions in northern Greece. We are hoping to
identify clues to help synchronize the life stage of B.
gilveolella being released with the phenology of C.
juncea, the local climate of southern Idaho, and other
factors which might affect its establishment, such as
coarseness of soil.
20
Lake Prespa Greece
15
Garden Valley Idaho
10
5
0
-5
1 5 9 13 17 21 25 29 33 37 41 45 49
82.83% match
of overlap
# of Week
Figure 1.
Composite for mean weekly soil temperature at
23 cm depth for 2002 to 2005. The profile for
Garden Valley is laid over the profile for Lake
Prespa.
Seasonal life history of B. gilveolella
Field sampling consisted of fortnightly visits to Lake
Prespa, where the sampling area was divided into three
parallel collection strips, each about 800 m long and
20 m wide according to the fineness of the sand grains
(coarse, top of beach; medium, middle of the beach;
fine, near the water). Field researchers walked through
one randomly selected transect at each collection strip,
and the closest C. juncea plant at every 3-m interval
was excavated. There were higher densities of plants
near the lake, and medium densities in the other two
sampling strips away from the water. At each sampling
time, a total of 51 plants from all three sampling strips
were collected. In the European Biological Control
Laboratory (EBCL) substation in Thessaloniki, Greece,
the larvae were extracted from their feeding tubes and
their instar was determined. All larvae were preserved
in alcohol, and 533 were found suitable for head capsule measurements to determine their development
stage. The total number of larvae collected each month
over the 4 years was determined and the percentage of
each of the five instars calculated. Additional studies
on the biology of this insect were made in quarantine
facilities in Bozeman, MT, USA, while it was undergoing host specificity testing (Littlefield et al., 2000).
Results
Examination of larvae in the field has given us information on their stage of development at different times
of the year, their method of over-wintering and time of
adult emergence (i.e. the presence of pupae or recently
empty pupae cases). Figure 2 shows the relative abundance of the different instars of Bradyrrhoa found at
Lake Prespa over the 4 years of this study. Pupae were
found beginning in mid June, peaking in July, and ending in early August. Presumably, adults were present in
the field at about the same time.
Adult females mate immediately after emerging but
usually require 23 days of feeding on sugar water in
569
XII International Symposium on Biological Control of Weeds
% Larva Distribution by Instar
100%
Pupa
0%
100%
5th Instar
0%
100%
0%
100%
4th Instar
3rd Instar
0%
100% 2nd Instar
0%
100% 1st Instar
Ja
n
Fe uar
br y
ua
ry
M
ar
ch
Ap
ril
M
ay
Ju
ne
Ju
A ly
Se ug
p t u st
em
b
O er
ct
o
N
ov ber
e
D mb
ec e
em r
be
r
0%
Figure 2.
Yearly occurrence of the various life stages of
533 Bradyrrhoa gilveolella larvae collected in
the field at Lake Prespa. Data used is based on
the monthly cumulative number of larvae collected between 2003 and 2005.
the laboratory before they begin oviposition. The ratio
of adult males to females is approximately equal. In
the laboratory, mating takes place readily in small mesh
emergence cages or plastic storage boxes containing
only paper towels. Oviposition in the laboratory lasts
57 days with a female producing on average 100 eggs
(Littlefield et al., 2000).
Eggs
In mesh cages over potted plants in the greenhouse,
females do not have a preferred oviposition site, and
eggs were found on a wide variety of plant parts. Unfortunately, despite intensive searching we did not find
eggs or empty egg shells at Lake Prespa so we cannot
confirm the normal location for oviposition in the field.
In the greenhouse on caged plants, eggs are firmly attached to the plant part that they are laid on, are 0.50.7
mm in diameter, flattened, and initially white, but as
they develop, become reddish, then darken 35 days
before hatching as the larvae becomes viable.
Larvae
Results of 3 years of field observations showed that
620 (30.4%) of the total larvae counted were collected
from the transect with plants growing in coarse sand,
48.4% in mixed sand and 21.2% fine sand.
First Instar: The newly emerged larva descends from
the plant by crawling down the leaf or stem to the
rosette. There it works its way down the space between
the crown and the soil and chews a small cavity approximately 0.5 mm in diameter into the root 23 cm
below the soil surface. From this wound, the plant
exudes latex, which penetrates into the surrounding
soil forming a porous clump. The larva lives in cavities
in this clump but does not form a distinct feeding
chamber.
Second Instar: The larva remains in the latex clump
and continues to feed in small pits in the root and forms
a chamber 1 mm in diameter and 34 mm long adjacent
to the root surface.
Third Instar: The larva abandons the latex clump and
feeding pits, and at a slightly lower point on the root
constructs the first true feeding chamber: a small 1- to
2-mm silk-lined tunnel attached to the outside of the
root. Feeding is restricted to a feeding groove 12 mm
wide and 35 mm long, penetrating the root cortex.
The larva feeds on the edges of this groove which have
a distinct yellow colour but show no sign of the latex
that, otherwise, oozes from any wound in the cortex.
Fourth Instar: The larva feeds by enlarging the groove
excavated in the root but a major change occurs in the
feeding tube. It is extended upwards to the soil surface,
either along the side of the plant or as a branch from the
main feeding tube. The feeding tube reaches 10 cm or
more in length, and by moving within it, the larva escapes extremely low soil temperatures to survive during winter when temperatures occasionally reaches 10C. Feeding is at a single feeding site which expands
to a groove up to 10 cm and more in length and 23 mm
wide. However, the depth remains shallow (1 mm), and
shows the distinct yellow edge where the larva has cut
into the cortex of the plant.
Fifth Instar: The majority of the larval growth (at
least 50%) occurs in this instar. There is no significant
change in the size or shape of the feeding tube although
the larva continues to strengthen and thicken the wall
using pellets of latex. Feeding remains concentrated in
the original feeding groove, which reaches its maximum length of 1015 mm. Upon completion of feeding, the opening between the tube and feeding groove
is closed with silk and the interior of the feeding tube is
lined with an additional layer of silk and latex to form
the pupal chamber.
Pupa
Pupation occurs in the special silk-lined chamber in
the feeding tubes, with all exuviate and frass sealed below. Upon completion of pupation, which in our greenhouse requires 714 days, the newly emerged adult
crawls up the feeding tube or its side branch. The end
of the tube consists of only a light layer of silk covered
with soil particles, which the adult traverses to emerge
at the soil surface.
570
Field studies of the biology of the moth Bradyrrhoa gilveolella
Feeding tube
The most distinct feature in the development of this
insect is the rather unique feeding tube attached to the
C. juncea root. Initially, this is a fine, silk-lined tunnel
constructed by the third instar larva down the outside
of the root of the plant. The length of this small tunnel apparently dictates the size of the subsequent feeding tube since as the larvae grow, they do not appear
to extend the tube, but only enlarge its diameter and
strengthen the wall with latex pellets. In loose loam
or fine loam, the wall of the tube is thin, flexible and
almost cloth-like, but in coarser soil and sand, grains
are incorporated into the wall making it much thicker
and firmer. The length of this feeding tube and the larvae in the soil are quite variable, but at Lake Prespa
it averaged 7.12 cm (range 121 cm). The method of
constructing the tunnel in which the tube is formed is
unknown. It is not excavated since the tunnel initially
does not reach the soil surface, so it possibly formed by
the larva forcing the sand grains or soil particles apart.
The tunnel is subsequently lined with the latex particles
and layers of silk, and in the fourth and fifth instar, an
excess of latex particles is often produced which often
form a distinct clump of white pellets in the bottom of
the feeding tube. The latex content of this feeding tube
attracted the attention of Russian scientists who discovered that it was a high-quality natural rubber, and
studied B. gilveolella (Kozulina and Rudakona, 1932)
to determine if these feeding tubes could be harvested
as a source of locally produced rubber (Dirsch, 1933;
Iljin, 1930).
Feeding
Any slight injury to the root of C. juncea results in
a copious flow of highly viscous white latex. B. gilveolella is highly co-evolved with the plant and not only
can compensate for this latex flow but may utilize it as
food. The feeding groove the larva creates in the roots
cortex is often no larger than the body of the larva and
after the second instar there is no sign of an uncontrolled flow of the latex. Apparently, the larva, feeding
on the outer edge of this groove, ingests the latex along
with other nutrients produced by the plant and excretes
the latex as small, distinct, white oval pellets, which it
incorporates into the wall of its feeding tube. The limited size of the groove itself appears to indicate that the
larva does not obtain its nutrients by consuming significant quantities of root tissues, but must subsist on the
sap and latex flowing from this wound.
Impact
In the field, there is no visible impact on mature
plants from the feeding of several larvae except on very
small plants (roots, 12 mm in diameter) which are
severed and killed, resulting in starvation of the larva.
In the field, the insect shows no preference for plant
size, and there seems to be uniform attack on plants
with roots as small as 2 mm up to 20 mm in diameter.
Since the larvae do not consume significant amounts of
root tissue, their impact appears to be primarily through
stressing the plant by diverting the nutrients that would
normally go into growth.
Generations per year
In early studies in Russia, Bradyrrhoa was found to
have two widely overlapping generations per year (Kozulina and Rudakova, 1932). Subsequently, an Australian study on the coast of Greece also reported two
generations per year (Careshe and Wapshere, 1975). At
Lake Prespa, we were initially confused since sampling
in spring or fall, produced a very wide range of larvae
(from as small as 5 mm to more 20 mm), but without
clear evidence of two generations. Once we began to
separate the larvae by instars and began sampling during the summer months, it became clear that we had
only a single generation per year. Adult emergence begins in mid-June and continues for almost 2 months.
At this time, the larger larvae (fourth and fifth instars)
disappear, marking the end of the first generation. First
and second instar larvae are found only from July to
September. Apparently, the first larvae to hatch complete most of their development and enter diapause in
early November as fifth-instar larvae, but those larvae
that hatch from the last eggs laid enter diapause as
fourth-instar or, a few, as late third-instar larvae. Larvae end diapause the following spring and complete
their feeding and development by that summer to give
rise to the next generation.
Discussion
Soil temperature comparisons between Lake Prespa,
Greece, and the Garden Valley sites in Idaho appeared
to be the closest match between Idaho and any of the
areas studied in the Balkans. B. gilveolella over-winters
as a late fifth-instar larva or as a pre-pupa, along the
coast of Greece, where it has two distinct generations
per year (Careshe and Wapshere, 1975). At Lake Prespa,
the population has no distinct over-wintering stage and
hibernates as any of the last three instars. Since an active larva would possibly be susceptible to freezing
temperatures, this would seem to be a disadvantage.
However, its survival is probably dependent on the
snow pack that covers the ground surface for the majority of the coldest part of winter. Similarly, our site in
southern Idaho has a similar snow cover which should
protect these larval stages.
Based on these observations, we recommend that
if first instar larvae are to be released in Idaho, they
should be released in July or August. However, soil
surface temperatures during the day can be extremely
high, so late evening releases of some type of shading
may be necessary. Late-instar larvae (fourth or fifth)
571
XII International Symposium on Biological Control of Weeds
should be released in September or October. In all the
above-mentioned cases, the surface of the release area
should be covered with loose sand to a depth of about 5
cm to allow fast and easy penetration of the larvae into
the ground and to protect them from heat and dehydration. If releases of adults are planned, they should be in
cages in July.
Acknowledgements
Since Lake Prespa and the surrounding valley lay within
a Greek national park, we are very grateful to the Greek
Ministry of Agriculture, Office of Development and
Protection of Forests and Natural Habitat and CITES
office, Region of Central Macedonia, Thessaloniki, for
authorizing our study and issuing us collection and export permits for C. juncea and B. gilveolella. In quarantine at Bozeman, Montana, we are very grateful to
technician Annie Demeij and graduate student Heather
Prody for assistance in rearing and maintaining our
colony and helping us with the laboratory studies.
References
Burdon, J.J., Groves, R.H., and Cullen, J.M. (1981) The impact of biological control on the distribution and abundance of Chondrilla juncea in south-eastern Australia.
Journal of Applied Ecology 18, 957966.
Careshe, L.A. and Wapshere, A.J. (1975) Biology and host
specificity of the Chondrilla root moth Bradyrrhoa gilveolella (Treitshke) (Lepidoptera, Phycitidae). Bulletin of
Entomological Research 65, 171185.
Cullen, J.M. (1981) Considerations in rearing Bradyrrhoa
gilveolella for control of Chondrilla juncea in Australia.
In: Delfosse, E.S. (ed.) Proceedings of the V International
Symposium on Biological Control of Weeds. CSIRO Australia, pp. 233239.
Dirsch, V. (1933) Pests of rubber producing plants in the
Ukraine. Zhurnal Cycle. Bio-Zool. Acad. Sci. Ukr. 4, 41
57 (in Russian).
Holm, L., Doll, J., Holm, E., Pancho, J. and Herberger, J.
(1997) World Weeds: Natural Histories and Distribution.
Chapter 22. Chondrilla juncea L., John Wiley & Sons
Inc., New York, pp. 18393.
Iljin, M.M. (1930) Chondrilla L. Geography, ecology and
rubber content. Trudy Po Prokladnoi Botanike, Genetike I
Seleksii 24, 147169 (in Russian).
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A world catalogue of agents and their target
weeds, 4th edn. CABI Publishing, Wallingford, 223 pp.
Kozulina, O.V. and Rudakova, K.V. (1932) The biology of
the rubber-moth Bradyrrhoa gilveolella Tr. Trudy Nauchno-Issled. Inst. Prom. No. 502, 2846 (in Russian).
Littlefield, J.L., Birdsall, J., Helsley, J. and Markin, G.P.
(2000) A petition for the introduction and field release of
the Chondrilla root moth, Bradyrrhoa gilveolella (Treitschke), for the biological control of rush skeletonweed
in North America. Unpublished USDA-APHIS Biological
Control of Weeds Petition 20002002, March 2000.
Littlefield, J.L., Markin, G.P., Kashefi, J., and Prody, H.D.
(2008) Habitat analysis of the rush skeletonweed root
moth, Bradyrrhoa gilveolella (Lepidoptera: Pyralidae). In
Julien, M.H., Sforza, R., Bon, M.C., Evans, H.C., Hatcher,
P.E., Hinz, H.L. and Rector, B.G. (eds) Proceedings of
the XII International Symposium on Biological Control of
Weeds. CAB International Wallingford, UK, p. 60.
Piper, G.L., Coombs, E.M., Markin, G.P. and Joley, D.B.
(2004) Rush skeletonweed, Chondrilla juncea. In Coombs,
E.M., Clark, J.K., Piper, G.L. and Cofrancesco, A.I. Jr.
(eds) Biological Control of Invasive Plants in the United
States. Oregon State University Press, Corvallis, USA,
pp. 293303.
572
The release and establishment of the tansy
ragwort flea beetle in the northern Rocky
Mountains of Montana
J.L. Littlefield1, G.P. Markin2, K.P. Puliafico3 and A.E. deMeij4
Summary
The flea beetle, Longitarsus jacobaeae (Waterhouse) (Coleoptera: Chrysomelidae), has successfully
suppressed tansy ragwort [Senecio jacobaea L. (Asteraceae)] in mild, mid-latitude climates of western North America. Attempts to establish this flea beetle in more continental climates, such as those
found in the interior of the Pacific Northwest, have failed. With the recent incursion of tansy ragwort
in northwestern Montana, a biological control program was implemented. Two populations of L.
jacobaeae from Oregon (collected from low and high elevation sites) and cold-adapted populations
from Switzerland were released between 1997 and 2006. Several release techniques using flea beetle
eggs, larvae or adults were tried, and those using eggs or larval flea beetles were less successful than
those using adults. Subsequent surveys indicate that the Oregon low-elevation population failed to
establish and that the Oregon high-elevation and Swiss populations have established and are dispersing from their original release sites.
Keywords: tansy ragwort, ragwort flea beetle, Longitarsus jacobaeae.
Introduction
The Eurasian weed species, tansy ragwort, Senecio
jacobaea L. (Asteraceae), readily invades disturbed
rangelands, pastures, open forests and other natural areas in areas of the western USA. Montana was considered tansy ragwort-free until several infestations were
located in Lincoln and Flathead Counties after the 1994
Little Wolf Fire (USFS, 1996). By 1997, despite extensive treatments with herbicides, the plant was found
to be too widely distributed and well established to be
eradicated or economically controlled. Several biological control agents were initially introduced by the US
Forest Service from established sites in Oregon (Markin and Birdsall, 2002); including the cinnabar moth,
Montana State University, Land, Resources and Environmental Sciences, Bozeman, MT 59717 USA.
2
USDA Forest Service, Rocky Mountain Research Station, Forestry Sciences, Montana State University, Bozeman, MT 59717, USA.
3
University of Idaho, Plant, Soil and Entomological Sciences, Moscow,
ID 83844, USA.
4
Montana State University, Land, Resources and Environmental Sciences, Bozeman, MT 59717, USA.
Corresponding author: J.L. Littlefield <[email protected]>.
CAB International 2008
1
Tyria jacobaeae L. (Lepidoptera: Arctiidae), the tansy ragwort seed fly, Botanophila seneciella (Meade)
(Diptera: Anthomyiidae), and the tansy ragwort flea beetle, Longitarsus jacobaeae (Waterhouse) (Coleoptera:
Chrysomelidae). An additional population of the flea
beetle was later introduced from Switzerland. This paper reports on the release and successful establishment
of two populations of the flea beetle in Montana.
Methods
Flea beetle populations
Three populations of the tansy ragwort flea beetle
were introduced into Montana. Our initial attempt to
establish L. jacobaeae used beetles from established
populations in western Oregon. These flea beetles (referred to as the Oregon low-elevation population) were
collected from the Willamette Valley (elevation, 75 m)
or along the Oregon coast near Florence. This population was established from the Italian strain of the flea
beetle collected near Rome, Italy in 1968 by K. Frick
(USDA-ARS) and initially introduced into northern
California in 1969 and in Oregon in 1971 (Frick, 1970;
Hawkes, 1980). This population has been very effective in controlling tansy ragwort in coastal areas of
573
XII International Symposium on Biological Control of Weeds
Oregon, Washington and northern California (Hawkes
and Johnson, 1978; McEvoy et al., 1991). The seasonal
life cycle of the Italian strain has been reported by Frick
and Johnson (1973) and Windig and Vrieling (1996).
Key aspects of this phenology are (1) the adults emerge
during the summer and undergo a summer aestivation,
(2) adults re-emerge in the autumn with the onset of
seasonal rains and begin to oviposite, (3) oviposition
continues through the winter months, (4) eggs do not
have a diapause and hatch within 10 days and (5) both
eggs and early instar larvae can be observed during the
winter months. This phenology, although well suited to
milder, mid-latitude climates, is ill-adapted to colder,
drier, continental climates. Previous introductions of
L. jacobaeae in the interior portions of the Pacific Northwest, east of the Cascade Mountains, have not been
successful (Coombs et al., 1996), and we suspected
that results for Montana would be similar.
A second population of the Italian strain was also
introduced from Oregon. This population (referred to
as the Oregon high-elevation population) was collected
from Mt. Hood, Oregon (elevation, 1100 m), and differs in phenology from the Oregon low-elevation population in that adults emerge in late summer and eggs
are the primary overwintering stage (Hawkes, 1980;
Markin and Birdsall, 2002). Although it has adapted to
colder climates found on Mt. Hood, population density
of this flea beetle appear to be less than at lower elevation sites (Markin and Birdsall, 2002).
A third population of L. jacobaeae from Switzerland was investigated, as it had been reported to be better adapted to continental climates (Frick, 1971). The
life history and biological attributes that would make
this population more cold-adapted were investigated
by Frick (1971), Frick and Johnson (1972) and Puliafico (2003). The phenology of the Swiss population
of L. jacobaeae differs from that of the Italian population in that (1) adults emerge in the later part of the
summer (e.g. starting in mid-July) and do not aestivate,
(2) oviposition occurs after 2 weeks and the beetles
overwinter as eggs in a semi-diapaused state and (3)
larvae hatch the following spring and complete their
development by mid-summer. Host-specificity testing of Senecio and Packera species endemic to northwestern Montana indicated no significant non-targets
impacts associated with the Swiss populations of L.
jacobaeae (Puliafico, 2003), and therefore, the beetle
could be introduced safely into Montana. Flea beetles
were collected in 2002 to 2004 by U. Schaffner (CABIEurope) from St. Imier and Mettembert, Switzerland,
from elevations of 820 and 640 m, respectively.
Release methods and monitoring
Several different release techniques were used to
establish L. jacobaeae in Montana. Adults were used
for the releases of the Oregon low-elevation popula-
tion. Beetles were collected from sites in Oregon in
early September and were field-released within large
cages (2 4 2 m) or smaller cages (1 1 1 m). L.
jacobaeae collected from Switzerland were screened
before release for possible parasites, pathogens and
other flea beetle species at the Montana State University
Biological Control Containment Facility. Eggs obtained
from field-collected adults were used for initial releases
and for maintaining a laboratory colony for subsequent
releases. Initial releases of the Swiss population used
eggs. Eggs were harvested and placed in groups of
25 eggs on a strip of filter paper, which was placed next
to the root crown of tansy ragwort plants at field sites
and held in place by moistened peat moss or soil. Eggs
were placed in the field in early November and, after
cold treatment, in late April or early May, June and July.
In subsequent years, we also inoculated plants with
newly hatched larvae, or adults were released uncaged
or into cages (2 4 2 m). The Oregon high-elevation
populations were released as adults in cages and also as
larvae transplanted into the field in infested plants from
a laboratory colony or as larvae in infested plant material, which was placed on tansy ragwort in the field.
The success of releases was determined by the collection and dissection of plants throughout the summer to look for larvae or to determine larval feeding.
This was conduced during the year of release and the
following year. Adults were vacuumed sampled on a
yearly basis using a modified leaf blower. Sites were
sampled from late July through late October or early
November. Due to uneven plant densities, adult counts
were expressed as adults per 100 plants.
Results
Oregon low elevation population
Releases were made in 1997 and 1998, with a total
of 435 adults released at six sites in Flathead County.
Approximately 170 to 280 adults were released per site,
although numerous smaller releases of 1015 adults
were also made. Adults were recovered at low levels
for 2 or 3 years after release. Flea beetle feeding was
observed in early September, and adults were observed
from mid-September to mid-October. By 2005, no
adults were recovered at release sites. During this time,
the tansy ragwort density at sites significantly declined
due to cinnabar moth feeding (Markin and Littlefield,
2006). This decline may have impacted any flea beetle
populations remaining at these sites. To our knowledge,
no long-term establishment of this population has occurred in Montana.
Oregon high elevation population
Before the release of the Swiss population, a second
population of L. jacobaeae was located on the slopes of
574
The release and establishment of the tansy ragwort flea beetle in the northern Rocky Mountains of Montana
Mt. Hood, Oregon. Releases were made starting in 1999
and continued through 2001. An estimated 410 adults
and 2905 larvae were released at 20 sites in Flathead
and Lincoln Counties. Releases made in 1999 were inadvertently sprayed with herbicides, and no recoveries of the flea beetle have been made from these sites.
Based on observations of the flea beetle habitats on
Mt. Hood, releases in 2001 occurred in more mesic habitats along the Little Wolf drainage in Lincoln County.
By 2002, larvae were recovered at four sites where either adult or transplanted infested plants were placed.
In 2006, we determined that the Oregon high-elevation
population had increased in numbers and dispersed
several hundred metres from the initial releases in the
Little Wolf drainage. Flea beetles were most evident in
moist micro-habitats. At one site, adults were sampled
along a moisture gradient, starting at a wet seep (the
site of release), then progressing out 60 m into a drier
habitat. The number of adults was higher along the wet
seep (92 beetles per 100 plants) but decreased rapidly
(two to nine beetles per 100 plants) as one moved into
drier areas. We are not certain if the presence of adults
in the wet area was due to available water, greater plant
density or if moisture conditions are favourable for egg
laying and/or survival.
Swiss population
A total of 27,560 eggs, 16,435 larvae and 2937
adults have been released at field sites from 2002 to
2006. Several release techniques were used in our effort to establish flea beetles. Our initial releases (2002
2004) were made using eggs obtained from our rearing
colony. As large numbers of eggs could be collected,
we thought this would be an efficient way to release the
insect. This technique did not appear to be successful
due to low establishment. In 2005, we switched tactics by releasing larvae that had just hatched and eggs
that were about to hatch. Larvae or eggs were placed
on tansy ragwort rosettes or at the base of developing
stems. This technique proved more difficult, as timing
of the larval hatch was critical and a large number of
larvae had to be placed on plants within a short time
span.
With improvements in our flea beetle rearing, we
were also able to release adequate number of adults
for the first time in 2005 and 2006. Adult releases have
several advantages in that they are less time consuming, adults are less vulnerable to handling damage and
they are more likely to select oviposition sites that
maximize offspring survival.
In general, the technique of using eggs to inoculate
plants proved unsatisfactory. Although larvae were recovered from plants with autumn and spring inoculations, no recoveries were made on those inoculated in
June or July. It is speculated that eggs may have desiccated, were more susceptible to predation or the plants
were unsuitable for larval establishment at these later
dates. Of the 21 sites with egg inoculations, three have
recoverable beetle populations after 4 years.
With the success of the cinnabar moth on the Flathead County side of the tansy ragwort infestation, most
of the sites were abandoned due to very low tansy ragwort plant densities. However, there was indication that
the flea beetle had established before the rapid decline
of tansy ragwort. Two sites in Flathead County were
retained, as these appeared to have persistent pockets
of tansy ragwort, and adults of the Swiss population
were released at these sites in 2005. In Lincoln County,
where the cinnabar moth has not been as successful,
at two sites that received larvae in 2005, adults were
recovered, and of the ten adult releases made before
2006, adults were recovered at nine.
Thirty-two release locations (a location may be comprised of several individual release plots) were visited
in August and September 2006. Flea beetle adults were
observed at 23 locations (72% of the locations). Beetles
were observed at all, except for one, of the 2005 releases.
At selected sites in the Little Wolf drainage in Lincoln
County, adult beetles were collected and returned to the
laboratory to confirm their population origin. From egg
hatch data, it appeared that both the Swiss and Oregon
high-elevation populations (see below) are present in
the Little Wolf drainage, and at some sites, both populations (and/or possible hybrids) are present. Sites in
which flea beetles have been recovered ranged from
seasonally moist (e.g. intermittent streams) to dry (e.g.
burnt slash piles).
There has been a gradual increase in the number of
adult flea beetles recovered in vacuum sampling. The
mean number of adults was 2.4/100 plants in 2004
(range, 15), 2.4/100 plants in 2005 (range, 15) and
7.8/100 plants in 2006 (range, 142). The number of
beetles collected represents only a small portion of the
beetles actually present due to sampling bias (i.e. the
sampling technique). Also the number of adults collected
may vary due to date, time of day or weather conditions. We consider the Swiss population of the flea
beetle to have become established since it has persisted
for several years and increased in density, although no
dispersal from the original sites has been observed.
Phenological development
Despite low flea beetle populations, a general indication of the phenological development of the three
L. jacobaeae populations in Montana can be inferred.
The life history of the Oregon low-elevation population
appears to be similar to that reported in the literature
(Frick and Johnson, 1973; Windig and Vrieling, 1996),
with the exception of a longer larval developmental period. Adults emerge from pupation in mid-September.
Eggs and first instar larvae were observed in mid-October
and early November. During this period, soil tempera-
575
XII International Symposium on Biological Control of Weeds
tures can drop as low as -2.5C. These temperatures
did not seem to adversely affect adults or larvae present at field sites. But by emerging in September, adults
may have a reduced period of time for oviposition before the onset of colder temperatures, which may limit
the subsequent population. This phenology in Montana
is very similar to that reported for the Oregon highelevation population from Mt. Hood (Markin and Birdsall, 2002). In Montana, the life history of the Oregon
high-elevation population differs in that adults emerge
in early August, rather than in September. This would
be more advantageous to population development, as
it extends the oviposition period of L. jacobaeae. The
life history of the Swiss population in Montana is very
similar to that reported by Frick (1971) and Puliafico
(2003). Adults have been collected from early August to November. Oviposition occurs approximately
2 weeks after adult emergence. Eggs remain in a semidiapaused state until spring, and larvae complete their development by late July. The main disadvantage of this life
history is that eggs are present in the soil during the driest
time of the year and may be subjected to desiccation.
Conclusions
The tansy ragwort flea beetle appears to be well established in Montana. This is the first report of establishment of this beetle east of the Cascade Mountains
of the United States. It is speculated that the Oregon
low-elevation population failed in Montana due to low
numbers released and phenological incompatibility. The
Oregon high-elevation population is well established
but may be environmentally limited to moist habitats.
The emergence of adults of this population occurs earlier in Montana than in Oregon, giving adults time to
lay larger numbers of eggs before winter. The Swiss
population also appears to have established but may
be less restricted in its habitat requirements, thereby
making it a superior control agent. Future research
will address the environmental suitability of the two
L. jacobaeae populations in Montana and their potential impact on tansy ragwort.
Acknowledgements
We thank E. Reneau, Y. Wang, J. Wolfe, C. Horning,
A. Schmidt and A. Hunter for assisting with field and
laboratory work; A. Odor, T. Barboulatos (USFS) and
W. Chalgren for locating release sites and their help
with releases; U. Schaffner (CABI) for the collection
of beetles in Switzerland; Kootenai and Flathead National Forests and Plum Creek Timber Company for
the use of their land for study sites; Members of Tansy
Ragwort Task Force for their cooperation and financial
support from the Montana Noxious Weed Trust Fund,
Montana Agricultural Experiment Station, US Forest
Service and M.J. Murdock Charitable Fund.
References
Coombs, E.M., Radtke, H., Isaacson, D.L. and Snyder, S.P.
(1996) Economic and regional benefits from the biological control of tansy ragwort, Senecio jacobaea, in Oregon.
In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of
the IX International Symposium on Biological Control of
Weeds, January 1926 1996, University of Cape Town,
Stellenbosch, South Africa, pp. 489494.
Frick, K.E. (1970) Longitarsus jacobaeae (Coleoptera:
Chrysomelidae), a flea beetle for the biological control of
tansy ragwort. I. Host plant specificity studies. Annals of
the Entomological Society of America 63, 284296.
Frick, K.E. (1971) Longitarsus jacobaeae (Coleoptera:
Chrysomelidae), a flea beetle for the biological control of
tansy ragwort. II. Life history of a Swiss biotype. Annals
of the Entomological Society of America 64, 834840.
Frick, K.E. and Johnson, G.R. (1972) Longitarsus jacobaeae
(Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy ragwort. 3. Comparison of the biologies of the egg stage of Swiss and Italian biotypes. Annals
of the Entomological Society of America 65, 406410.
Frick, K.E. and Johnson, G.R. (1973) Longitarsus jacobaeae
(Coleoptera: Chrysomelidae), a flea beetle for the biological control of tansy ragwort. 4. Life history and adult aestivation of an Italian biotype. Annals of the Entomological
Society of America 66, 358366.
Hawkes, R.B. (1980) Biological control of tansy ragwort
in the state of Oregon, USA. In: Del Fosse, E.S. (eds)
Proceedings of the 5th International Symposium on the
Biological Control of Weeds. CSIRO, Brisbane, Australia,
pp. 623626.
Hawkes, R.B. and Johnson, G.R. (1978) Longitarsus jacobaeae aids moth in the biological control of tansy ragwort.
In: Freeman, T.E. (ed) Proceedings of the 4th International Symposium on the Biological Control of Weeds. University of Florida, Gainesville, FL, USA, pp. 193196.
Markin, G.P. and. Birdsall, J.L. (2002) Biological control of
tansy ragwort in Montana: status of work as of December
2001. Unpublished Report. USFS, Rocky Mountain Research Station, 13 pp.
Markin, G.P. and Littlefield, J.L. (2006) Biological control of
tansy ragwort in Montana: status of work as of December
2005. Unpublished Report. USFS, Rocky Mountain Research Station, 19 pp.
McEvoy, P.B., Cox, C.S. and Coombs, E.M. (1991) Successful biological control of ragwort. Ecological Applications
1, 430442.
Puliafico, K. (2003) Molecular taxonomy, bionomics and
host specificity of Longitarsus jacobaeae (Waterhouse)
(Coleoptera: Chrysomelidae): the Swiss population revisited. MSc thesis in Entomology, 119 pp.
USFS (1996) Tansy Ragwort Control Project, Tally Lake
Ranger District, Flathead National Forest, Flathead and
Lincoln Counties, State of Montana. Federal Register, 61,
6752767530 (December 23, 1996).
Windig J.J. and K. Vrieling. (1996) Biology and ecology
of Longitarsus jacobaeae and other Longitarsus species feeding on Senecio jacobaea. In: Jolivet, P.H.A.
and Cox, M.L. (eds) Chrysomelidae Biology, Vol. 3 General Studies. SPB Academic Publishing, Amsterdam,
pp. 315326.
576
Factors affecting mass production of
Duosporium yamadanum in rice grains
D.M. Macedo, R.W. Barreto and A.W.V. Pomella
Summary
Duosporium yamadanum (Matsuura) Tsuda & Ueyama is a pathogenic fungus that attacks purple
nutsedge, Cyperus rotundus, L. in Brazil. Occasionally, it is found causing severe natural epiphytotics of leaf blight on that host in the field. Exploratory studies have already indicated that this fungus
has potential for the development of a mycoherbicide. To help confirm its potential, we studied mass
production of inoculum of D. yamadanum in solid fermentation using polished rice as the substrate
for cultivation of D. yamadanum. The effects of the following factors were investigated for their
influence on conidial production: water content, length of incubation period before the opening of
the plastic bags containing the substrate and addition of supplements to the substrate. Maximum
production of conidia was obtained with a water content of 4060% in the substrate (w/v), with average production at each harvest of 2.5 105 conidia per gram of substrate. Water contents above 60%
inhibited growth and sporulation. Opening bags containing the inoculated substrate after 3 to 4 days
resulted in the highest levels of sporulation; average of 3.0 105 conidia per gram of substrate. The
supplementation of the substrate with calcium carbonate did not significantly increase the sporulation
levels compared to controls, whereas the addition of urea or sucrose led to a significant reduction in
sporulation; average of 9.0 104 conidia per gram of substrate.
Keywords: Cyperus rotundus, purple nutsedge, weed biological control, bioherbicide.
Introduction
Purple nutsedge, Cyperus rotundus L., is often considered the worlds worst agricultural weed (Holm et al.,
1977). Infestations are extremely difficult to control
through mechanical methods, and chemical control has
either been ineffective or limited by cost, environmental or management problems associated with the most
promising products. This weed has been a target by
several biological control programmes involving insect
(Frick and Quimby, 1977; Frick et al., 1979; Phatak et al.,
1987) and fungal (Phatak et al., 1982; Upadhyay et al.,
1991; Prakash et al., 1996; Neto, 1997; Okoli et al., 1997;
Ribeiro et al., 1997; Kadir et al., 1999, 2000a,b; Rosskopf et al., 1999; Kadir and Charudattan, 2000) natural
enemies. Although some experimental results have been
particularly promising, no commercial mycoherbicide
is available, and there are no classical biological control
agents in the pipeline. A combination of methods in an
Universidade Federal de Viosa, Departamento de Fitopatologia, Viosa, MG, 36571-000, Brazil
Corresponding author: D.M. Macedo <[email protected]>.
CAB International 2008
integrated management approach is often mentioned as
ideal to minimize the problems associated with intensive chemical control (Bariuan et al., 1999). Although
Brazil is known to be outside the centre of origin of
C. rotundus, a survey carried out in this country revealed
a series of fungal pathogens showing clear potential
for utilization for the development of a mycoherbicide
(Barreto and Evans, 1994, 1996). Among these, the
dematiaceous hyphomycete, Duosporium yamadanum
(Matsuura) Tsuda & Ueyama, was selected as specially
promising, as it was capable of spontaneously causing
severe leaf blight epihytotics in C. rotundus in the field
(Barreto and Evans, 1994). A series of intensive studies on the biology and management of D. yamadanum
was initiated in 1995 by Pomella (1999) and continued
by Macedo (2006) and mostly confirmed that this fungus has potential as a biological control agent. Among
the most common limitations hampering the development of mycoherbicides are those related to the mass
production of abundant, good quality inoculum to be
used as the active ingredient. Large-scale production
of fungal biological control agents for weed biological
control is mainly through techniques of liquid, diphasic
or solid fermentation (Jackson et al., 1996). Liquid fermentation is the favoured technique because it is nor-
577
XII International Symposium on Biological Control of Weeds
mally the most economically viable method (Tebeest
and Templeton, 1985). The diphasic method is usually
not viable economically because it often involves one
growth phase on a medium solidified with agar (Walker,
1980), but this is not necessarily true, as there are cheap
alternatives to agar. Finally, solid fermentation utilizes
substrates such as grains for fungal colonization and
sporulation. This technique is commonly used for
small-scale production of fungi that do not sporulate
well in liquid media (Pandey, 2003). Pomella (1999)
reported good results in mass production of D. yamadanum through solid fermentation. Additional investigations were undertaken aimed at perfecting the technique
originally developed by Pomella (1999). Investigations
on the effects of some parameters on sporulation such
as use of nutritional supplements, length of period of
incubation before initiation of harvesting and water
content in substrate were performed.
Material and methods
General conditions for experiments
Our experiments used a selected strain of D. yamadanum (RWB 476). The fungus was grown for 7 days in
Petri plates containing V8-juice agar, at 25C, under a
light regime of 12 h/day. Three culture discs (diameter,
20 mm) were cut from the margin of actively growing
cultures and transferred to flasks containing 100 ml of
sterile (autoclaved) liquid medium (200 ml V8-juice
diluted with 800 ml of water). Flasks were maintained
for 4 days under agitation at 140 rpm. The resulting
mycelium was aseptically blended within the remaining medium, and 20 ml of the mycelial suspension
was used as seed and transferred to each of a series of
polypropylene bags containing 150 g of polished rice +
75 ml of water. The bags with rice were autoclaved before seeding with mycelium of D. yamadanum. After
seeding, the bag were closed and left in a controlled
temperature room at 25C and a 12 h/day light regime
(light from two fluorescent lamps and one NUV BLB
60 W lamp placed 40 cm above the bags) for either
3 days or (in one specific experiment) for a series of five
periods of incubation of different lengths. The mass of
grains was loosened by gently pressing the bags with
the hands to allow for a uniform colonization of the
substrate by the fungus. After 3 days, the bags were
opened, and a first conidial harvest was performed. Colonized rice grains were then transferred to a flask containing 250 ml of sterile water supplemented with an
antibiotic (chloranphenicol at 2.5 ppm) and vigorously
stirred with a glass rod. The rice was sieved out of this
suspension and placed on aluminium trays (19 28 2
cm) previously cleaned with 70% alcohol and held at
25 3C for further periods of conidial production. An
interval of 24 h between each harvesting episode was
maintained. The quantity of conidia in the remaining
suspension was estimated with a haemocytometer. At
24-h intervals, the harvesting procedure was repeated,
and conidial production was evaluated. The number of
harvests performed varied for different experiments.
Effects of different levels of water
content in the substrate on sporulation
of D. yamadanum
Pomella (1999) used an arbitrary proportion of 80%
water in the substrate with seemingly adequate results.
A range of different proportions of water were tested
in this experiment, namely 40%, 50%, 60%, 70% and
80% of volume of water per weight of substrate (polished rice grains). Our experimental unit consisted of
one plastic bag containing 150 g of rice seeded with
D. yamadanum. The experiment was arranged in a
completely randomized scheme with four repetitions.
Influence of different supplements
added to the substrate on sporulation
of D. yamadanum
The effect of nutritional supplementation of the basic
substrate (polished rice) on sporulation of D. yamadanum was investigated by following the general procedure described above. The treatments consisted in
supplementing the water added to the basic substrate
(polished rice) before autoclaving with either calcium
carbonate (1.5 g/l), urea (2.0 g/l) or sucrose (20 g/l).
Control consisted of a group of bags containing the
basic substrate without any supplement. Evaluation
of sporulation was performed as previously described.
Nine harvests were done in this experiment, which was
arranged in a completely randomized scheme with four
repetitions.
Effects of different lengths of incubation
of D. yamadanum, before initiation of
harvesting, on sporulation
In this experiment, the general procedure described
above was followed, but different periods of incubation between seeding the substrate and the first harvest
were tested: 3, 4, 5, 6, 7 and 8 days of incubation at the
aperture of the recipients (DIAR). The evaluation was
performed as described above. The experiment was arranged in a completely randomized scheme with four
repetitions.
Statistical analysis
Variance analysis were performed with the software SAS (version 8.3; Statistical Analysis System,
Cary, NC, USA). Conidial production was evaluated
by calculations of areas under the curve as described
by Madden et al. (2007). The assays with an independent quantitative variable were analyzed by comparing
areas under the curve of conidial production (AUCCP)
obtained for each treatment. Variance analysis of the
effects of treatments was performed and compared
with Tukeys test at 5% of probability.
578
Mean concentration, log 10
conidia/mL
Factors affecting mass production of Duosporium yamadanum in rice grains
40%
50%
60%
70%
80%
1
0
50
100
150
200
250
300
350
400
Hours of incubation
Figure 1.
Mean concentration of log 105 conidia per millilitre of Duosporium yamadanum produced for treatments with different percentages of water added to the substrate (for a
total of 13 harvests). Bars represent the standard error.
Results
Conidia production
1600
Effects of different levels of water
content in the substrate on sporulation
of D. yamadanum
The highest sporulation levels were obtained with
a water content ranging from 40% to 60%. The largest yield of conidia was obtained in the second harvest.
A clear reduction in yields was observed on the tenth
harvest, that is, at hour 314 (Fig. 1). No statistical difference was observed for total yield obtained for treatments with 40%, 50% and 60% of water. No colonization or sporulation was obtained for the treatments with
70% and 80% of water (Fig. 2).
1400
CV = 3.97
1200
1000
800
600
400
200
Figure 2.
50%
60%
40%
70% 80%
Conidial production (AUCCP) for different water percentage on the substrate (mean of four
repetitions). Columns with the same letters did
not differ statistically under Tukeys test at 5%
of probability.
highest level of sporulation was obtained for the treatment where calcium carbonate was added to the substrate, there was no statistical difference between this
treatment and the control. Treatments involving supplementation with urea and sucrose led to significantly
less sporulation (Fig. 4).
Influence of different supplements
added to the substrate on sporulation
of D. yamadanum
The highest levels of sporulation were attained for
the first harvest in all treatments (Fig. 3). Although the
4
Calcium carbonate
Mean concentration, log 10 conidia/mL
Sucrose
Control
Urea
0
50
74
98
122
146
170
194
218
242
266
290
Hours of Incubation
Figure 3.
Mean concentration of log 105 conidia per millilitre of Duosporium yamadanum for substrates supplemented with different substances (total of nine harvests). Bars represent the standard error.
579
1040
1020
1000
980
960
940
920
900
880
860
CV =2.20
A
B
Figure 4.
U
re
a
Su
cr
os
e
l
C
on
tro
Ca
rb
on
at
Conidia production
XII International Symposium on Biological Control of Weeds
Conidial production (AUCCP) for substrates
supplemented with different substances (mean
of four repetitions). Columns with the same
letters did not differ statistically under Tukeys
test at 5% of probability.
Effects of different lengths of incubation
of D. yamadanum, before initiation of
harvesting, on sporulation
Allowing D. yamadanum to grow on rice within
the bags for 3 or 4 days was shown to be significantly
better, in terms of sporulation, than the other periods
that were tested. Longer periods of incubation led to
a significant reduction in sporulation. Sporulation was
maintained for longer periods of time for treatments
subjected to shorter periods of incubation and dropped
to zero after eight harvests (for 8 days of incubation)
and at the last harvest (for 7 days of incubation; Figs. 5
and 6).
Discussion
Viability of a mycoherbicide depends heavily on the
development of an adequate methodology for largescale production of fungal propagules (Jackson et al.,
1996). Each fungus has different requirements for opti-
mal production (quantity and quality), and often minor
adjustments in a mass production protocol may have
significant impacts on final results. Although almost
all commercial mycoherbicides have relied on liquid
fermentation, mass production of entomopathogens
has generally relied on solid fermentation, particularly
based on grains, such as rice, as a substrate (Wyss et
al., 2001; Tarocco et al., 2005).
Preliminary attempts by Pomella (1999) to mass
produce D. yamadanum in liquid media failed to yield
any sporulation. Later, the same author demonstrated
that solid fermentation might offer an adequate alternative for mass production. The protocol described by
Pomella (1999) for this purpose is in contrast to the
present study in some respects. For instance, water content in the substrate of 40%, 50% and 60% gave better
sporulation compared to 70% to 80% found to be best
by Pomella (1999). In the present study, higher levels
of water content led to an inadequate consistency of
the substrate. In high water-content treatments, grains
became too soft and water soaked and lack of aeration within the substrate mass probably did not allow
proper colonization of the substrate and sporulation of
the fungus. Other differences between the two production protocols may explain the discrepancies between
observations made in this work and those made by Pomella (1999). In the latter, seeding of the medium was
done with culture disks, harvests were initiated much
later (13 days) and the plastic bags were not completely
sealed but had a cotton plug closing the bags openings.
This might have allowed a progressive dehydration of
the substrate generating conditions that were less favourable for the fungal growth than those provided by
the set of conditions adopted in the present study.
Of particular interest was the persistence of sporulation on colonized rice grains observed in this study. After 13 harvesting episodes, conidia were still relatively
abundant for the best treatments, despite the progressive decline in sporulation. Our utilization of blended
Mean concentration, log 10 5
conidia/mL
3 days
4 days
6 days
5 days
8 days
7 days
1
0
74
98
122
146
170
194
218
242
266
290
Hours of incubation
Figure 5.
Mean concentration of log 105 conidia per millilitre of Duosporium yamadanum for different periods of incubation within sealed plastic bags before initiation of harvesting. Bars represent the standard error.
580
Factors affecting mass production of Duosporium yamadanum in rice grains
1100
1000
900
Conidia production
CV = 3.28
AB
800
700
600
500
400
300
200
100
0
3 Days
4 Days
6 Days
5 Days
7 Days
8 Days
Incubation period
Figure 6.
Conidial production (AUCCP) for different periods of incubation
within sealed plastic bags before initiation of harvesting (mean of
four repetitions). Columns with the same letters did not differ statistically under Tukeys test at 5% of probability.
mycelium as seed allowed for a quick colonization of
the substrate, shortening considerably the process of conidium production for this fungus. The period of spore
production and harvest number was increased.
The levels of moisture that are more appropriate for
mass production of fungi vary from species to species.
For example, the ideal moisture content in the substrate
was shown to be 30% to 40 % for Metarhizium anisopliae (Metsch) Sorok. var. acridum (Arzumanov et
al., 2005) and also for Penicillium oxalicum Currie &
Thom (Larena et al., 2002), whereas for Mucor bacilliformis Hesselt, the ideal is 90% (Lareo et al., 2006).
The long period of incubation within the bags adopted
in Pomellas protocol was shown to be unnecessary.
Pomella waited for a thorough and visible colonization
of the rice mass before opening the bags and starting
the conidial harvest to avoid problems with contamination. However, using the techniques described in
this paper, contamination was not a problem and before D. yamadanum colonies are visible with naked
eye, colonization is well advanced. Keeping the bags
closed for longer periods was shown to be harmful for
sporulation, perhaps because it leads to stress such as
reduced exchange of gases, lack of oxygen, poor heat
dissipation or others.
Supplementing the substrate with urea or sucrose is
known to increase sporulation for many fungi, but in
the case of D. yamadanum, it was clearly harmful. The
addition of calcium carbonate resulted in a statistically
negligible increase in sporulation. Higher concentrations of calcium carbonate might have a more significant effect on sporulation, an aspect deserving further
investigation.
The results obtained in this study provide improvement on the protocol for mass production proposed by
Pomella (1999).
Acknowledgements
This work forms part of research projects submitted as
a D.Sc. dissertation to the Departamento de Fitopatologia/Universidade Federal de Viosa by A.W. Pomella
and as an MSc dissertation to the same department by
D.M. Macedo. The authors thank the Brazilian Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq) and CAPES for financial support.
References
Arzumanov, T., Jenkins, N. and Roussos, S. (2005) Effect of
aeration and substrate moisture content on sporulation of
Metarhizium anisoplidae var. acridum. Process Biochemistry 40, 10371042.
Bariuan, J.V., Reddy, K.N. and Wills, G.D. (1999) Glyphosate injury, rainfastness, absortion, and translocation in
purple nutsedge (Cyperus rotundus). Weed Technology
13, 112119.
Barreto, R.W. and Evans, H.C. (1994) Mycobiota of the weed
Cyperus rotundus in the state of Rio de Janeiro, with an
elucidation of its associated Puccinia complex. Mycological Research 98, 11071116.
Barreto, R.W. and Evans, H.C. (1996) Fungal pathogens of
weeds collected in the Brazilian tropics and subtropics and
their biocontrol potential. In: Delfosse, E.S. and. Scott,
R.R. (eds) Proceedings of the VIII International Symposium on Biological Control of Weeds. DSIR/CSIRO, Melbourne, Australia, pp. 121126.
Frick, K.E. and Quimby, Jr, P.C. (1977) Biocontrol of purple
nutsedge by Bactra verutana Zeller in a greenhouse. Weed
Science 25, 1317.
Frick, K.E., Williams, R.D., Quimby, Jr, P.C. and Wilson,
R.F. (1979) Competitive biocontrol of purple nutsedge
(Cyperus rotundus) and yellow nutsedge (C. esculentus)
with Bactra verutana under greenhouse conditions. Weed
Science 27, 178183.
581
XII International Symposium on Biological Control of Weeds
Holm, L., Plucknett, D.L., Pancho, J.V. and Herberger, J.P.
(1977) The Worlds Worst Weeds. Distribution and Biology. University Press of Hawaii, Honolulu, HI, 609 pp.
Jackson, M.A., Schisler D.A., Slininger, P.J., Boyette, C.D.,
Silman R.W. and Bothast, R.J. (1996) Fermentation strategies for improving the fitness of a bioherbicide. Weed
Technology 10, 645650.
Kadir J.B., Charudattan R., Stall W.M. and Bewick, T.A. (1999)
Effect of Dactylaria higginsii on interference of Cyperus
rotundus with L-esculentum.Weed Science 47, 682686.
Kadir, J.B. and Charudattan, R. (2000) Dactylaria higginsii,
a fungal bioherbicide agent for purple nutsedge (Cyperus
rotundus). Biological Control 17, 113124.
Kadir, J.B., Charudattan, R. and Berger, R.D. (2000a) Effects of some epidemiological factors on levels of disease
caused by Dactylaria higginsii on Cyperus rotundus.
Weed Science 48, 6168.
Kadir, J.B., Charudattan. R. Stall, W.M. and Brecke, B.J.
(2000b) Field efficacy of Dactylaria higginsii as a bioherbicide for the control of purple nutsedge (Cyperus rotundus). Weed Technology 14, 16.
Larena, I., Melgajero, P. and De Cal, A. (2002) Production,
survival, and evaluation of solid-state inocula of Penicillium oxalicum, a biocontrol agent against Fusarium wilt
of tomato. Phytopathology 92, 863869.
Lareo, I., Sposito, A.F., Bossio, S.L. and Volpe, D.C. (2006)
Characterization of growth and sporulation of Mucor bacilliformis in solid fermentation on an inert support. Enzyme
and Microbial Technology 38, 391399.
Macedo, D.M. (2006) Duosporium yamadanum: Produo
massal, formulao e associao com herbicidas para o
controle de tiririca. MSc thesis. Universidade Federal de
Viosa, Viosa, Brazil, 54 pp.
Madden, L.V., Hughes, G., and van den Bosch, F. (2007)
Study of Plant Disease Epidemics. American Phytopathological Society, Saint Paul, USA, 421 pp.
Neto, C.R.B. (1997) Estudos sobre Cercospora caricis Oudem, como agente potencial de biocontrole de tiririca
(Cyperus rotundus L.). MSc thesis. Universidade de Braslia, Brazil, 122 pp.
Okoli, C.A.N., Shilling, D.G., Smith, R.L. and Bewick, T.A.
(1997) Genetic diversity in purple nutsedge (Cyperus rotundus L) and yellow nutsedge (Cyperus esculentus L).
Biological Control 8, 111118.
Pandey, A. (2003) Solid-state fermentation. Biochemical Engineering Journal 13, 8184.
Phatak, S.C., Sumner D.R., Wells, H.D., Bell D.K. and Glaze,
N.C. (1982) Biological control of yellow nutsedge with
the indigenous rust fungus Puccinia canaliculata. Science
219, 14461447.
Phatak, S.C, Callaway, M.B. and Vavrina, C.S. (1987) Biological control and its integration in weed management
systems for purple and yellow nutsedge (Cyperus rotundus and C. esculentus). Weed Technology 1, 8491.
Pomella, A.W.V. (1999) Avaliao do fungo Duosporium
yamadanum no controle biolgico da tiririca (Cyperus
rotundus). DSc thesis. Universidade Federal de Viosa,
Brazil, 183 pp.
Prakash, O., Kumar, R., Dev, J. and Chakrabarti, D.K. (1996)
Biological control of nutgrass (Cyperus rotundus) in greengram (Phaseolus radiatus). Indian Journal of Agricultural Sciences 66, 490493.
Ribeiro, Z.M.A., Mello, S.C.M., Furlanetto, C., Figueiredo,
G. and Fontes, E.M. (1997) Characteristics of Cercospora
caricis, a potential biocontrol agent of Cyperus rotundus.
Fitopatologia Brasileira 22, 513519.
Rosskopf, E.N.R., Charudattan, R. and Kadir, J. B. (1999)
Use of plant pathogens in weed control. In: Bellows, T.S.
and Fisher, T.W. (eds) Handbook of Biological Control.
Academic, San Diego, USA, pp. 891918.
Tarocco, F., Lecuona, R.E., Couto, A.S. and Arcas, J.A.
(2005) Optimization of erythritol and glycerol accumulation in conidia of Beauveria bassiana by solid-state fermentation, using response surface methodology. Applied
Microbiology Biotechnology 68, 481488.
Tebeest, D.O. and Templeton, G.E. (1985) Mycoherbicides:
progress in the biological control of weeds. Plant Disease
69, 610.
Upadhyay, R.K., Kenfield, D., Strobel, G.A. and Hess, W.M.
(1991) Ascochyta cypericola sp. nov. causing leaf blight
of purple nutsedge (Cyperus rotundus). Canadian Journal
of Botany 69, 797802.
Walker, H.L. (1980) Production of spores for field studies.
Advances in Agricultural Technology 12, 15.
Wyss, G.S., Charudattan, R. and Devalerio, J.T. (2001)
Evaluation of agar and grain media for mass production
of conidia of Dactylaria higginsii. Plant Disease 85,
11651170.
582
Biological control of tansy ragwort (Senecio
jacobaeae, L.) by the cinnabar moth, Tyria
jacobaeae (CL) (Lepidoptera: Arctiidae),
in the northern Rocky Mountains
G.P. Markin1 and J.L. Littlefield2
Summary
The control of tansy ragwort on the coast of western North America is a major success story for weed
biological control. However, tansy ragwort is still expanding into the colder interior regions of the
Pacific Northwest of the United States where previous efforts to establish the same complex of agents
have failed. We have successfully established one of the agents, the cinnabar moth, Tyria jacobaeae
L., on a major new tansy ragwort infestation in the mountains of northwestern Montana. The cinnabar
moth is still expanding its range, but in the areas where first released, it has given excellent control,
having eliminated tansy ragwort as a visible component in the forest ecosystem while not impacting
native Senecio species. Although establishment in other areas has been slower, we predict that we
will eventually control tansy ragwort over most of its range in the northern Rocky Mountains of the
United States.
Keywords: tansy ragwort, Senecio jacobaea, cinnabar moth, Tyria jacobaeae.
Introduction
In the Pacific Northwest corner of the United States,
tansy ragwort, Senecio jacobaea L., (Asteraceae), an
introduced European forb, is an invasive weed in pastures, native meadows and open forests (Coombs et al.,
1991, 1999). It is a particularly serious problem for
grazing livestock because it contains toxic alkaloids.
Along the Pacific Northwest coast in the 1960s and
1970s, a USDA-ARS program successfully established
three biological control agents and resulted in one of
the most successful biological control weed programs
in North America (Turner and McEvoy, 1995; Coombs
et al., 1996, 2004; Julien and Griffith, 1998). Tansy
ragwort, however, is still spreading east of the Cascade
US Forest Service, RMRS, Forestry Sciences Laboratory, Bozeman,
MT 59717, USA.
2
Montana State University, Department of LRES, Bozeman, MT 59717,
USA.
Corresponding author: G.P. Markin <[email protected]>.
CAB International 2008
1
Mountains into eastern Oregon, Washington and northern Idaho.
In 1994, a wild fire burned 6100 ha of fir and pine
forests in a mountainous area straddling the boundary
between Lincoln and Flathead Counties in northwestern
Montana. Tansy ragwort was probably already present,
but after the fire, an explosive flush of new plants occurred, which was first noticed in 1996. Management
plans were immediately initiated with the goal of herbicide spraying to begin in 1997. In attempting to obtain
funding for the program, it was necessary to develop an
integrated management strategy, and biological control
was added as an afterthought.
Initially, we felt the three biological control agents
used on the west coast would fail in Montana, but the
funding provided the chance to investigate why these
agents had not established previously when released
east of the Cascade Mountains. Our preliminary studies showed that all three agents would establish, but
in Montana, the cinnabar moth, Tyria jacobaeae (L.)
(Lepidoptera: Arctiidae) was particularly suitable. This
paper describes the success we observed in subsequent
years to the release of this moth.
583
XII International Symposium on Biological Control of Weeds
Materials and methods
Source of cinnabar moths
The first colonies of cinnabar moths, obtained from
Oregon in 1996 from the Willamette Valley and along
the coast near Florence, were used for preliminary nontarget host studies. Additional shipments from the Willamette Valley in 1997 were used for three field-cage
releases within the area burned by the forest fire in Flathead County in Montana. In 1998, a new moth population was located near Mount Hood at 1300-m elevation
in the Cascade Mountains of Oregon in a forested site
with a heavy winter snow cover, similar to northwestern
Montana. From 1998 to 2000, egg masses from the Mount
Hood population were shipped to Bozeman, sterilized
by soaking in a 0.1% sodium hypochloride for 5 min
and rinsed in running water to eliminate bacterial or viral contaminants. After hatching, first-instar larvae were
held, ten individuals to a Petri dish, for 10 to 14 days,
fed tansy ragwort leaves treated with 100 ppm of the
fungicide Benomyl (Benlate sp. fungicide, DuPont Agricultural Products, Willmington, DE, USA) to eliminate microsporidium that were reported to be present in
the west coast cinnabar moth populations (Bucher and
Harris, 1961; Hawkes, 1973). If any larvae died, the entire batch was discarded. Larvae that reached third instars without mortality were used for subsequent studies
in the laboratory and for field release in 1998 to 2000.
Non-target plant testing
Initially, there were questions concerning using the
cinnabar moth because, in early laboratory testing, it
fed on several North American Senecio sp. (Bucher and
Harris, 1961) and had been reported attacking a field
population of Senecio triangularis Hook in Oregon
(Diehl and McEvoy, 1989). It was therefore necessary
to determine whether Montana varieties of S. triangu
laris might also be at risk. In 1996, no-choice feeding
tests using third-instar larvae in Petri dishes offering
either Montana S. triangularis or tansy ragwort leaves
were set up, and the rate of development, survival and
weight of any pupae produced was determined. In 1997
and 1998, the tests were repeated but expanded to look
at other species of Senecio found in the Montana tansy
ragwort area. Those species were Senecio (sym: Pac
kera) pseudaureus Rydb., Senecio hydrophilus Nutt.,
Senecio integerrimus Nutt. and Senecio (sym: Packera)
canus Hook (Hanson, 2000). When these tests indicated
that S. pseudaureus could be fed on, the first field releases in 1997 and 1998 (of 300 larvae each) were made
in 2 4 m field cages (three each year) containing intermixed S. pseudaureus and tansy ragwort plants.
Open field release
By 1999, studies indicated that the cinnabar moth
was not a threat to native Senecio, and the rapid popu-
lation increases in cages indicated it might be a useful
biological control agent for Montana. The original six
cages were removed that year so the moths could disperse naturally, and we began a program of additional
releases to spread them as rapidly as possible through
the remaining tansy ragwort area.
The tansy ragwort infestation in Montana occurs in
two counties, Flathead County and Lincoln County,
each differing environmentally and in land ownership.
In Flathead County, the infestation was restricted to
National Forest lands in the eastern half of a 4000 ha
area of remote rugged fir and pine forest (elevation,
13501560 m) that was burned in a wildfire in 1994
and subsequently salvage logged. Initially, the tansy
ragwort infestation was thought to cover only 100 ha,
and the program was aimed at controlling and, if possible, eradicating the infestation using herbicides. By
1997, additional surveys indicated that there were at
least 500 ha to be sprayed. There were also numerous
small environmentally sensitive sites located close to
water where herbicide was not allowed to be used, and
these were used for the initial biological control studies. Besides the six caged releases in 1997 and 1998,
uncaged releases of 300 larvae each at eight new field
sites in the burned area and three in the surrounding
unburned forest were made in 1999.
In 2000, the program was expanded westward into
the Little Wolf Creek drainage in Lincoln County
where the remaining third of the area burned by the
wildfire also contained dense stands of unsprayed
tansy ragwort. This area was primarily the property
of a timber company, which thought chemical control
was not economically justifiable. In 2000 and 2001,
12 releases of 300 early instar larvae, collected from the
now well-established populations in Flathead County,
were made. From 2002 to 2004, emphasis shifted to
a third area approximately 12 km to the southwest of
Little Wolf Creek in the vicinity of Island Lake, a small
mountain lake. The land ownership here was a mixture
of National Forest, private timber company and private
ranches. The area differed from the first two targeted
areas, as it was unburned and consisted of an 8000-ha
mosaic of different aged fir, pine and larch forests and
open meadows. The area was extensively disturbed by
heavy grazing and logging. Tansy ragwort was concentrated in the more open and disturbed sites and was
as abundant and dense as in the burned area. Seven
releases of 300 early instar larvae from the Flathead
County population were made in 2002.
Monitoring impact
During the three caged releases made in 1997, and
repeated in 1998, all tansy ragwort plants in the cages
were recorded as either seedlings, rosettes (would not
bolt that year) or mature plants that had bolted. These
gave estimations of populations within the six caged
areas and were compared to populations at six similar
584
Biological control of tansy ragwort (Senecio jacobaeae, L.) by the cinnabar moth
but uncaged areas between 50 and 100 m away, at
which no releases were made. In 1999, when the cages
were removed, a permanent marker was placed at the
centre of each of these 12 study sites. At 1 m radius out,
all tansy plants found within four 25-cm2 quadrats were
counted to determine site density. Monitoring through
the remainder of this program continued at the six ini
tial caged release sites and the six uncaged areas. Within
a few years, all uncaged areas had been inundated by
the cinnabar moth so all data has been combined for an
average of the 12 study sites.
When the program changed from research in 1999
to an operational program in which we were trying to
redistribute the cinnabar moth as rapidly as possible,
no further detailed monitoring was conducted, although all new releases were marked. These sites were
visited annually, and visual estimates were made of the
abundance of tansy ragwort and cinnabar moth larvae
in mid- to late July when the plants were in flower and
the larvae were most active and visible. Larvae population estimates were made by walking a 50-m transect
in 5 min while counting or estimating the number of
larvae seen feeding on the flower heads and, later in the
study, on any surviving rosettes.
Results
Non-target feeding
Laboratory feeding tests showed that S. triangularis
from Montana was an unsuitable host (Table 1). Feeding occurred in starvation tests, but development was
much slower and produced only a few small pupae.
Even poorer development was seen on S. hydrophilus,
S. integerrimus and S. canus. Subsequent observations
in the field on natural populations of S. triangularis
and S. hydrophilus intermixed with tansy ragwort at
Table 1.
five locations in Flathead County showed that, even
during the peak population of the cinnabar moth and
its collapse after the elimination of the tansy ragwort,
both species were totally ignored. Why S. triangularis
should be fed on in the field in Oregon but not in Montana has not been resolved.
By contrast, S. pseudaureus in laboratory no-choice
feeding tests supported almost normal development
of larvae, although the larvae would not feed on it in
choice tests when tansy ragwort was also available.
Subsequent field observations showed that, during the
first 2 or 3 years of the cinnabar moth build up, S. pseud
aureus was ignored. However, when the supply of
tansy ragwort was exhausted, females would occasionally lay egg masses on S. pseudaureus. The resulting
larvae skeletonized the leaves that contained the egg
mass and then disappeared. Furthermore, if late-instar
larvae totally consumed adjacent tansy ragwort plants
and dispersed in search of food, they occasionally fed
on S. pseudaureus. This minor feeding was observed
for a year or two until the tansy ragwort had disappeared, and when the cinnabar moth population collapsed, feeding on S. pseudaureus ceased. At no time
after the disappearance of tansy ragwort in 2003 have
any cinnabar moths been found utilizing S. pseudau
reus as a permanent host. S. pseudaureus in Montana
therefore will not support a permanent population of
cinnabar moths and suffers only temporary attack when
adjacent tansy plants are stripped of foliage. The potential for it to attack other Senecios exists, however, so
this moth should not be released in new areas without
preliminary host testing of local Senecio.
Establishment of the cinnabar moth
Of the six original cage releases, five built up populations that, by the second year, were causing 50%
omparison of survival rate and development times for larvae and pupal weight of the cinnabar moth raised on
C
tansy ragwort, Senecio jacobaeae, and other species of Senecio native to Montana.
Species
1997
Senecio jacobaea
Senecio pseudaureus
Senecio hydrophilus
Senecio triangularis
Senecio integerrimus
Senecio canus
1998
Senecio jacobaea
Senecio pseudaureus
Senecio hydrophilus
Senecio triangularis
Senecio integerrimus
Senecio canus
Larvae no.
Pupae no.
Pupate %
26
24
28
21
22
25
24
16
16
2
0
0
92.3(a)a
66.7(b)
57.1(b)
9.5(c)
0.0
0.0
22.6(a)a
25.0(a)
32.6(a,b)
46.0(b)
0.0
0.0
0.122(a)a
0.110(a)
0.070(b)
0.070(b)
0.000
0.000
43
42
29
21
30
33
34
38
22
2
1
1
90.5(a)
79.1(a)
75.9(a)
9.5(b)
3.3(b)
4.4(b)
20.4(a)
21.9(a)
28.2(b)
39.5(b,c)
29.0(c)
40.0(c)
0.130(a)
0.110(a)
0.080(b)
0.080(b)
0.090(a,b)
0.090(b,b)
Each year, grouping is compared based on Tukey HSD at 0.05 level.
585
Days to pupate
Pupae weight (g)
XII International Symposium on Biological Control of Weeds
defoliation. Of the eight additional open releases, all
established and spread and, by 2004, had reached 95%
of the infested area. In Flathead County, the cinnabar
moth population collapsed after the disappearance of
the tansy ragwort around 2003. While plants continue
to sprout from the soil seed bank and may escape detection while small rosettes, within a year of flowering,
they are usually found by the moth and destroyed. This
pattern of low numbers of cinnabar moth and low plant
number equilibrium has been observed in Oregon for
the last 20 years and we expect is now the situation in
Flathead County. The exception is a small (less than
a quarter hectare) site in which a dense stand of tansy
ragwort has, so far, escaped attack in the moist bottom
of a narrow valley.
In the Little Wolf area, Lincoln County, six of seven
releases, which were made on a hillside, established
but population build up was slower than in the Flathead County sites. A large amount of tansy ragwort,
resulting from the lack of a spray program, is present.
Consequently, it will take longer for the cinnabar moth
population to reach a level capable of overwhelming
this huge biomass, a problem not encountered in Flathead County where the majority of the tansy ragwort
biomass had been eliminated by herbicide spraying. By
contrast, the five releases made within 30 m of Little
Wolf Creek failed. It is presumed that these and the failure in Flathead County are due to micro-climatic effects
restricted to these narrow mountain valley bottoms.
In Flathead County, establishment occurred at three
sites in unburned forest, but our initial six releases, of
300 larvae each, in the unburned area around Island
Lake failed. In 2004, releases of between 1000 and
Figure 1.
2000 late instar larvae were made at ten sites around
the lake. By 2005, most of these releases established.
By 2006, five had expanded, and they were totally defoliating the tansy in areas ranging from 0.5 to 5 ha; three
had low populations that were not causing significant
defoliation, and two appeared to have failed. We have
no explanation why small releases of early instar larvae
failed, but large releases of late instar larvae gave rise
to rapidly expanding populations.
Impact
Tansy ragwort density at the 12 original sites in
Flathead County remained constant for the first 3 or
4 years, while the populations of the cinnabar moth built
up. They then declined to a low in 2003 when tansy
ragwort was almost undetectable (Fig. 1). The cinnabar
moth population then collapsed due to the lack of food.
The tansy ragwort re-sprouted from the seed bank, and
the peak in 2004 represents only new seedlings or very
small rosettes. However, the cinnabar moth soon reappeared at all sites and suppressed the plant. Today, the
tansy ragwort and the cinnabar moth appear to be in
equilibrium, maintaining the plant population at a much
suppressed level. The major visible impact of our program has been the almost total disappearance of flowering tansy ragwort plants since 2002. The Little Wolf
and Island Lake areas in Lincoln County now have
well-developed moth populations, and in many areas,
they have caused the demise of mature plants. We predict that these populations will continue to increase and
spread and that they will reduce tansy ragwort to levels
comparable to those in Flathead County.
Density of tansy ragwort, Senecio jacobaeae, showing the total numbers of plants and the numbers that were
mature and bolted. Data are the mean of 12 permanent plots at Flathead County, MT. The estimated numbers of
late-instar larvae of the cinnabar moth, Tyria jacobaeae, counted in 5 min along a 50-m transect at the same 12
sites. No estimate for 1997 or 1999 since the cinnabar moth was confined in field cages.
586
Biological control of tansy ragwort (Senecio jacobaeae, L.) by the cinnabar moth
Comparison of effectiveness between
Montana and Oregon
The success of the cinnabar moth in Montana raises
the question why this insect fails to control tansy ragwort along the west coast (van der Meijden, 1979;
Myers, 1980; Crawley and Gillman, 1989). This may
be due to climate differences affecting the phenology
of the plant. In Oregon, the cinnabar moth was often
observed defoliating large mature plants. However,
the plant compensated with re-growth during the following mild, wet fall and winter (Hawkes, 1981; Cox
and McEvoy, 1983; Turner and McEvoy, 1995). In
Montana, plants that survive defoliation do not recover
noticeably before the onset of cold weather in October and permanent snow cover in November (Fig. 2).
Surviving plants emerge the following spring as small
rosettes that seldom flower. Several consecutive years
of cinnabar moth feeding on even the strongest plants
in Montana continue to reduce them in size until they
are eventually killed. It is interesting that the only other
location where the cinnabar moth is credited with controlling tansy ragwort is the eastern Maritime Provinces
of Canada (Harris et al., 1973, 1978), an area similar to
Montana with a long cold winter and snow cover that
probably protect the over wintering pupae in the soil
from freezing (Fig. 2).
Future of the cinnabar moth in Montana
An integrated control program
The rapid success obtained in eliminating tansy ragwort in Flathead County was due to a combination of
herbicide application and biological control. There was
an intensive herbicide spray program between 1997
and 1999 that probably killed 99% of mature, flowering plants. The only unsprayed tansy ragwort was in the
buffers around springs, moist seeps or riparian zones,
where spraying was prohibited and which was used for
the biological control study. However, after spraying, a
flush of seedlings was observed, and by 2001 and 2002,
flowering plants began to reappear as the replacement
generation matured. At this time, the cinnabar moth
population was well established at our research sites,
spreading rapidly through the surrounding tansy ragwort area and, within a few years, eliminated the need
for additional chemical treatments. Chemical treatment
of tansy in Flathead County is now limited to roadside
spraying of any isolated plants found to prevent their
86
30
77
25
68
20
59
15
50
10
41
32
23
-5
6/13/1998
7/13/1998
8/13/1998
9/13/1998
10/13/1998
11/13/1998
12/13/1998
1/13/1999
2/13/1999
3/13/1999
4/13/1999
5/13/1999
6/13/1999
7/13/1999
8/13/1999
9/13/1999
10/13/1999
11/13/1999
12/13/1999
1/13/2000
2/13/2000
3/13/2000
4/13/2000
5/13/2000
6/13/2000
7/13/2000
8/13/2000
9/13/2000
10/13/2000
Temp (F)
On the west coast, the cinnabar moth populations
are affected by severe diseases (Bucher and Harris,
1961; Hawkes, 1973). We took great efforts to eliminate diseases from the populations that we introduced
to Montana, and our monitoring there has detected
no sign of disease that could limit the cinnabar moths
effectiveness.
To counter this possibility, we are continuing our
effort to establish the tansy ragwort flea beetle, Lon
gitarsus jacobaeae (Waterhouse), in those areas where
the cinnabar moth has not established. Hopefully, flea
beetle colonies will be numerous enough that, if the
cinnabar moth population eventually collapses, they
will be ready to replace it (see Littlefield et al., this
volume).
Temp (C)
Discussion
Figure 2.
Soil temperatures 2 cm below the surface at site 1 in the Flathead National Forest. Relatively flat lines
from November to April indicate snow cover.
587
XII International Symposium on Biological Control of Weeds
seeds from being carried out of the area and the spraying of isolated satellite population, as they are found in
outlying areas.
The successful control in Flathead County, although
unplanned, is an excellent example of how an integrated
control program for a new weed can be implemented.
In this case, herbicides contained a new infestation and
suppress seed production in the core area long enough
for the biological control agents to establish and build
up populations capable of dispersing and overwhelming the suppressed population.
Acknowledgements
We are deeply grateful to the efforts of Carol Horning
who supported our program by collecting and shipping
the cinnabar moth to Montana and to Eric Coombs of
the Oregon Department of Agriculture who provided
much useful information based on his extensive personal experience with the tansy ragwort program and
the cinnabar moth in Oregon. Finally, we wish to thank
Terry Carter, the late vegetation manager for Flathead
National Forest, and Ann Odor, Bill Chalgren and Dan
Williams, vegetation managers in Lincoln County for
their invaluable support that made this biological control program possible.
References
Bucher, G.E. and Harris, P. (1961) Food-plant spectrum and
elimination of disease of cinnabar moth larvae, Hypocrite
jacobaeae (L.) (Lepidoptera: Arctiidae). Canada Ento
mologist 93, 931936.
Coombs, E.M., Bedell, T.E., and McEvoy, P.B. (1991) Tansy
ragwort (Senecio jacobaea): importance, distribution and
control in Oregon. In: James, L.F., Evans, J.O., Ralphs,
M.H. and Child, R.D. (eds) Noxious Range Weeds. Westview Press, Boulder, CO, USA, pp. 419428.
Coombs, E.M., Radtke, H., Isaacson, D.L., and Snyder, S.P.
(1996) Economic and regional benefits from the biological control of tansy ragwort, Senecio jacobaea, in Oregon.
In: Moran, V.C. and Hoffmann, J.H. (eds) Proceedings of
the 9th International Symposium on Biological Control
of Weeds. University of Cape Town, Stellenbosch, South
Africa, pp. 489494.
Coombs, E.M., McEvoy, P.B., and Turner, C.E. (1999) Tansy
ragwort. In: Sheley, R.L. and Petroff, J.K. (eds) Biology
and Management of Noxious Rangeland Weeds. Oregon
State University Press, Corvallis, OR, USA, pp. 389
400.
Coombs, E.M., McEvoy, P.B., and Markin, G.P. (2004) Tansy
ragwort, Senecio jacobaea. In: Coombs, E.M., Clark,
J.K., Piper, G.L. and Cofrancesco, Jr., A.F. (eds) Bio
logical Control of Invasive Plants in the United States.
Oregon State University Press, Corvallis, OR, USA,
pp. 335344.
Cox, C.S. and McEvoy, P.B. (1983) Effect of summer moisture
stress on the capacity of tansy ragwort (Senecio jacobaea)
to compensate for defoliation by cinnabar moth (Tyria ja
cobaea). Journal of Applied Ecology 20, 225234.
Crawley, M.J. and Gillman, G.P. (1989) Population dynamics of cinnabar moth and ragwort in grassland. Journal of
Animal Ecology 58, 103550.
Diehl, J.W. and McEvoy, P.B. (1989) Impact of the cinnabar
moth (Tyria jacobaeae) on Senecio triangularis, a nontarget native plant in Oregon. In: Delfosse, E.S. (eds)
Proceedings of the 8th International Symposium on Bio
logical Control of Weeds. Istituto Sperimentale per la Patologia Vegetale, MAF. Rome, Italy, pp. 119126.
Hanson. E. (2000) Plants, database 3/2000 Alphabetical
listing of scientific and synonyms (Old Names). USDA
Forest Service Handbook, Forest Inventory and Analy
ses. Portland Forestry Sciences Lab, Portland, OR, USA
(p. 436).
Harris, P., Wilkinson, A.T.S., Thompson, L.S. and Neary,
M. (1978) Interaction between the cinnabar moth, Tyria
jacobaeae L. (Lep.: Arctiidae) and ragwort, Senecio ja
cobaea L. (Compositae) in Canada. In: Freeman, T. (eds)
Proceedings of the 6th International Symposium on Bio
logical Control Weeds. University of Florida, Gainesville,
FL, USA, pp. 174180.
Harris, P., Wilkinson, A.T.S. and Myers, J.H. (1984) Senecio
jacobaeae, tansy ragwort (Compositae). In: Kelleher, J.S.
and Hulme, M.A. (eds) Biological Control Programmes
Against Insects and Weeds in Canada 19691980. Commonwealth Agricultural Bureaux, UK, pp. 195201.
Hawkes, R. B. (1973) Natural mortality of cinnabar moth in
California. Annals of the Entomological Society of Amer
ica 66, 137146.
Hawkes, R.B. (1981) Biological control of tansy ragwort in
the state of Oregon, U.S.A. In: Delfosse, E.S. (ed) Pro
ceedings of the 5th International Symposium on Bio
logical Control of Weeds. CSIRO Entomology, Brisbane,
Australia, pp. 623626.
Julien, M.H. and Griffith, M.W. (1998) Biological Control
of Weeds. A World Catalogue of Agents and their Tar
get Weeds, 4th edn. CABI Publishing, Wallingford, UK,
223 pp.
Myers, J.H. (1980) Is the insect or the plant the driving force
in the cinnabar moth-tansy ragwort system? Oecologia
47, 1621.
Turner, C.E. and McEvoy, P.B. (1995) 71/Tansy ragwort. In:
Nechols, J.R., Andres, L.A., Beardsley, J.W., Goeden,
R.D., and Jackson, C.G. (eds) Biological Control in the
Western United States. Publication 3361. Division of Agriculture and Natural Resources, University of California,
USA, pp. 264269.
van der Meijden, E. (1979) Herbivore exploration of the fugitive plant species: Local survival and extinction of the
cinnabar moth and ragwort in a heterogeneous environment. Oecologia 42, 307323.
588
Establishment, spread and initial impacts
of Gratiana boliviana (Chrysomelidae)
on Solanum viarum in Florida
J. Medal,1 W.A. Overholt,2 P. Stansly,3 A. Roda,4 L. Osborne,5
K. Hibbard,6 R. Gaskalla,7 E. Burns,7 J. Chong,4 B. Sellers,8 S.D. Hight,9
J.P. Cuda,1 M. Vitorino,10 E. Bredow,11 J.H. Pedrosa-Macedo11 and C. Wikler12
Summary
Solanum viarum Dunal (Solanaceae) is an invasive perennial shrub in southeastern USA. Native to
South America, it was first found in Florida in 1988, and it has already invaded more than 400,000
ha of grasslands and conservation areas in 11 states. Currently recommended control tactics for this
weed in pastures are based on herbicide applications combined with mechanical (mowing) practices.
These control tactics provide a temporary solution and can cost as much as $188/ha for dense infestations of the weed. A biological control project against S. viarum was initiated in 1997. After 3 years
of intensive host-specificity testing, the South American leaf beetle Gratiana boliviana was approved
for field release by the United States Department of Agriculture (USDA)-Animal and Plant Health
Inspection Service (APHIS)-Plant Protection and Quarantine (PPQ) in 2003, and its release in Florida
began in summer 2003. Up to now, approximately 120,000 beetles have been released in 25 counties
in Florida. The beetles established at virtually all the release sites in Florida. Beetle dispersal has been
based on plant availability with annual dispersal from 1.6 to 16 km/year from the release sites. Initial
impacts of the beetles range from 30% to 100% plant defoliation. The fruit production declined from
40 to 55 fruits per plant in summer 2003, when beetles were released, to zero or a few deformed fruits
(one to four per plant) 2 years post release in five of the release sites monitored. Mass rearing, field
release and post-release evaluation of G. boliviana and the target plant will continue during 2008.
Keywords: invasive plant, weed biological control, monitoring.
Introduction
Solanum viarum Dunal (Solanaceae) is a perennial
shrub from South America that has been spreading
throughout Florida at an alarming rate during the last
two decades. The pastureland infested in 1992 was estimated in approximately 60,000 ha (Mullahey et al.,
University of Florida, POB 110620. Gainesville, FL 32611, USA.
University of Florida, Indian River REC. 2199 S. Rock Rd. Ft. Pierce,
FL 34945, USA.
3
University of Florida, Southwest FL-REC, 2686 Hwy 29N. Immokalee, FL 34142, USA.
4
USDA-APHIS-PPQ-CPHST, Subtropical Horticulture Research Station, 13601 Old Cutler Rd., Miami, FL 33158, USA.
5
University of Florida, Mid-Florida-REC, 2725 Binion Rd., Apopka, FL
32703, USA.
6
Florida Department of Agriculture and Consumer Service-Division of
Plant Industry, 3513 South US-1. Ft. Pierce, FL 34982, USA.
CAB International 2008
1
2
1993), and this infested area increased to more than
300,000 ha in 19951996 (Mullahey et al., 1997).
Currently, the infested area is estimated at more than
400,000 ha (Medal et al., 2004; Medal, 2005). S.
viarum was first reported in the United States in Glades
County, FL, in 1988 (Coile, 1993; Mullahey and Colvin,
1993). This weed also is present in Alabama, Arkansas,
Florida Department of Agriculture and Consumer Service-Division of
Plant Industry, 1911 SW. 34th Street. POB 147100, Gainesville, FL
32614, USA.
8
University of Florida, Range Cattle-REC, 3401 Experiment Station,
Ona, FL 33865, USA.
9
USDA-ARS-CMAVE, 6383 Mahan Dr., Tallahassee, FL 32308, USA.
10
Universidade Regional de Blumenau. Blumenau, Santa Catarina,
Brazil.
11
Universidade Federal do Paran. Rua Lothario Meissner, 3400, Curitiba, PR, Brazil.
12
Universidade do Centro-Oeste. Irati, PR 84500, Brazil.
Corresponding author: J. Medal <[email protected]>.
7
589
XII International Symposium on Biological Control of Weeds
Georgia, Louisiana, Mississippi, North Carolina, Pennsylvania, South Carolina, Tennessee, Texas and Puerto
Rico (Bryson and Byrd, 1996; Dowler, 1996, Mullahey
et al., 1997; Medal et al., 2003). However, infestations
in these states have still not reached high levels. The
potential range of S. viarum in the United States can be
extended even further based on studies of the effects
of temperature and photoperiod conducted by Patterson (1996) in controlled environmental chambers. This
invasive exotic weed was placed on the Florida and
Federal Noxious Weed Lists in 1995.
In addition to its invasion of pasture lands and
reduction of cattle carrying capacity (Mullahey et al.,
1993), S. viarum is a host of at least six viruses that
affect vegetable crops including tomato, tobacco and
pepper (McGovern et al., 1994a,b, 1996). Furthermore,
it is also an alternate host for agricultural pests such
as the Colorado potato beetle, Stilodes (=Leptinotarsa)
decemlineata (Say) (Coleoptera: Chrysomelidae), a
major defoliating insect pest of potato in North America; tomato hornworm Manduca quinquemaculata
(Haworth) (Lepidoptera: Sphingidae), a major pest of
tomato and tobacco plants; and the silverleaf whitefly
Bemisia argentifolii Bellows and Perring (Homoptera:
Aleyrodidae), one of the most troublesome insect pests
worldwide of many field and vegetable crops (Habeck
et al., 1996; Medal et al., 1999). Although it is very
difficult to determine the real (direct and indirect) economic losses due to this invasive weed, Mullahey et al.
(1996) estimated the annual production loss to Florida
ranchers was US$11 million in 1993.
Native to southern Brazil, Paraguay, northeastern
Argentina and Uruguay (Nee, 1991), S. viarum has
spread into other parts of South and Central America
including Mexico, Nicaragua, Honduras and Costa Rica
(J. Medal, personal communication). This weed also
has spread into other regions including the Caribbean
(confirmed in Puerto Rico), Africa, India, Nepal and
China (Chandra and Srivastava, 1978; Coile, 1993).
The rapid spread in Florida can be partially attributed
to the high reproductive potential (Akanda et al., 1996;
Pereira et al., 1997) and effective seed dispersal by
cattle and wildlife, such as deer, feral hogs, raccoons
and birds that feed on the fruits (Mullahey et al., 1993;
Bryson et al., 1995; Brown et al., 1996).
One S. viarum plant can produce on average from
100 to 160 fruits and 41,000 to 50,000 seeds with a
germination rate of at least 75% (Mullahey et al., 1993;
Pereira et al., 1997). The extent of the infestation is
increasing rapidly in the United States, making this a
national rather than just a Florida problem.
Current management practices for S. viarum in Florida are based on herbicide applications combined with
mechanical (mowing) practices (Mislevy et al., 1996,
1997; Sturgis and Colvin, 1996; Akanda et al., 1997).
These control tactics provide temporary weed suppression at an estimated cost of US$185/ha to control dense
infestations of S. viarum (Mullahey et al., 1996). In ad-
dition to being expensive, the application of herbicides
is not always feasible in rough terrain or inaccessible
areas.
A biological control project on this highly invasive
non-native weed was initiated in January 1997 by the
University of Florida in collaboration with the Universidade Estadual Paulista, Jaboticabal campus, Brazil;
Universidade Federal do Paran in Curitiba, Brazil;
Universidade Regional de Blumenau, Santa Catarina state, Brazil; Universidade Centro-Oeste, in Irati,
Paran state, Brazil; Instituto Nacional de Tecnologa
Agropecuaria (INTA-Cerro Azul), Misiones province,
Argentina; and the United States Department of Agriculture (USDA)-Agricultural Research Service (ARS)
Biological control laboratory in Hurlingham, Buenos
Aires province, Argentina. Foreign explorations in
South America identified several insects as potential
biological control agents of S. viarum including three
leaf beetles, Gratiana boliviana Spaeth, Metriona
elatior Klug and Gratiana graminea Klug (Chrysomelidae) and the flower-bud weevil Anthonomus tenebrosus (Boheman) (Curculionidae). These potential
agents were initially selected for screening because
of the extensive foliage/flower bud plant damage attributed to these beetles in their native range (Medal
et al., 1996, 1999, 2006). Two other promising biological control candidates that are currently undergoing
open-field host-specificity tests in Brazil are the leaf
beetle Platyphora sp. and a flea beetle (Chrysomelidae)
(H. Medal, unpublished data).
G. boliviana was approved for field release in the
United States by USDA-Animal and Plant Health Inspection Service (APHIS)-Plant Protection and Quarantine (PPQ) in May 2003. A high level of specificity
and significant defoliation of S. viarum was indicated
in host-specificity tests (Gandolfo et al., 1999, 2007;
Medal et al., 2002, 2004). Field releases of G. boliviana in the United States began in May 2003. Requests
for field releases of the leaf-feeder beetles M. elatior
and G. graminea in the United States were submitted to
TAG (Technical Advisory Group for Biological Control Agents of Weeds) in September and October 2006,
respectively.
Release of G. boliviana in Florida
A total of 120,000 beetles have been released in 25
Florida counties since the summer of 2003. Florida
counties where beetles have been released are shown
in Fig. 1. The number of beetles released at each location varied from 30 to 2000 based on beetle availability
and density of the S. viarum infestation.
A new release technique was used for the first time
in the S. viarum biological control project. S. viarum
plants that were infested with beetles (approximately
100 per plant) in the greenhouse at the Southwest Florida rearing facility in Immokalee were taken to the field
and transplanted in October 2005 and August 2006 in
590
Establishment, spread and initial impacts of Gratiana boliviana (Chrysomelidae) on Solanum viarum in Florida
Figure 1.
Florida counties (dark) where Gratiana boliviana have been released during the period 20032007.
Lee County, Florida. This has proven to be an efficient
technique with less labor involved for insect field release. Beetle releases have been made both on private
and public lands. The G. boliviana demand by cattle
ranchers for field release each year exceeded the beetle
production by our team of collaborators. We plan to
increase the beetle production in 2007 by establishing
field insectaries at different locations in Florida. Cattle
ranchers interested in obtaining beetles for release in
their farms are being provided with adult beetles, and
we are conducting monthly or bimonthly post-release
evaluations on the dispersion of the beetles, on the extent of the feeding damage and changes in the beetle
population at selected release sites. The post-release
evaluations also include observations on possible nontarget effects on closely related plant species growing
in the release area and on the regeneration of native
plant species and/or improved pastures that have been
displaced by the S. viarum plants.
Post-release evaluations of
G. boliviana in Florida
Evaluation of the feeding effects of the beetles on S.
viarum plants (percent defoliation, fruit production)
and number of beetles on plants began in the summer
of 20032004 in Polk and Okeechobee counties, in
St. Lucie and Okeechobee counties and in Collier and
Hendry counties. Monitoring also was initiated at the
Eagle Creek Conservation area in Orange County by
K. Peterman (Environmental Scientist) and J. Medal.
For the post-release evaluation in Polk County, where
approximately 1000 beetles were released in August
591
XII International Symposium on Biological Control of Weeds
2003, 20 marked plants within 100 m of the initial release site have been thoroughly examined every 2 to
3 months since the summer of 2003. The estimated
(visual) defoliation increased on average from 46%
(December 2003) to 94% (December 2004), and it
was directly associated with the increase in number of
adults and immature beetles observed on the plants during the same period, except from August to December
2004 when the number of beetles decreased. In 2005,
the plant defoliation was high on average from 69% to
96%. At least half of the 20 marked plants were unable
to regrow after complete defoliation by the beetles in the
previous year and also due to the competition by other
plant species. The number of S. viarum fruits produced
per plant defoliated by the beetles has significantly decreased with none or very few small fruits compared
with the large number of fruit (4055) observed during the summer of 2003 at the time the beetles were
released. Most of the plants on the 4-ha release site
have been replaced by other plant species including
bahiagrass (Paspalum notatum Fluegg), Rubus sp.,
dayflower (Commelina diffusa Burm), Caesar weed
(Urena lobata L.), air-potato (Dioscorea bulbifera L.),
roadside flatsedge (Cyperus sphacelatus, Rottb.), oak
(Quercus sp.) and other herbaceous vegetation. The estimated S. viarum density at the release area (4 ha) at
the end of November 2005 was only 510%, which is
significantly lower than the initial population density
(80% to 90%) that was observed in the summer 2003
before the beetles were released. The relatively low
number (<100) of beetles recorded on the 20 marked S.
viarum plants on each monitoring date in 2005 can be
attributed to beetle dispersal to S. viarum plants as far
as 1600 m away (September 2005) from the initial release site. Dispersal of the beetles was associated with
the low availability of foliage on the S. viarum plants at
the release site caused by extensive beetle defoliation
during the previous growing season. Dispersal ability
of the beetles at five of the Florida release sites ranged
from 1.6 to 16 km/year. After 3 years post-release at the
Polk County site, beetle defoliation is having a great
impact. Fruit production has declined to one to five deformed or no fruits per plant if the beetles start feeding
on the plants before fruit formation. Follow-up studies include observations on possible non-target effects
on closely related plant species growing in the release
area. To date, no non-target effects (J. Medal, personal
communication) have been observed even on plants in
the same genus such as the non-natives red soda apple
(Solanum capsicoides All.), wetland nightshade (Solanum tampicense Dunal) and turkey berry (Solanum torvum Sw.) that are growing intermixed with or in close
proximity to S. viarum.
Conclusion
Post-release evaluations of the South-American leafbeetle G. boliviana, first biological control agent whose
releases in Florida began in summer 2003 against the
invasive non-native spiny shrub S. viarum, have indicated an extensive weed defoliation and reduction of
fruit production in five of the release sites monitored.
The beetle established at almost all the release sites and
is spreading to adjacent weed-infested areas. Field observations also confirmed the specificity of the beetle
on the target weed, and to date, no non-target effects
have been observed even on plants closely related.
Acknowledgments
We thank Zundir Buzzi (Universidade Federal do
Paran, Curitiba, Brazil) for the identification of G.
boliviana. We thank Howard Frank (University of
Florida) for reviewing the manuscript. This research
is being funded by the United Sates Department of
Agriculture-Animal Plant Health Inspection Services
and by the Florida Department of Agriculture and
Consumer Services, Division of Plant Industry.
References
Akanda, R.A., Mullahey, J.J. and Shilling, D.G. (1996)
Growth and reproduction of tropical soda apple (Solanum
viarum Dunal) in Florida. In: Mullahey, J (ed) Proceedings of Tropical Soda Apple Symposium. University of
Florida-IFAS. Bartow, FL, USA, pp. 1522.
Akanda, R.A., Mullahey, J.J. and Shilling, D.G. (1997) Tropical soda apple (Solanum viarum) and bahiagrass (Paspalum notatum) response to selected PPI, PRE, and POST
herbicides. In: Abstracts of the Weed Science Society of
America Meeting, vol. 37. WSSA Abstracts, Orlando, FL,
USA, p.35.
Brown, W.F., Mullahey, J.J. and Akanda, R.A. (1996) Survivability of tropical soda apple seed in the gastro-intestinal
tract of cattle. In: Mullahey, J. (ed) Proceedings of Tropical Soda Apple Symposium. University of Florida-IFAS.
Bartow, FL, USA, pp. 3539.
Bryson, C.T. and Byrd Jr., J.D. (1996) Tropical soda apple in
Mississippi. In: Mullahey, J. (ed) Proceedings of Tropical Soda Apple Symposium. University of Florida-IFAS.
Bartow, FL, USA, pp. 5560.
Bryson, C.T., Byrd Jr., J.D. and Westbrooks., R.G. (1995)
Tropical soda apple (Solanum viarum Dunal) in the
United States. Mississippi Department of Agriculture and
Commerce-Bureau of Plant Industry Circular, USA, 2 pp.
Chandra, V. and Srivastava, S.N. (1978) Solanum viarum Dunal syn. Solanum khasianum Clarke, a crop for production
of Solasadine. Indian Drugs 16, 5360.
Coile, N.C. (1993) Tropical soda apple, Solanum viarum Dunal: the plant from hell. Botany Circular No. 27. Florida
Dept. Agric. and Consumer Services, Division of Plant
Industry, Gainesville, FL, USA.
Dowler, C.C. (1996) Some potential management approaches
to tropical soda apple in Georgia. In: Mullahey, J. (ed)
Proceedings of Tropical Soda Apple Symposium, Bartow,
Florida. University of Florida-IFAS, Bartow, FL, USA,
pp. 4154.
Gandolfo, D., Sudbrink, D. and Medal, J. (1999) Biology and
host specificity of the tortoise beetle Gratiana boliviana,
592
Establishment, spread and initial impacts of Gratiana boliviana (Chrysomelidae) on Solanum viarum in Florida
a candidate for biocontrol of tropical soda apple (Solanum
viarum), In: Spencer, N. (ed) Program Abstract, Xth International Symposium on Biological Control of Weeds.
USDA-ARS/Montana State University, Bozeman, MT,
USA, p. 130.
Gandolfo, D., McKay, F., Medal, J.C., and Cuda, J.P. (2007)
Open-field host specificity test of Gratiana boliviana
(Chrysomelidae), a biocontrol agent of Tropical soda apple in the USA. Florida Entomologist 90, 223228.
Habeck, D.H., Medal, J.C. and Cuda, J.P. (1996) Biological
control of tropical soda. In: Mullahey, J. (ed) Proceedings
of Tropical Soda Apple Symposium.University of FloridaIFAS, Bartow, FL, USA, pp. 6971.
McGovern, R.J., Polston, J.E., Danyluk, G.M., Heibert, E.,
Abouzid, A.M. and Stansly, P.A. (1994a) Identification
of a natural weed host of tomato mottle geminivirus in
Florida. Plant Disease 78, 11021106.
McGovern, R.J. Polston, J.E. and Mullahey, J.J. (1994b) Solanum viarum: weed reservoir of plant viruses in Florida.
International Journal of Pest Management 40, 270
273.
McGovern, R.J., Polston, J.E. and Mullahey, J.J. (1996) Tropical soda apple (Solanum viarum Dunal): host of tomato,
pepper, and tobacco viruses in Florida. In: Mullahey, J.
(ed) Proceedings of Tropical Soda Apple Symposium. University of Florida-IFAS, Bartow, FL, USA, pp. 3134.
Medal, J. (2005) A super beetle fighting the plant from hell:
tropical soda apple. The Florida Cattleman and Livestock
Journal 69(8), 4041.
Medal, J.C., Charudattan, R., Mullahey, J.J. and Pitelli, R.A.
(1996) An exploratory insect survey of tropical soda apple, Solanum viarum in Brazil and Paraguay. Florida Entomologist 79, 7073.
Medal, J.C., Pitelli, R.A., Santana,A., Gandolfo, D., Gravena, R.
and Habeck, D.H. (1999) Host specificity of Metriona
elatior Klug (Coleoptera: Chrysomelidae) a potential
biological control agent of tropical soda apple, Solanum
viarum) in the USA. BioControl 44, 432436.
Medal, J.C., Sudbrink, D., Gandolfo, D., Ohashi, D. and
Cuda, J.P. (2002) Gratiana boliviana, a potential biocontrol agent of Solanum viarum: quarantine host-specificity
testing in Florida and field surveys in South America. BioControl 47, 445461.
Medal, J.C., Gandolfo, D. and Cuda, J.P. (2003) Biology of
Gratiana boliviana, the first biocontrol agent released
to control tropical soda apple in the USA. University of
Florida-IFAS Extension Circular ENY, USA, 3 pp.
Medal, J., Ohashi, D., Gandolfo, D., McKay, F. and Cuda, J.
(2004) Risk assessment of Gratiana boliviana (Chrysomelidae), a potential biocontrol agent of tropical soda
apple, Solanum viarum (Solanaceae) in the USA. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M.,
Morin, L. and Scott, J.K. (eds) Proceedings of the XI In-
ternational Symposium on Biological Control of Weeds.
CSIRO Entomology, Canberra, Australia, pp. 292296.
Medal, J., Overholt, W., Stansly, P., Osborne, L., Roda, A.,
Chong, J., Gaskalla, R., Burns, E., Hibbard, K., Sellers, B.,
Gioeli, K., Munyan, S., Gandolfo, D., Hight, S. and Cuda,
J.P. (2006) Classical Biological Control of Tropical Soda
Apple in the USA. University of Florida-IFAS Extension
Circular IN-457, USA, 7 pp.
Mislevy, P., Mullahey, J.J. and Colvin, D.L. (1996) Management practices for tropical soda apple control: Update.
In: Mullahey, J. (ed) Proceedings of Tropical Soda Apple
Symposium. University of Florida-IFAS, Bartow, FL,
USA, pp. 6167.
Mislevy, P., Mullahey, J.J. and Martin, F.G. (1997) Tropical
soda apple (Solanum viarum) control as influenced by
clipping frequency and herbicide rate. In: Abstracts of the
Weed Science Society of America Meeting, vol. 37. WSSA
Abstracts, Orlando, FL, USA, p. 30.
Mullahey, J.J. and Colvin, D.L. (1993) Tropical soda apple:
a new noxious weed in Florida. University of Florida,
Florida Cooperative Extension Service, Fact Sheet WRS7, 3 pp.
Mullahey, J.J., Nee, M., Wunderlin, R.P. and Delaney, K.R.
(1993) Tropical soda apple (Solanum viarum): a new
weed threat in subtropical regions. Weed Technology 7,
783786.
Mullahey, J.J., Mislevy, P., Brown, W.F. and Kline, W.N.
(1996) Tropical soda apple, an exotic weed threatening
agriculture and natural systems. Dow Elanco. Down to
Earth 51(1), 18.
Mullahey, J.J., Akanda, R.A. and Sherrod, B. (1997) Tropical
soda apple (Solanum viarum) update from Florida. In: Abstracts of Weed Science Society of America Meeting, vol.
37. WSSA Abstracts, Orlando, FL, USA, p. 35.
Nee, M. (1991) Synopsis of Solanum section Acanthophora:
a group of interest for glyco-alkaloides. In: Hawkes, J.G.,
Lester, R.N., Nee, M., Estrada, N. (eds) Solanaceae III:
Taxonomy, Chemistry, Evolution. Royal Botanic Gardens
Kew, Richmond, Surrey, UK, pp. 258266.
Patterson, D.T. (1996) Effects of temperature and photoperiod
on tropical soda apple (Solanum viarum Dunal) and its
potential range in the United States. In: Mullahey, J. (ed)
Proceedings of Tropical Soda Apple Symposium. University of Florida-IFAS. Bartow, FL, USA, pp. 2930.
Pereira, A., Pitelli, R.A., Nemoto, L.R., Mullahey, J.J. and
Charudattan, R. (1997) Seed production by tropical soda
apple (Solanum viarum Dunal) in Brazil. In: Abstracts of
the Weed Science Society of America Meeting, vol. 37.
WSSA Abstracts. Orlando, Florida, USA, p. 29.
Sturgis, A.K. and Colvin, D.L. (1996) Controlling tropical
soda apple in pastures. In: Mullahey, J. (ed) Proceedings
of Tropical Soda Apple Symposium.University of FloridaIFAS. Bartow, FL, USA, p. 79.
593
Dissemination and impacts of the fungal
pathogen, Colletotrichum gloeosporioides
f. sp. miconiae, on the invasive alien tree,
Miconia calvescens, in Tahiti (South Pacific)
J.-Y. Meyer,1 R. Taputuarai2 and E. Killgore3
Summary
Long-term monitoring of biological control agents in their areas of introduction is essential to assess
their effectiveness. There is a need to monitor and evaluate agent dispersal and impacts so that the
degree of success can be quantified or reasons for failure can be clearly understood. A fungal pathogen, Colletotrichum gloeosporioides f. sp. miconiae Killgore & L. Sugiyama (Melanconiales, Coelomycetes, Deuteromycetinae), found in Brazil in 1997 was released in the tropical oceanic island of
Tahiti (Society Islands, French Polynesia, South Pacific) to control miconia, Miconia calvescens DC
(Melastomataceae), a small tree native to Tropical America, which has invaded native rainforests. The
plant pathogen, proven to be highly specific to miconia, causes leaf spots, defoliation and eventually
death of young seedlings in laboratory conditions. Two permanent plots in Tahiti (Taravao Plateau
and Lake Vaihiria) were monitored for a period of 6 years to assess the pathogens dispersal and
impacts on miconia in the wild. Leaf spots were observed approximately 30 days after inoculation.
Percentage of infected plants reached 100% after 3 months, and between 90% and 99% of leaves were
infected. Subsequent re-infection occurred after 3 months at Vaihiria and 18 months at Taravao. Mortality rate for monitored plants was 15% and reached 30% for seedlings less than 50 cm tall. Within
3 years, the fungus had disseminated throughout the island of Tahiti and had infected nearly all the
miconia plants up to 1400 m in montane rainforests. It was also found on the neighbouring island of
Moorea without any intentional release there. Leaf damage on miconia canopy trees increased from
4% to 34% with elevation in permanent plots set up between 600 and 1020 m. Our study showed
that rainfall and temperature were two limiting environmental factors that affected fungal spread and
disease development. Although this plant pathogenic agent is successfully established, has spread efficiently and has caused significant impacts on seedlings, additional biocontrol agents are still needed
to fully control the massive invasion by miconia in the Society Islands.
Keywords: biological control, island, monitoring; rainforest.
Introduction
The greatest weakness of biological control has been
the failure to quantify success adequately and to monitor
programs effectively (Myers and Bazely, 2003: 171).
Government of French Polynesia, Dlgation la Recherche, B.P.
20981 Papeete, Tahiti, French Polynesia.
2
Institut Louis Malard, B.P. 30 Papeete, Tahiti, French Polynesia.
3
Hawaii Department of Agriculture, Plant Pathology Quarantine Facility, 1428 South King Street, Honolulu, HI, USA.
Corresponding author: J.-Y. Meyer <jean-yves.meyer@recherche.
gov.pf>.
CAB International 2008
1
Long-term monitoring of biological control agents in
areas of introduction is essential to evaluate their effectiveness, i.e. their establishment (or acclimatization), their
reproduction (or replication), their dissemination and their
impacts on the target invasive alien species. Success of
biocontrol programmes against weeds in agro-ecosystems
or invasive plants in natural ecosystems is considered to
be achieved if there is a notable reduction in population
density through a decrease in the plant vigour, growth rate,
ability to reproduce or to germinate (Briese, 2000; Myers
and Bazely, 2003). It is also important to verify that the released natural enemy, although proven to be host-specific
in laboratory conditions, does not affect other non-target
species in natural conditions (Barton, 2004).
594
Dissemination and impacts of the fungal pathogen, Colletotrichum gloeosporioides f. sp. miconiae
This paper summarizes the results of monitoring a
fungal pathogen introduced in a tropical island ecosystem to control a dominant invasive tree in native
rainforests. We documented the establishment and dissemination of the plant pathogenic agent, quantified its
direct impacts on the target species in the wild and verified the absence of damages on non-target plant species
over a 6-year period. Factors suspected to have influenced the dynamics and impacts of the plant pathogen
are discussed, and its general efficiency is assessed.
Methods and materials
Miconia, the green cancer of Pacific
islands native forests
Miconia, Miconia calvescens DC (Melastomataceae), represents one of the most dramatic and devastating cases of a documented plant invasion into island
ecosystems. A small tree 412 m tall (up to 16 m in its
native range), miconia is native to tropical rainforests
of Central and South America where it is an understory
species in dense forest and a colonizer of small forest gaps (Meyer, 1994). It was introduced to the tropical oceanic island of Tahiti (French Polynesia, South
Pacific Ocean) in 1937 as a garden ornamental plant
because of its large, striking leaves with purple undersides (under the horticultural name Miconia magnifica
Triana). Rapid vegetative growth (up to 1.5 m per
year), early sexual maturity (reached in four to five
years), self-pollination and independence from specific pollinators, three flowering and fruiting peaks per
year, prolific fruit and seed production (millions of tiny
seeds per tree), active dispersal of the small fleshy berries by frugivorous native and alien birds over longdistances, high rate of seed germination (90% in 1520
days in laboratory conditions) even under very low
light conditions, large size and persistence of the soil
seed bank (up to 10,000 seeds/m2 and longevity more
than 15 years) and shade-tolerance make this species a
particularly aggressive colonizer in undisturbed native
forests and a competitor with native and endemic insular species (Meyer, 1994, 1996, 1998). In less than 50
years, miconia has successfully invaded all the native
mesic and wet forests of Tahiti (rainfall >2000 mm/
year) from sea level to 1400 m elevation and covers
approximately 70% of the island. Between 40 and 50
species of the 100 plants endemic to Tahiti are believed
to be on the verge of extinction due to the invasion of
miconia (Meyer and Florence, 1996). Miconia is also
invasive in the rainforests of Hawaii (Medeiros et al.,
1997), New Caledonia and north Queensland in Australia, and remains a potential threat for many other wet
tropical Pacific islands.
Because conventional manual and chemical control
methods have shown their limits in heavily invaded
islands such as Tahiti and Moorea (more than 80,000
and 3500 ha, respectively), where miconia is form-
ing dense monospecific stands on steep, mountainous
slopes, biological control is viewed as the only effective alternative.
Cgm, a host-specific fungal plant pathogen
Colletotrichum (Order Melanconiales, Class Coelomycetes, Subdivision Deuteromycetinae) is one of the
most important genera of plant pathogenic fungi worldwide, particularly in subtropical and tropical regions
(Prusky et al., 2000). There are several form species
(or strains) of Colletotrichum gloeosporioides (Penz.)
Penz. & Sacc. which are very specific to their plant
hosts, including the well-known C. gloeosporioides
forma specialis aeschynomene, which is used as a mycoherbicide registered as Collego for control of the
invasive legume, Aeschynomene virginica (L.) Britton,
Sterns & Poggenb. (Fabaceae), in soybeans and rice in
several Mississippi River delta states, and C. gloeosporioides (Penz) Sacc. f. sp. jussiaeae successfully used
as a biological agent for control of winged water primrose, Jussiaea decurrens DC (Onagraceae) on rice fields
of Arkansas (Templeton, 1982; TeBeest, 1991). In the
Pacific Islands, C. gloeosporioides (Penz) Sacc. f. sp.
clidemiae, found in Panama, was released in Hawaii in
1986 to control the invasive shrub, Clidemia hirta D.
Don (Melastomataceae) (Trujillo et al., 1986; Trujillo,
2005).
C. gloeosporioides (Penz) Sacc. f. sp. miconiae
Killgore & L. Sugiyama (hereafter, Cgm) was discovered in the State of Minas Gerais in Brazil in 1997 and
isolated by Dr Robert Barreto (Universidad de Viosa).
It reproduces by asexual spores or conidia, 14.717.5
m in length and 5.06.25 m in width, which are produced in acervuli that arise on the abaxial surface leaf
(Killgore et al., 1999). Conidia of Colletotrichum fungi
are produced under high moisture conditions and are
disseminated by wind-driven rain. In the laboratory,
test plants were inoculated using a spore concentration of 1 105 conidia millilitre in sterile water and
incubated for 48 h in an enclosed chamber at 100%
relative humidity (Killgore, 2002). The Cgm causes
foliar anthracnose and necrosis, which lead to premature defoliation. When the pathogen is inoculated onto
injured stems, cankers develop causing a dieback of the
branch. Under aseptic laboratory conditions, the fungal
pathogen attacked germinating miconia seeds and also
killed emergent seedlings (Killgore, 2002). Mortality rate of very young seedlings (1 to 1.5 months old,
less than 5 mm tall and with two leaves) in laboratory
conditions (at room temperature between 24 and 30C,
hygrometry between 40 and 70%, 12 h light and 12 h
darkness lightning regime) was 74% only 1 month after
inoculation (Table 1).
Host-specificity tests were conducted at the quarantine facilities of the Hawaii Department of Agriculture in Honolulu following Wapsheres phylogenetic
centrifugal method (Wapshere, 1974) on 28 plant spe-
595
XII International Symposium on Biological Control of Weeds
Table 1.
ortality rate of young Miconia calvescens seedlings (1 to 1.5 months old) infected by Colletotrichum gloeospoM
rioides f. sp. miconiae in laboratory conditions.
Treatment
Number of dead seedlings 1 week after inoculation date (%)
Number of dead seedlings 2 weeks after inoculation date (%)
Number of dead seedlings 4 weeks after inoculation date (%)
cies in the botanical order Myrtales (species from the
families Combretaceae, Lythraceae, Melastomataceae,
Myrtaceae and Thymelaeaceae), including native and
endemic Tahitian Melastomataceae (Astronidium spp.
and Melastoma denticulatum Bonpl. ex Naudin). Repeated tests showed that the Cgm was highly specific
to miconia (Killgore et al., 1999). It was released in
the Hawaiian Islands in June 1997 after approval by
the US Department of Agriculture, the US Department
of Interior and the Hawaii State Department of Agriculture and in Tahiti with the approval of the French
Polynesian Government. The two release sites in Tahiti
were the Taravao Plateau on the peninsula of Tahiti Iti
in April 2000 and near Lake Vaihiria at the centre of
Tahiti Nui in September 2002. Both sites are located at
approximately 600 m elevation with an annual rainfall
of 3300 and 7000 mm/year, respectively.
Inoculated (N = 163)
Control (N = 166)
99 (61%)
115 (71%)
120 (74%)
1 (0.6%)
6 (4%)
6 (4%)
Astronidium, an endemic shrub of the same family as
miconia, were examined.
A total of 20 transects, 10 m long, were set up in
2002 and 2003 on different parts of the island of Tahiti
(leeward side, windward side, peninsula of Tahiti Iti)
between 70 and 1100 m elevation to evaluate Cgm dispersal across the island. Observations on isolated plants
were also made at higher elevation (up to 1400 m) and
in other mountainous areas above 1000 m elevation in
the centre of Tahiti during field surveys in 2003 (Fig. 1).
Maximum leaf damages on mature, reproductive miconia trees (called canopy leaves) was measured in
2005 and 2006 on 28 cut trees (three to eight trees per
site for a total of 1571 leaves) in eight other permanent
plots located in different sites on the island of Tahiti
between 600 and 1020 m elevation.
Results
Monitoring impact and dissemination
Monitoring was conducted in two permanent plots
of approximately 100 m2 set up in the two release sites
in Tahiti for a 6-year period (20002006). A total of
110 seedlings and juvenile miconia plants between 20
cm and 1.5 m tall on Taravao and between 10 cm and
2.8 m tall on Vaihiria were tagged per plot. It was almost impossible to study smaller seedlings and larger
plants for technical reasons (size of the tags and access
to higher branches and leaves). Miconia plants were
inoculated by spraying a solution of Cgm spores (concentration of 1 108 conidia millilitre in sterile water),
which were mass-cultured in laboratory conditions (on
10% V8 juice agar and under constant fluorescent illumination at 2024C) at the Institut Louis Malard
in Tahiti from an inoculum sent by the Hawaii Department of Agriculture.
Every week for the first 3 months after the release,
then every 6 months during the monitoring period, we
recorded the number of dead plants per plot and the
percentage of infected leaves (with one or more foliar
lesions) for each plant; we counted the number of leaf
spots on the most infected leaf for each plant and visually estimated the percentage of leaf damages (i.e.
leaf area loss due to necrosis) on the most infected leaf
for each plant (called maximum damages). A visual
check for non-target damage was done at the same
time, in and around the two permanent plots for disease
symptoms. In particular, species belonging to the genus
Establishment and reproduction
Typical symptoms of Cgm infection appeared between 21 (Taravao) and 33 days (Vaihiria) after the initial inoculation. Subsequent re-infection by the fungus
on miconia plants located outside the release plot (i.e.
fungal reproduction and dispersal) varied between 3
(Vaihiria) and 18 months (Taravao) after inoculation.
The difference in time to onset of re-infection between
the two sites (located at the same elevation) may be explained by a much lower mean annual rainfall in Taravao (3300 mm/year) compared to Vaihiria (7000 mm/
year) and by a serious drought period, which occurred
in Tahiti when the pathogen was released on Taravao in
2000 (rainfall of 144 and 216 mm during the months
of September and November 2000, respectively, vs a
mean annual rainfall of 212 and 355 mm in September
and November during the last 10 years. Data received
from Mto-France in French Polynesia, personal communication).
Dissemination
Within 3 years (2003), the fungus was detected 15
km from the release sites and spread throughout the island of Tahiti. It has infected almost all the miconia
trees, juvenile plants and seedlings between sea level
and 1400 m elevation. Miconia plants found on the
neighbouring island of Moorea (located approximately
596
Dissemination and impacts of the fungal pathogen, Colletotrichum gloeosporioides f. sp. miconiae
Figure 1.
Map of Miconia calvescens invasion in Tahiti and the evolution of Colletotrichum gloeosporioides f. sp. miconiae
dissemination from the two release sites.
20 km from Tahiti) were also infected by Cgm in 2003
without being intentionally introduced (Fig. 1). Spores
may have been carried there by the wind or on contaminated clothing or equipment.
Impacts
Three months after inoculation, 100% of the tagged
plants and between 90% and 99% of their leaves were
infected by Cgm (i.e. presence of leaf spots) in the two
permanent plots (Figs. 2 and 3). The mean number of
leaf spots on the most infected leaf increased from 10
(1 month after the inoculation) to 90 (6 years after the
inoculation) in Taravao, and from 10 (1 month after
the inoculation) to 55 (4 years after the inoculation) in
Vaihiria. In 2006, the mean mortality rate was approximately 15% for the monitored plants (14.1% in Taravao and 16.7% in Vaihiria), reaching approximately
120
100
80
Mean % 60
40
20
0
Apr-00 May-00 Jun-00 Aug-00 Sep-00 Nov-00 Jun-02 Jan-03 May-03 Aug-03 Dec-04 Oct-05 Aug-06
Dates
Dead plants
Figure 2.
Infected plants
Infected leaves
Leaf damages
Evolution of Colletotrichum gloeosporioides f. sp. miconiae impacts on monitored Miconia calvescens plants in
the Taravao release site (2000).
597
XII International Symposium on Biological Control of Weeds
120
100
80
Mean % 60
40
20
0
Sep-02
Oct-02
Dec-02
Apr-03
Aug-03
Dec-04
May-06
Dates
Dead plants
Figure 3.
Infected plants
Infected leaves
Leaf damages
Evolution of Colletotrichum gloeosporioides f. sp. miconiae impacts on monitored Miconia calvescens plants in
the Vaihiria release site (2002).
30% for small seedling less than 50 cm tall (28.6% in
Taravao and 32.3% in Vaihiria). Multiple damages on
the surviving tagged miconia plants, including rotting
stems and deformed leaves, ranged between 14% (Taravao) and 47% (Vaihiria). Mean maximum leaf damages
on the tagged miconia plants in the two monitored sites
was approximately 25% (27.4% in Taravao and 21.3%
in Vaihiria), and maximum damages on miconia canopy leaves in eight other permanent study sites (100 m2
quadrats) varied between 4% (at 600 m elevation) and
34% (at 970 m elevation; Fig. 4). None of the nontarget alien and native plants, located within or outside
the permanent plots, displayed any symptoms of Cgm
infection after 6 years of monitoring.
40
Canopy leaves maximum damages (%)
35
30
25
20
15
10
5
0
400
500
600
700
800
900
1000
1100
1200
Elevation (m)
Figure 4.
Mean percentage of Miconia calvescens canopy leaves maximum damages caused by Colletotrichum
gloeosporioides f. sp. miconiae according to elevation in eight different plots set up in Tahiti in 20052006
(N = 1571 leaves).
598
Dissemination and impacts of the fungal pathogen, Colletotrichum gloeosporioides f. sp. miconiae
Discussion
Only a very few biocontrol programmes using plant
pathogens with the goal of preserving tropical island
native forest ecosystems are documented (Gardner,
1992; Smith et al., 2002; Trujillo, 2005) compared to
their common use in agriculture ecosystems (see, e.g.
Templeton, 1982). In natural areas, the need to carefully monitor and evaluate agent establishment, dissemination and impact is particularly important, so that
the degree of success can be quantified or reasons for
failure can be understood (Briese, 2000).
The Cgm, a fungal pathogen proven to be highlyspecific to miconia in laboratory conditions, was successfully introduced to the tropical oceanic island of
Tahiti in 2000 (Meyer and Killgore, 2000). Our results
show that in 3 years, it has established, reproduced and
spread throughout the island of Tahiti and even to the
neighbouring island of Moorea, located approximately
20 km away, without any intentional release there. The
Cgm has succeeded in infecting almost all the miconia plants on both islands without causing any apparent harm to non-target plant species. It was capable of
distributing itself by natural means, particularly at high
elevation in montane rainforests or cloudforests (up to
1400 m elevation), and to infect both dense monospecific stands and isolated miconia plants. Several biocontrol
programmes elsewhere in the world have been considered unsuccessful (or only partially successful) because
of the large habitat range of the target species but the
narrower ecological range of their natural enemies (e.g.
Lantana camara L. in Hawaii; Broughton, 2000).
Leaf damage caused by Cgm is more severe in highelevation areas of Tahiti (Moorea and Raiatea, unpublished observation) where cooler and wetter conditions
prevail, suggesting that temperature and moisture (as
humidity or free water) are important factors for disease development. The reproduction and dissemination
of the pathogen was delayed at the Taravao site due to
a drought period, which occurred when the pathogen
was released in 2000. The same pattern was observed
in Hawaii after the release of the Cgm in 1997. The
importance of air temperature was demonstrated for
other C. gloeosporioides with an optimum temperature
for many form species at about 28C. Disease development was severely limited at 36C (TeBeest, 1991).
Defoliation of C. hirta caused by C. gloeosporioides
f. sp. clidemiae over contiguous areas only occurs
when weather conditions are favourable, i.e. cool (16
24C), windy and rainy (Conant, 2002; Trujillo, 2005;
R. Hauff, personal communication).
Mortality rate was high for very young miconia
seedlings (74% 1 month after inoculation) in laboratory
conditions but was relatively low and slow for larger
miconia plants in the field (15% for plants between 10
cm and 2.80 m tall, up to 30% for seedlings less than
50 cm tall, 4 to 6 years after the release). Although the
plant pathogen may slow the growth of established miconia plants (between 17% and 35 % of the surviving
miconia plants showed multiple damages, especially
rotten stems and curved leaves) and cause the dieback
of young seedlings, it alone will not control the massive
invasion of miconia. Partial defoliation of reproductive
canopy trees (up to 35%) favoured the recruitment of
native plants, including rare threatened endemic plants,
by enhancing the light availability in the understory
(Meyer et al., 2007).
The Cgm was released on the island of Raiatea
(Society Islands) in 2004, where manual and chemical control operations have been conducted since 1992
on a 450-ha infested area, and in Nuku Hiva (Marquesas Islands) in 2007, where small miconia populations
(<1 ha) were recently discovered. On these islands,
manual and chemical control will be necessary to
achieve complete eradication, but the fungal pathogen
is expected to reduce the number of miconia plants, especially those at the seedling stage. Additional biological control agents are still much needed to fully control
the massive invasion by miconia in the Society Islands
and the Hawaiian Islands.
Acknowledgements
This research program was funded by the Contrat de
Dveloppement Etat-Pays (France-French Polynesia
Development Contract 20002006) and conducted in
collaboration with the Institut Louis Malard (Tahiti).
We deeply thank the previous and current chiefs of the
Dlgation la Recherche (Dr Isabelle Bella Perez,
Dr Edouard Suhas and Dr Priscille Tea Frogier), and
ministers of research (Dr Patrick Howell, Prof Louise
Peltzer, Dr Jean-Marius Raapoto, Dr Keitapu Maamaatuaiahutapu and Dr Tearii Alpha) for their support.
Mauruuru roa to our graduate students Valrie Tchung,
Anne Duplouy, Sylvain Martinez, Marie Fourdrigniez
and Van-Mai Lormeau for their precious help on the
field. We also thank Robert Hauff (Division of Forestry
and Wildlife, Division of Land and Natural Resources,
Honolulu) for his personal communication on C. hirta
biocontrol monitoring in Hawaii, Dr Patrick Conant
(Hawaii Department of Agriculture, Hilo, HI) for his
helpful comments and two anonymous reviewers for
improving the manuscript.
References
Barton, J. (2004) How good are we at predicting the field
host-range of fungal pathogens used for classical biological control of weeds? Biological Control 31, 99122.
Briese, D. T. (2000) Classical biological control. In: Sindel,
S.M. (ed) Australian Weed Management Systems. R.G.
and F.J. Richardson, Melbourne, Australia, pp. 161186.
Broughton, S. (2000) Review and evaluation of lantana biocontrol programs. Biological Control 17, 272286.
599
XII International Symposium on Biological Control of Weeds
Conant, P. (2002) Classical biological control of Clidemia
hirta (Melastomataceae) in Hawaii using multiple strategies. In: Smith, C.W., Denslow, J. and Hight, S. (eds) Proceedings of a Workshop on Biological Control of Invasive
Plants in Native Hawaiian Ecosystems. Technical Report
129. Pacific Cooperative Studies Unit, University of Hawaii at Manoa, Honolulu, USA, pp 1320.
Gardner, D.E. (1992) Plant pathogens as biocontrol agents in
native Hawaiian ecosystems. In: Stone, C.P., Smith, C.W.
and Tunison, J.T. (eds) Alien Plant Invasions in Native
Ecosystems of Hawaii: Management and Research. University of Hawaii Press, Honolulu, USA, pp. 432451.
Killgore, E.M., Sugiyama, L.S., Baretto, R.W. and Gardner,
D.E. (1999) Evaluation of Colletotrichum gloeosporioides for biological control of Miconia calvescens in Hawaii. Plant Disease 83, 964.
Killgore, E.M. (2002) Biological control potential of Miconia calvescens using three fungal pathogens. In: Smith,
C.W., Denslow, J. and Hight, S. (eds) Proceedings of a
Workshop on Biological Control of Invasive Plants in Native Hawaiian Ecosystems. Technical Report 129. Pacific
Cooperative Studies Unit, University of Hawaii at Manoa,
Honolulu, USA, pp 4552.
Medeiros, A.C., Loope, L.L., Conant, P. and McElvaney, S.
(1997) Status, ecology and management of the invasive
plant Miconia calvescens (Melastomataceae) in the Hawaiian Islands. Bishop Museum Occasional Paper 48,
2336.
Meyer, J.-Y. (1994) Mcanismes dinvasion de Miconia calvescens DC. en Polynsie franaise. PhD Dissertation.
Universit des Sciences et Technique du Languedoc,
Montpellier, 126 pp.
Meyer, J.-Y. (1996) Status of Miconia calvescens (Melastomataceae), a dominant invasive tree in the Society Islands
(French Polynesia). Pacific Science 50, 6676.
Meyer, J.-Y. (1998) Observations on the reproductive biology of Miconia calvescens DC (Melastomataceae), an invasive tree on the island of Tahiti (South Pacific Ocean).
Biotropica 30, 609624.
Meyer, J.-Y. and Florence, J. (1996) Tahitis native flora
endangered by the invasion of Miconia calvescens DC.
(Melastomataceae). Journal of Biogeography 23, 775
781.
Meyer, J.-Y. and Killgore, E. (2000) First and successful release of a bio-control pathogen agent to combat the invasive alien tree Miconia calvescens (Melastomataceae)
in Tahiti. Invasive Species Specialist Group of the IUCN
Aliens 12, 8.
Meyer, J.-Y., Duplouy, A. and Taputuarai, R. (2007) Dynamique des populations de larbre endmique Myrsine longifolia (Myrsinaces) dans les forts de Tahiti
(Polynsie franaise) envahies par Miconia calvescens
(Melastomataces) aprs introduction dun champignon
pathogne de lutte biologique: premires investigations.
Revue dEcologie (Terre Vie) 62, 1733.
Myers, J.H. and Bazely, D. (2003) Ecology and Control of
Introduced Plants, Chapter 7. Biological Control of Introduced Plants. Cambridge University Press, Cambridge,
UK, pp 164194.
Prusky, D., Freeman, S. and Dickman, M.B. (2000) Colletotrichum: Host Specificity, Pathology, and HostPathogen
Interaction. The American Phytopathological Society, St.
Paul, MN, USA, 393 pp.
Smith, C.W., Denslow, J. and Hight, S. (2002) Proceedings
of a Workshop on Biological Control of Invasive Plants
in Native Hawaiian Ecosystems. Technical Report 129.
Pacific Cooperative Studies Unit, University of Hawaii at
Manoa, Honolulu, USA.
TeBeest, D.O. (1991) Ecology and epidemiology of fungal
plant pathogens studied as biological control agents of
weeds. In: TeBeest, D.O. (ed) Microbial Control of Weeds.
Chapman and Hall, New York, USA, pp. 97114.
Templeton, G. E. (1982) Status of weed control with plant
pathogens. In: Charudattan, R. and Walker, H.L. (eds)
Biological Control of Weeds with Plant Pathogens. Wiley,
New York, pp. 2944.
Trujillo, E.E. (2005) History and success of plant pathogens
for biological control of introduced weeds in Hawaii. Biological Control 33, 113122.
Trujillo, E.E., Latterell, F.M. and Rossi, A.E. (1986) Colletotrichum gloeosporioides, a possible biological control
agent for Clidemia hirta in Hawaiian forests. Plant Disease 70, 974976.
Wapshere, A.J. (1974) A strategy for evaluating the safety of
organisms for biological weed control. Annals of Applied
Biology 77, 200211.
600
One agent is usually sufficient for
successful biological control of weeds
J.H. Myers
Summary
A review of the recent literature reveals ten new cases of successful biological control of weeds. In
each of these a single species of biological control agent has apparently caused the decline in the
density and biomass of the target weed. Each of these agents also dramatically damages the plants and
reduces their survival. While it remains difficult to predict which species of insect herbivore or plant
pathogen will be successful, history suggests that those that cause serious damage to the plants are
most likely to reduce target weed biomass and density. Clearly single agents can be successful.
Keywords: biological control efficacy, natural enemies, invasive plants.
Introduction
Recent, highly publicized examples of non-target impacts in weed biological control programmes (Louda
et al., 2003) have resulted in more stringent regulations
and public apprehension about the introduction of new,
exotic biological control agents. In addition, interest in
the conservation of natural biodiversity calls attention
to the potential impacts of more introductions of exotic species such as those occurring in weed biological
control programmes. Given these concerns, in the future, it will be important for the practice of biological
control to be conservative in terms of the number of
agents introduced and to be efficient in regard to the
degree of control achieved with each new species introduced.
Although it is a commonly held view that biological control success requires the introduction of several
species of agents attacking different parts of a target
weed, we have shown in the past that a majority of
weed biological control successes has been attributed
to a single species of agent (Denoth et al., 2002). To
determine if the pattern in more recent weed biological control successes is in accordance with the earlier
analysis, I have reviewed the recent literature for successful weed control projects to determine if single or
University of British Columbia, Departments of Zoology and Agroecology, 6270 University Blvd., Vancouver, BC, Canada V6T 1Z4
Corresponding author: J.H. Myers <[email protected]>.
CAB International 2008
multiple agents were required to achieve biological
control success.
Methods
The definition of success that I have used in this evaluation is that the population density of the target weed
was greatly reduced by the activity of the introduced
agent. Even if the scale at which the agent had established was small and geographically limited, I considered this to be successful control if plant density
at sites with established agents declined significantly
as compared to control sites lacking the agent. Other
definitions of success might include factors such as the
geographical extent of the impact of the control agent
or other biological parameters such as the reduction of
seed production (review in Myers and Bazely, 2003).
The recent literature was searched using Biosis and
Science Citation Index and the key words biological
control success. I also reviewed the journal Biological
Control from 2003 to the present for examples of successful control programmes.
Results
Nine cases were found of biological control of weeds
programmes recorded in the recent literature, and we
have been studying an additional example that is successful (Table 1). In all these cases, success was attributed to a single agent, and in no recent cases were
multiple agents reported to be necessary or involved in
success. These examples are briefly described below.
601
XII International Symposium on Biological Control of Weeds
Table 1.
ummary of recently reported successful biological control of weed programs. Est. indicates the number of
S
agents that are established in the program. Successful agents are those to which success has been attributed in the
literature.
Weed
Country
Est.
Successful agents
Reference
Mimosa pigra
Asparagus
asparagoides
Linaria dalmatica
Australia
Australia
8
3
Carmenta mimosa
Puccinia myrsiphylii
Paynter (2005)
Morin and Edwards (2006)
CanadaBC, USA
Mecinus janthinus
Azolla filiculoides
Ageratina riparia
South Africa
New Zealand
Hawaii
Hawaii
1
2
3
1
Stenopelmus rufinasus
Entyloma ageratinae
De Clerck-Floate and Harris (2002),
Nowerski (2004), Hansen (2004)
McConnachie et al. (2004)
Barton et al. (2007)
Trujillo (2005)
Trujillo (2005)
Canada
Mogluones cruciger
Canada
Galerucella calmariensis
Montana
British Columbia,
Colorado, Montana
12
12
Cyphocleonus achates
Larinus minutus
Passiflora
tarminiana
Cynoglossum
officinale
Lythrum salicaria
Centaurea stroebe
Centaurea diffusa
Septoria passiflorae
Mimosa, Mimosa pigra, L.
M. pigra is a woody legume that creates impenetrable stands in the Northern Territory of Australia
(Paynter, 2005). Six biological control agents were introduced and established, but of these, only the twig
boring moth, Carmenta mimosa Echlin and Passoa, has
been found to reduce the density of M. pigra. Attack by
C. mimosa can cause severe defoliation and kill twigs
and branches, and this can reduce seed production by
90%. Reduced seed banks are associated with increased
under storey from competing plants, and this further
hinders seedling success. Higher densities of understorey plants also increase fuel loads and the susceptibility
of mimosa to fire. Mimosa populations did not expand
in any of the nine sites in which C. mimosa was established and three of the sites contracted in size. In
contrast, mimosa populations expanded in four of eight
areas that lacked C. mimosa. From this study (Paynter,
2005), it appears that C. mimosa has the potential to be
a successful biological control agent.
Bridal creeper, Asparagus
asparagoides, L. Druce,
Bridal creeper, A. asparagoides, is a South African
vine that has invaded natural areas with Mediterranean
climates in Australia. Initially, it was considered that
multiple agents would be required for biological control success, and three agents were introduced: a rust,
Puccinia myrsiphylli (Thm), a leafhopper in the genus
Zygina and a beetle in the genus Crioceris (Morin and
Edwards, 2006). Of these, the rust has become widely
established in coastal areas and has significantly reduced bridal creeper populations (Morin et al., 2006).
De Clerck-Floate and Schwarzlnder
(2002)
Lindgren et al. (2002), Denoth and
Myers (2005)
Story et al. (2006)
Myers (unpublished), Seastedt et al.
(2003), Smith (2004)
Although defoliation by the leafhopper also appears to
have the potential to reduce the population density of
bridal creeper, an egg parasite has reduced its success
and populations are unstable. It may have a greater impact in drier, interior sites where the conditions may
be less favourable for the fungus (Morin and Edwards,
2006). Presently, it appears that the fungus is able
to achieve successful biological control on its own in
coastal areas of Australia.
Dalmatian toadflax, Linaria dalmatica,
L. (Mill.)
Dalmatian toadflax, Linaria dalmatica, infests large
areas of rangeland in western North America. The stemboring weevil, Mecinus janthinus,Germar, has been
highly successful in reducing densities of Dalmatian
toadflax, particularly in British Columbia, Canada (De
Clerk-Floate and Harris, 2002; Hansen, 2004; Nowerski, 2004; McClay and Hughes, 2007). In some areas
such as Alberta, Canada, the climate is less favourable
for the weevils, and they do not have sufficient time to
develop to the over wintering adult stage (McClay and
Hughes, 2007). Thus, M. janthinus is an example of a
species that is a successful biological control agent on
its own in some regions, but it is not adapted to all of
the areas invaded by the target weed.
Red floating fern, Azolla filiculoides, Lam.
Red floating fern, A. filiculoides, is native to South
America but became a serious aquatic weed in South
Africa. The frond-eating weevil, Stenopelmus rufina
sus, Gyllenhall was introduced from Florida, USA,
602
One agent is usually sufficient for successful biological control of weeds
to South Africa in 1997 and caused extinction of the
water fern at 81% of the original 112 release sites usually within 7 months of its release (McConnachie et al.,
2004). This programme with a single species of agent
is considered to be highly successful.
has reduced the density of houndstongue at most of
the release sites (De Clerck-Floate and Schwarzlnder,
2002; De Clerck-Floate et al., 2005) and thus is considered to be a successful biological control agent.
Purple loosestrife, Lythrum salicaria L.
Mistflower, Ageratina riparia (Regel)
R.M. King & H. Rob
Mistflower, A. riparia, is an ornamental that is native to Central America. It has been spread to many
tropical areas where it reaches dense stands in warm,
moist habitats along forest edges, wetlands and poorly
managed pastures. In HawaiI, four species of biological control agent were introduced, and success was attributed to all of them, although no quantitative data
were collected (Trujillo, 1985). Two species of biological control agents were introduced to New Zealand: the
white fungus, Entyloma ageratinae Barreto and Evans,
in 1998, and a gall fly, Procecidochares alani, Steyskal,
in 2001. The fungus spread rapidly, and within several
years, mistflower cover declined from 81% to 1.5% at
some heavily infested sites, apparently before the gall
flies were widely established (Barton et al., 2007). The
success of E. ageratinae in reducing mistflower suggests that this species may be sufficient for successful
biological control, and the contribution of other agents
in Hawaii may have been overestimated.
Banana poka, Passiflora tarminiana,
Copens and Barney
Banana poka, P. tarminiana, is native to the high
Andes and became a major weed in high elevations of
Hawaii where it infested more that 50,000 ha (Trujillo,
2005). In 1993, a fungus, Septoria passiflorae Syd.,
was introduced to Hawaii from Columbia for hostrange testing under quarantine. After it was shown to
be specific, S. passiflorae was sprayed on banana poka
in the Hilo Forest Reserve. Densities of the weed were
reduced by 95% within 4 years in many areas but not
in regions in which the fungus was killed by acid rain
(Trujillo, 2005). This plant pathogen has preserved endangered species and allowed the regeneration of koa
forests at high elevations on the islands of Kauai, Maui
and Hawaii.
Houndstongue, Cynoglossum officinale L.
Houndstongue, C. officinale, is a native of Europe
and Asia Minor and a serious rangeland weed in British
Columbia, Canada. The seeds of this plant attach to the
faces and hides of cattle and the foliage is very toxic
to large mammals. The root-boring weevil, Mogluones
cruciger Herbst., was introduced in 1997 and became
effectively established at many sites where it may attack more than 90% of flowering plants and kill over
half the rosettes it attacks in the autumn. M. cruciger
Purple loosestrife, L. salicaria, is a European plant
that has become widely spread in wetland areas across
southern Canada and the northern USA. Two species
of leaf-feeding beetles have been introduced widely
as biological control agents, Galarucella calmarien
sis and Galarucella pusilla Dust., and in some locations, the root-boring beetle, Hylobius transversovit
tatus Goeze, has also been introduced. In southern
British Columbia, the species composition of releases
is not certain but is likely to have been dominated
by G. calmariensis (R. DeClerk-Floate, personal communication). This has effectively reduced loosestrife
at several sites that were monitored (Denoth and Myers, 2005). Similarly, in Manitoba, G. calmariensis
reduced loosestrife populations at several locations
(Lindgren, 2000). In Michigan, both species of beetles
were introduced, but only G. calmariensis became established and here too the species effectively reduced
loosestrife density at many locations (Landis et al.,
2003). Both Galarucella spp. were established in
western Oregon and were considered to be ecological
equivalents. They contributed to a dramatic decline in
loosestrife plant size and density over 3 years (Schooler,
1998). In New York, G. pusilla was the dominant species at many locations, but neither species together
or alone had an apparent impact on loosestrife density (Grevstad, 2005). After 5 years, G. calmariensis
was spreading and increasing in density compared to
G. pusilla. In another study, in New York, the two species of leaf-feeding beetles were introduced, and after
5 years, a significant decline in loosestrife densities occurred (Albright et al., 2004). The species composition
of the two beetle populations was not monitored over
time in this study so it is not clear if both species were
involved in the decline of host plant density. Comparisons of the impacts of H. transversovittatus and
G. calmariensis showed that the leaf-feeding beetle
had the greatest impact on loosestrife reproductive
effort and above-ground biomass, and this damage allowed improved growth of other plants in plots (HuntJoshi et al., 2004). In conclusion, reduction of purple
loosestrife density has been achieved in programs involving only G. calmariensis, and this shows that a
single agent can be successful. The failure of Grevstad
(2005) to find a reduction in loosestrife density in programs that initially involved both leaf-feeding beetle
species may suggest that there can be a negative interaction of the species or that the environment in New
York is less conducive to control by these leaf-feeding
beetles. The two species were successful in combination in western Oregon however.
603
XII International Symposium on Biological Control of Weeds
Spotted knapweed, Centaurea stoebe
Linnaeus micranthos (S. G. Gmelin
ex Gugler) Hayek
Spotted knapweed, C. stoebe micranthos, is a serious rangeland weed in western North America where it
infests more than 3 million ha. Twelve Eurasian insect
species have been introduced as potential biological
control agents. Seven of these species are considered
to have an impact through the reduction in seed production. This has not, however, been translated into
successful reduction of plant density. One species, the
root-boring beetle Cyphocleonus achates Fahraeus can
kill spotted knapweed plants and caused a significant
reduction in plant density, 99% and 77% over 11 years,
at two sites where it reached high densities (Story et al.,
2006). This weevil does not fly, so its spread is slow,
but nevertheless, the population did expand on average
by 99 m per year. Thus, this species is capable of at
least local control of spotted knapweed.
Diffuse knapweed, Centaurea
diffusa Lamarck
Diffuse knapweed, C. diffusa, like spotted knapweed, is a rangeland weed in western North America.
It infests drier areas at lower elevations than spotted
knapweed and shares some of the introduced biological control agents with spotted knapweed (Bourchier
et al., 2002). Twelve species of exotic, potential control agents have been introduced for this species as
Figure 1.
well, and the most widely established agents include
two gall flies, Urophora affinis Frauenfeld and U.
quadrifasciata, the root-boring beetle, Sphenoptera
jugoslavica Obenberger, and most recently the weevil,
Larinus minutus Gyllenhal. Since the establishment
and spread of L. minutus in the late 1990s, densities of
diffuse knapweed have declined in many sites in British Columbia, Canada and Colorado and Montana in
the United States (Seastedt et al., 2003; Smith, 2004;
J. Myers, unpublished data). While the other established
biological control agents reduce seed production of diffuse knapweed, L. minutus reduces both seed production and plant survival. Damage to flowering plants can
be severe particularly in dry summers. The impact of
L. minutus on diffuse knapweed in British Columbia
was demonstrated after a fire removed both knapweed
and biological control agents from a site. The seed bank
of knapweed was sufficient for a resurgence of dense
plant populations, and over 3 years, L. minutus returned
to once again suppress the knapweed (J. Myers, unpublished data). These observations suggest that L. minutus
is an effective control agent and sufficient to reduce the
density of diffuse knapweed.
Discussion
These examples of recent biological control successes
can be incorporated with the data previously summarized by Denoth et al. (2002) (Fig. 1). In that review,
based on Julien and Griffiths (1998) catalogue of biological control agents and weeds, diffuse knapweed
Number of agents considered to have contributed to the successful biological
control of weeds. In this figure, ten new successful projects are added to those
summarized by Denoth et al. (2002), and instead of attributing success to the
eight agents established in biological control of diffuse knapweed programs,
with the success of Larinus minutus, this project is moved to the category in
which one agent was successful.
604
One agent is usually sufficient for successful biological control of weeds
control was attributed to the combined impact of eight
agent species. This is curious because diffuse knapweed densities in British Columbia, upon which this
value was based, had not declined before 1998. Therefore, I suggest that this be changed, as the decline of
diffuse knapweed occurred after the establishment of
L. minutus that is likely the agent most responsible for
the successful control of diffuse knapweed, and I have
incorporated this change in Fig. 1.
One situation in which multiple species of agents
may be necessary is when the target weed has a greater
environmental range than the agent. For example, the
failure of M. janthinus to survive in Alberta, Canada
suggests that another agent or strain of control agent
may be necessary in the toadflax control programme
(McClay and Hughes, 2007).
In this review, I only considered published cases of
biological control success. Less successful programmes
are unlikely to be described in publications, and thus,
we cannot evaluate the impacts of the biological control agents in these. Cases of modest control by several
agents may therefore be overlooked and understudied
(M. Julien, personal communication). An exception is
a recent report on the impact of the mite Tetranychus
lintearius Dufor, on gorse, Ulex europaeus L. (Davies
et al., 2007). This agent was originally recommended
for use in gorse control by Zwlfer (1963) who observed
dead plants in their native habitat that he thought had
been killed by the mites. In the study of Davies et al.
(2007) and others reviewed there, the mites have been
found to reduce the growth and sometimes flower production of gorse. Damage from T. lintearius and five
other species of biological control agents has been insufficient, however, to successfully control gorse measured
as a reduction in the density of plants (Hill et al., 2000).
Although a majority of successful weed programmes
can be attributed to a single agent, this observation
does not help to identify the characteristics that will
allow agents to be successful before their release. One
factor that does stand out however is the ability of the
agent to kill the target weed. It appears that accumulated impact of several types of feeding damage from
different agents is not necessary for successful control.
To reduce the number of releases of exotic species for
biological control, the paradigm should change from
the assumption that multiple agents are required to a
greater focus on identifying potential biological control agents that can kill host plants. Introducing fewer
agents per programme would make biological control
of weeds more efficient, less costly and more environmentally benign. More focus on the potential efficacy
of biological agents could be cost effective (McClay
and Balciunas, 2005).
Acknowledgements
Thanks to Jenny Cory for comments on this manuscript. Research was supported by NSERC.
References
Albright, M., Harman, W., Fickbohm, S., Meehan, H.,
Graoff, S. and Austin, T. (2004) Recovery of native flora
and behavioral responses by Galerucella spp. following
biocontrol of purple loosestrife. American Midland Natu
ralist 152, 248251.
Barton, J., Fowler, S., Gianotti, A., Winks, C., de Beurs, M.,
Arnold, G. and Forrester, G. (2007) Successful biological
control of mist flower (Ageratina riparia) in New Zealand: Agent establishment, impact and benefits to the native flora. Biological Control 40, 370385.
Bourchier, R.S., Mortensen, K. and Crowe, M. (2002). Cen
taurea diffusa Lamarck, Diffuse Knapweed, and Centau
rea maculosa Lamarck, Spotted Knapweed (Asteraceae).
In: Mason, P.G. and Huber, J.T. (eds) Biological Control
Programmes in Canada, 19812000. CABI, Wallingford,
UK, pp. 302313.
Davies, J. T., J. E. Ireson and G. R. Allen. (2007). The impact
of the gorse spider mite, Tetranychus lintearius, on the
growth and development of gorse, Ulex europeaus. Bio
logical Control 41, 8693.
De Clerck-Floate, R., Wikeem, B.M. and Bourchier, R.
(2005) Early establishment and dispersal of the weevil,
Mogulones cruciger (Coleoptera: Curculionidae) for biological control of houndstongue (Cynoglossum officinale)
in British Columbia, Canada. Biocontrol Science and
Technology 15, 173190.
De Clerck-Floate, R.A. and Harris, P. (2002) Linaria dalmat
ica (L.) Miller, Dalmatian toadflax (Scrophulariaceae).
In: Mason, P.G. and Huber, J.T. (eds) Biological Control
Programmes in Canada, 19812000. CABI, Wallingford,
UK, pp. 368374.
De Clerck-Floate, R.A. and Schwarzlnder, M. (2002) Cy
noglossum officinale (L.), Houndstongue (Boraginaceae).
In: Mason, P.G. and Huber, J.T. (eds) Biological Control
Programmes in Canada, 19812000. CABI, Wallingford,
UK, pp. 337343.
Denoth, M., Frid, L. and Myers, J.H. (2002) Multiple agents
in biological control: improving the odds? Biological
Control 24, 2030.
Denoth, M. and Myers, J.H. (2005) Variable success of biological control of Lythrum salicaria in British Columbia.
Biological Control 32, 269279.
Grevstad, F. (2005) Ten-year impacts of the biological control agents Galerucella pusilla and G. calmariensis (Coleoptera:Chrysomelidae) on purple loosestrife (Lythrum
salicaria) in Central New York State. Biological Control
39, 13.
Hill, R. L., A. H. Gourlay and S. V. Fowler. (2000) The biological control program against gorse in New Zealand,
In: Spencer, N. (ed) Proceedings of the X International
Symposium Biological Control of Weeds. Montana State
University Bozeman, MO, USA, pp. 909917.
Hansen, R. (2004) Biological Control of Dalmatian Toadflax.
USDA-APHIS-PPQ-CPHST, Washington, DC, USA.
Hunt-Joshi, T. and Blossey, B. (2005) Interactions of root and
leaf herbivores on purple loosestrife (Lythrum salicaria).
Oecologia 142, 554563.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A World Catalogue of Agents and their Target
Weeds, 4th edn. CAB International, Wallingford, Oxon,
223 pp.
605
XII International Symposium on Biological Control of Weeds
Landis, D., Sebolt, D., Haas, M. and Klepinger, M. (2003)
Establishment and impact of Galerucella calmariensis
L. (Coleoptera: Chrysomelidae) on Lythrum salicaria L.
and associated plant communities in Michigan. Biological
Control 28, 7891.
Lindgren, C.J. (2000) Performance of a biological control
agent, Galerucella calmariensis L. (Coleoptera: Chrysomelidae) on purple loosestrife Lythrum salicaria L. in
Southern Manitoba (19931998). In: Spencer, N.R. (ed.)
Proceedings of the X International Symposium on Biolog
ical Control of Weeds. Montana State University, Bozeman, pp. 367382.
Lindgren, C.J., Corrigan, J. and De Clerck-Floate, R.A.
(2002). Lythrum salicaria L., purple loosestrife (Lythraceae). In: Mason, P.G. and Huber, J.T. (eds) Biological
Control Programmes in Canada, 19812000. CABI,
Wallingford, UK, pp. 383390.
Louda, S.M., Pemberton, R.W., Johnson, M.T. and Follett,
P.A. (2003) Nontarget effects the Achilles Heel of biological control? Retrospective analyses to reduce risk associated with biocontrol introductions. Annual Review of
Entomology 48, 365396.
McClay, A.S. and Balciunas, J.K. (2005) The role of prerelease efficacy assessment in selecting classical biological control agents for weeds applying the Anna Karenina
principle. Biological Control 35, 197207.
McClay, A.S. and Hughes, R. (2007) Temperature and hostplant effects on development and population growth of
Mecinus janthinus (Coleoptera:Curculionidae), a biological control agent for invasive Linaria spp. Biological
Control 40, 405410.
McConnachie, A., Hill, M. and Byrne, M. (2004) Field assessment of a frond-feeding weevil, a successful biological control agent of red waterfern, Azolla filiculoides, in
southern Africa. Biological Control 29, 326331.
Morin, L. and Edwards, P.B. (2006) Selection of biological
control agents for bridal creeper: a retrospective review.
Australian Journal of Entomology 45, 287291.
Morin, L., Evans, K. and Sheppard, A. (2006) Selection of
pathogen agents in weed biological control: critical issues
and peculiarities in relation to arthropod agents. Austra
lian Journal of Ecology 45, 349365.
Myers, J.H. and Bazely, D.R. (2003) Ecology and Control
of Introduced Plants. Cambridge University Press, Cambridge, UK.
Nowerski, R. (2004) Mecinus janthinus. In: Coombs,
E., Clark, J., Piper, G. and Cofrancesco, A. (eds) Bio
logical Control of Invasive Plants in the United States.
Oregon State University, Corvallis, OR, USA, pp.
392394.
Paynter, Q. (2005) Evaluating the impact of a biological control agent Carmenta mimosa on the woody wetland weed
Mimosa pigra in Australia. Journal of Applied Ecology
42, 10541062.
Schooler, S. (1998) Biological control of purple loosestrife
Lythrum salicaria by two chrysomelid beetles Galerucella
pusilla and G. calmariensis. MSc thesis. Oregon State
University, Corvallis, OR.
Seastedt, T.R., Gregory, N. and Buckner, D. (2003) Effect
of biocontrol insects on diffuse knapweed (Centaurea
diffusa) in a Colorado grassland. Weed Science 51,
237.
Smith, L. (2004) Impact of biological control agents on diffuse knapweed in central Montana. In: Cullen, J.M.,
Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L.
and Scott, J. (eds) XI International Symposium on Bio
logical Control of Weeds. CSIRO Entomology, Canberra,
Australia, pp. 589593.
Story, J., Callan, N., Corn, J. and White, L. (2006) Decline
of spotted knapweed density at two sites in western Montana with large populations of the introduced root weevil,
Cyphocleonus achates (Fahraeus). Biological Control 38,
227232.
Trujillo, E. (1985) Biological control of Hamakua pa-makani
with Cercosporella sp. in Hawaii. In: Delfosse, E.S. (ed)
Proceedings of the VI International Symposium on Bio
logical Control of Weeds. Agriculture Canada, Vancouver
Canada, pp. 661671.
Trujillo, E. (2005) History and success of plant pathogens for
biological control of introduced weeds in Hawaii. Biolog
ical Control 33, 113122.
Zwlfer, H. 1963. Ulex europaeus project: European investi
gations for New Zealand. Report 2. Commonwealth Institute of Biological Control, Delemont, Switzerland.
606
Evaluating implementation success for
seven seed head insects on Centaurea
solstitialis in California, USA
M.J. Pitcairn,1 B.Villegas,1 D.M. Woods,1,2 R. Yacoub3 and D.B. Joley4
Summary
A stage-based invasion model proposed by Colautti and MacIsaac (2004) is used to evaluate the
success of implementation efforts for seven seed head insects introduced on Centaurea solstitialis
L. (Asteraceae) in California, USA. Six insects were introduced intentionally; one was introduced
unintentionally. Establishment of initial foreign material (stage III invasion) at field nursery sites
was very successful (86%) with six of seven species establishing (five of six intentional, one of one
unintentional). For the intentional species, initial releases occurred at multiple locations, and establishment success among locations was 100% for four of the five species that established. Statewide
distribution of seed head insects introduced intentionally (stage IV invasion) was performed through
the network of County Agricultural Commissioners. Training workshops provided knowledge and
insects to county biologists for release in their county. Hundreds of releases occurred in this way, and
follow-up monitoring estimated establishment success above 87% for four of six species. Regional
spread and population increase to high density (stage V invasion) was examined in an independent
statewide survey of C. solstitialis seed heads in 2001 and 2002. Results showed that only one of the
intentional species, Eustenopus villosus (Boheman) and the unintentional species, Chaetorellia succinea (Costa), appear to have achieved high densities over large areas. The fly, C. succinea, appears
to be the most successful seed head insect, while E. villosus was second in abundance statewide. The
other seed head insects were recovered in low numbers statewide and appear to contribute little to the
herbivore pressure on this weed.
Keywords: invasion, stage-based implementation model.
Introduction
Development of a biological control organism may be
grouped into pre-release and post-release research efforts. Pre-release efforts include the foreign exploration and host-specificity testing of candidate natural
enemies, among other efforts (Harley and Forno, 1992;
Briese, 2000). Post-release efforts begin with the approval to release the control organism from quarantine
California Department of Food and Agriculture, Biological Control
Program, Sacramento, CA 95832, USA.
2
University of Wyoming, Department of Plant Sciences, Laramie, WY
82071-2000, USA.
3
California Department of Food and Agriculture, Division of Plant
Health and Pest Prevention Services, GIS Laboratory, Sacramento, CA
95814, USA.
4
California Department of Food and Agriculture, Seed Inspection Laboratory, Sacramento, CA 95832, USA.
Corresponding author: M.J. Pitcairn <[email protected]>.
CAB International 2008
1
and consist of release and establishment efforts, field
impact studies and non-target surveys (Hansen, 2004).
The multi-year release and establishment effort is the
implementation phase of a new biological control organism and is commonly carried out by state and local government entities. This phase occurs in several
steps: initial establishment of foreign material in field
nursery sites; distribution of domestic material from
nursery sites to satellite locations throughout the region; and eventual spread of the natural enemy away
from release sites into the surrounding infestations of
the target weed. The goal is to obtain self-sustaining
populations of the biological control agent throughout
the area infested by its host plant.
Several researchers have pointed out the similarity
between the intentional introduction of a biological
control organism and the invasion of an unintentionally introduced exotic organism (Crawley, 1989; Simberloff, 1989; Grevstad, 1999; McEvoy et al., 2000).
Colautti and MacIsaac (2004) and Colautti (2005)
607
XII International Symposium on Biological Control of Weeds
proposed a stage-based description of the invasion process over large geographic areas for invasive species.
Here, we modify their model to describe the stages
of the implementation phase of a biological control
organism and use this model to evaluate the success
of implementation efforts for a guild of exotic seed
head insectsintroduced against Centaurea solstitialis
L. (Asteraceae), commonly named yellow starthistle,
in California (USA).
Methods
The invasion model
Colautti and MacIsaac (2004) and Colautti (2005)
proposed a model of six stages to describe the invasion of an exotic organism into a new region. Stage 0
is the organism in its area of origin; in stage I, the organism is taken up by a transport vector (ship ballast,
contaminant in a seed shipment, etc.); in stage II, the
organism is introduced into the new region; in stage III,
the organism establishes a small, incipient population;
in stage IV, the organism spreads throughout the new
region; and in stage V, the organism has spread over a
large region and generally occurs in high density. Stage
IV can be achieved through two processes: formation
of several small satellite populations away from the initial infestation (stage IVa) or simple expansion of the
initial established population (stage IVb). Both stages
IVa and IVb expand to stage V.
The development and utilization of a biological control organism may be described in a similar
way. Stage 0 is the natural enemy in its hosts area
of origin; in stage I, a sample of the natural enemy
population is collected and held in quarantine for hostspecificity testing; in stage II, a sample of the material in quarantine is released on the target weed in the
new region; in stage III, the organism has established
small populations in field nursery sites; in stage IV,
biological control workers collect the organism from
the initial nursery sites and distribute it throughout the
region infested by the target weed. The goal is stage
V, where the biological control organism forms selfsustaining populations of high density throughout the
regions infested by the target weed. In this model, the
implementation phase of a biological control organism
begins with its approval to release (stage II) and ends
with the biological control organism occurring in high
density throughout the area infested by the target weed
(stage V).
Biology of yellow starthistle in California
Yellow starthistle is an invasive exotic weed from
the Mediterranean region of Europe and was likely introduced into California as a contaminant in shipments
of alfalfa seed (DiTomaso and Gerlach, 2000). It was
first recorded near the San Francisco Bay in 1859 and
now infests over 5.7 million ha in California alone (Pitcairn et al., 2006). Yellow starthistle is a winter annual
that invades rangelands, orchards, vineyards, pastures,
parks and natural areas. It is favored by soil disturbance
but can invade areas that have not been disturbed by
humans or livestock. Germination begins with the onset
of the winter rains; flowering begins in early summer
and continues into the fall. Individual plants may produce from one to 100 capitula (hereafter, seed heads);
infested areas commonly produce 650 to 700 seed
heads per square metre, but densities as high as 3000
seed heads per square metre have been reported (DiTomaso et al., 2003). In uninfested heads, approximately
30 to 40 achenes (hereafter, seeds) per head are produced. Yields of 50 million seeds/ha are common
(M. J. Pitcairn, unpublished data), but estimates of 120
to 500 million seeds/ha have been recorded in heavily
infested areas (Maddox, 1981; DiTomaso and Gerlach,
2000).
Yellow starthistle is the target of an ongoing biological control effort in the western United States. Research
efforts resulted in the release approval of six exotic
insects that oviposit in the immature seed heads, and
their larvae feed on developing seeds. The first insect
released was the gall fly, Urophora jaculata (Rondani)
(Dipt.: Tephritidae), in 1969, but it failed to establish
(Turner et al., 1994). Five other insects were released
from 1984 through 1992, and all have established:
Bangasternus orientalis (Capiomont) (Col.: Curculionidae), Eustenopus villosus (Boheman) (Col.: Curculionidae), Urophora sirunaseva (Hering) (Dipt.:
Tephritidae), Chaetorellia australis (Hering) (Dipt.:
Tephritidae) and Larinus curtus Hochhut (Col.: Curculionidae) (Turner et al., 1994). In California, three
species, B. orientalis, E. villosus and U. sirunaseva,
are now widespread. The two other insects, C. australis and L. curtus, occur in low numbers at a limited
number of locations. A seventh insect, Chaetorellia
succinea (Costa) (Dipt.: Tephritidae), was accidentally
introduced into southern Oregon in 1991 and is also
widely established throughout California (Balciunas
and Villegas, 1999).
Evaluation of the implementation stages
Stage III - releases of foreign material: The six insect species intentionally released as biological control organisms were collected from their area of origin
(Greece, Italy and Turkey) as larvae in seed heads, sent
to the United States Department of Agriculture, Agricultural Research Service quarantine facility in Albany, CA,
USA and held for emergence in sleeve cages (Turner et
al., 1996). Emerged adults were collected, identified and
a subsample killed and examined for internal parasites
and diseases. When ready, the remaining adults were released into field nursery sites (Turner et al., 1996). Most
nursery sites were monitored annually for recovery of
the insect and, when population levels were considered
608
Evaluating implementation success for seven seed head insects on Centaurea solstitialis in California, USA
high enough, the sites were used for training and a source
of adults for distribution in stage IV.
Stage IV - releases of domestic material: Stage IV
involves the transfer of knowledge and biological control organisms from state control to local control. The
State of California is divided into 58 counties for the
purposes of supporting local governmental activities.
Each county supports an Agricultural Commissioner
who maintains a small staff of biologists that assist
with issues concerning local agriculture. The California Department of Food and Agriculture performs
workshops and training sessions for the county biologists on the biology of biological control organisms
and the damage they cause to the target weed (Villegas, 1998). Training occurs through oral presentations
at field nursery sites and through active participation
of the attendees in the collection, counting and packaging of biological control organisms. Afterwards,
workshop attendees return to their counties and release the organisms at new field sites. This network of
trained county biologists has been very effective in the
distribution of biological control organisms throughout California.
A sample of stage IV release sites (n = 60120)
were monitored 3 to 4 years following their initial releases. The sites were chosen to represent the different
climatic regions where yellow starthistle grows. Plants
were swept with a sweep net at late bud stage or early
flowering to collect any active adult seed head insects,
and any evidence of oviposition and feeding damage
was recorded.
Stage V - local spread and population increase: It
was expected that each of the seed head insects would
increase their densities locally and spread throughout
the populations of yellow starthistle located nearby.
The regional spread of the seed head insects away from
their release sites was evaluated by a survey performed
in 2001 and 2002 where plant samples were collected along roads throughout areas infested with yellow starthistle. Samples occurred approximately every
16 km. Over 100 seed heads from at least ten plants
were collected, and the date and latitude and longitude
coordinates for each sample location were recorded. All
seed heads were returned to the laboratory, and a minimum of 100 heads from each sample was dissected,
and the presence of insects was recorded by species.
Each biological control organism produces a characteristic type of feeding damage that is easily recognized
upon dissection of the head. The exception was the
damage caused by the two species of Chaetorellia. A
second survey where adult flies were reared from seed
heads showed that C. australis was infrequently recovered, and its infestation rate was very low (<5%).
Thus, for the analysis reported here, we consider all
heads damaged by Chaetorellia spp. to be infested with
C. succinea.
Typically, immature seed head production begins in
May with peak production in early July (unpublished
data). Seed heads accumulate on plants and provide a
record of the cumulative attack by insects throughout
the season. Survey collections began after peak seed
head production (late July). A total of 421 samples
were collected during the two summers of this study.
A data file of insect abundance based on the proportion
of seed heads infested was plotted using a geographic
information system (GIS), and values between points
were interpolated using an inverse-distance weighting (IDW) algorithm that covered the entire state of
California.
Results
Stage III releases of foreign material
Five of the six seed head insects released as approved
biological control organisms established (Table 1). For
each species, releases of foreign material occurred in
three to seven locations within California (additional
releases occurred in Oregon, Washington and Idaho;
Turner et al., 1994). Establishment rates among locations were 100% except for U. jaculata, which failed
to establish, and for C. australis, which failed to establish at all five release sites in California. An overall
establishment rate of 27% for C. australis is estimated
from its rate of establishment among all locations in the
other western states (Turner et al., 1996).
The unintentional release of C. succinea occurred
when foreign material from quarantine consisting of a
mixture of C. succinea and C. australis adults was released at a field site near Merlin, Oregon (Balciunas and
Villegas, 1999). Both flies apparently established. Examination of the voucher specimens retained from the
overseas shipments received in quarantine suggested
that the accidental release of C. succinea occurred just
once (Balciunas and Villegas, 1999). Thus, the establishment rate of C. succinea in stage III is 100% (one
of one).
Stage IV releases of domestic material
Three insects, U. sirunaseva, B. orientalis and E.
villosus, built up high populations at field nursery sites
and were distributed statewide through a distribution
program. The program operated primarily through a series of workshops designed to train county biologists in
the identification, collection and release of the biological control organisms. Based on the training received
at the workshops, county biologists collected available
biological control organisms from their release sites in
stage III then returned to their county and released the
insects at their own nursery sites. The new county sites
would serve as sources for further distribution of insects
within their counties. Training workshop sites began 2
to 3 years following initial release of foreign material
in stage III. A total of 41,380 domestically produced
U. sirunaseva was released at 163 locations, 80,290
609
XII International Symposium on Biological Control of Weeds
Table 1.
ummary for the implementation of seven seed head insects on yellow starthistle (Centaurea solstitialis L.) in CalS
ifornia (USA). The values for stages III and IV represent the proportion of releases that successfully established.
Species
Urophora jaculata (Rondani)
Urophora sirunaseva (Hering)
Bangasternus orientalis (Capiomont)
Chaetorellia australis (Hering)
Eustenopus villosus (Boheman)
Larinus curtus Hochhut
Chaetorellia succinea (Costa)
Stage III (%)
Stage IV (%)
Stage V
0
100
100
27b
100
100
100c
87
92
19
92
25
100
No
No
No
Yes
No
Yes
Turner et al. (1994).
Turner et al. (1996).
c
Balciunas and Villegas (1999).
a
B. orientalis at 371 locations, and 316,000 E. villosus
at 1024 locations. Post-release monitoring at a subset
of these locations estimated establishment at 87% (45
of 52, U. sirunaseva), 92% (118 of 128, B. orientalis)
and 92% (119 of 129, E. villosus; Table 1).
Stage IV distribution of C. australis and C. succinea
in California resulted from a release effort during 1994
through 1996 (Balciunas and Villegas, 1999). Encouraged by high populations of what was thought to be
C. australis in southern Oregon, 9463 Chaetorellia
adults were collected and released at 21 locations in
California. It was during this release effort that the
presence of C. succinea was discovered, and releases
were discontinued. Post-release surveys of the 21 release sites recovered C. australis at four locations (establishment rate = 19%) and C. succinea at 21 locations
(establishment rate = 100%).
The performance of the weevil, L. curtus, following establishment in stage III was disappointing as
populations failed to increase and were recovered in
low numbers several years afterward. A small number of field-collected adults from several release sites
were found to be infested with a species of Nosema,
a microsporidian parasite of the alimentary canal. A
population of L. curtus in northern Oregon had readily established and built up high population numbers.
Field samples later revealed no infection by Nosema
at this location. It was thought that the microsporidian
hindered population growth, so over 5600 adult L. curtus were collected from northern Oregon and released
at 23 locations in California from 1997 to 1999. Postrelease monitoring showed that the weevil established
at five of 20 locations (establishment rate = 25%), but
densities of the established populations remained extremely low.
Stage V local spread and
population increase
Four species of seed head insects were recovered
during an independent survey of yellow starthistle
plants away from release sites in 2001 and 2002: B.
orientalis, U. sirunaseva, E. villosus and C. succinea.
All were found widespread throughout the state but in
varying levels of abundance. Both B. orientalis and U.
sirunaseva were recovered from 62% of the locations
but occurred mostly in low numbers (usually <15%
of seed heads attacked). In contrast, E. villosus was
recovered at 80% of the locations, and its abundance
ranged from 0% to 93% of the heads attacked. The fly,
C. succinea, was recovered at 99% of the locations, and
its abundance ranged from 0% to 96% of the heads attacked. The weevil, L. curtus, was not recovered. The
fly, C. australis, was recovered in a follow-up survey
where adult flies were reared from sampled seed heads;
however, adult emergence rarely exceeded five flies per
100 heads.
Insect abundance, estimated as the number of heads
attacked per sample, was interpolated between sample
locations using a GIS system, and the resulting maps
showed regional differences in abundance. For each
insect species, the frequency distribution of abundance
values was sorted by magnitude and divided into ten
quantiles. The top quantile consisted of values in the
top 10% of those recorded for the species. Areas of
highest abundance were identified as those areas with
interpolated values within the top quantile (Fig. 1). For
B. orientalis, locations with abundance values above
18% heads attacked represented the top quantile, and
these occurred in the mountain ranges of the most
northern areas of California. For U. sirunaseva, locations with values above 9% attack rate represented the
top quantile, and these areas were scattered within the
northern and southern regions of the Coastal Mountain
Range. For E. villosus, its top quantile consisted of locations with values above 41% attack rate, and these
were observed throughout the mountains of northern
California and along the foothills of the Sierra Nevada
Mountains. Areas with the lowest abundance of E. villosus were observed in the San Joaquin and Sacramento Valleys. The fly, C. succinea, was the most common
seed head insect recovered in the survey. Its top quantile consisted of abundance values above 70%. However, in comparison with the other seed head insects,
610
Evaluating implementation success for seven seed head insects on Centaurea solstitialis in California, USA
Figure 1.
Plot of interpolation values for the four seed head insects recovered from Centaurea solstitialis L. in California during 20012002. All plots show areas of highest abundance as indicated by seed head attack rates in the top 10% of observed values (top quantile). The weevil
Bangasternus orientalis and the fly Urophora sirunaseva rarely exceeded attack rates of
18% and 9%, respectively. The weevil Eustenopus villosus and the fly Chaetorellia succinea occurred in higher numbers (>41% and 50%, respectively). Sample locations are
indicated by the filled dots.
abundance values above 50% are considered high. For
C. succinea, areas with interpolated abundance values
above 50% occurred throughout central and southern
California (Fig. 1).
Discussion
The stage-based implementation model identified three
main stages in the release and establishment of a biological control organism: establishment of foreign material
from quarantine (stage III), establishment of domestic
material (stage IV) and the increase and spread of the
control organism throughout the region (stage V). Examination of establishment rates among stages shows
that implementation of the six seed head insects intentionally introduced as biological control organisms was
very successful. Five of six species established (83%)
and establishment rates of the foreign material were
100% for four of five species. Collection and distribution of material produced domestically (stage IV) was
also fairly successful with three of five species showing
establishment rates above 87% throughout California.
Interestingly, the accidentally introduced fly, C. succinea, also showed a high rate of establishment (100%
in stages III and IV). However, the regional spread of
the seed head insects at stage V was not as successful
as only two species, E. villosus and C. succinea, appear
to have built up populations approaching high densities
611
XII International Symposium on Biological Control of Weeds
over large regional areas. The fly, C. succinea, appears
to be the most successful organism with estimated attack rates greater than 50% over half of the area infested by yellow starthistle (Fig. 1). The weevil, E.
villosus, was second in abundance with approximately
35% of the yellow starthistle infestation experiencing
attack rates greater than 41%. None of the other insects
appear to contribute much to the overall natural enemy
pressure on this weed.
Several studies have identified propagule pressure
as an important factor in the successful establishment
of invading organisms. For example, Kolar and Lodge
(2001) reviewed published studies on invasive species
in an effort to identify attributes that might predict probable invaders. While the information they presented
was limited to only a few taxa, one strong pattern was
that successful establishment was positively related to
propagule pressure. Propagule pressure, either number
of individuals released or number of releases, is one
factor that biological control workers have increasing
control over due to improvements in rearing technology
and faster, more efficient shipping abilities. For implementation efforts, establishment rates in stages III and
IV would benefit from increases in the number of insects used. The only stage over which we appear to have
little control is stage V. Reviews of past biological control projects (e.g. Hall and Ehler, 1979; Crawley, 1989;
Simberloff, 1989; Coombs, 2004) have concentrated on
establishment success in stages III and IV but provided
little examination of projects in stage V. While establishment is an important component in the implementation of a biological control organism, ultimate success in
controlling the target weed may reside in the attributes
of control organisms that result in their ability to obtain
high densities and regional spread. More research on
the transition of biological control organisms to stage V
may greatly improve our understanding of the attributes
of an effective biological control agent and may lead to
releases of more effective agents in the future.
Acknowledgements
We thank Kathy Chan for providing release information from unpublished quarantine records located at
the USDA Agricultural Research Service facility in
Albany, CA.
References
Balciunas, J.D. and Villegas, B. (1999) Two new seed head
flies attack yellow starthistle. California Agriculture 53,
811.
Briese, D.T. (2000) Classical biological control. In: Sindel,
B.M. (ed) Australian Weed Management Systems. R.G.
and F.J. Richardson, Melbourne, Australia, pp. 161192.
Colautti, R.I. (2005) In search of an operational lexicon for
biological invasions. In: S. Inderjit (ed) Invasive Plants:
Ecological and Agricultural Aspects. Birkhauser Verlag,
Basel, Switzerland, pp. 115.
Colautti, R.I. and MacIsaac, H.J. (2004) A neutral terminology to define invasive species. Diversity and Distributions 10, 135141.
Coombs, E.M. (2004) Factors that affect successful establishment of biological control agents. In: Coombs, E.M.,
Clark, J.K., Piper, G.L. and Cofrancesco, A.F. Jr. (eds)
Biological Control of Invasive Plants in the United States.
Oregon State University Press, Corvallis, OR, USA,
pp. 8594.
Crawley, M.J. (1989) The successes and failures of weed biocontrol using insects. Biocontrol News and Information
10, 213223.
DiTomaso, J.M. and Gerlach, J.D. Jr. (2000) Centaurea solstitialis L. (yellow starthistle). In: Bossard, C.C., Randall,
J.M. and Hoshovsky, M.C (eds) Invasive Plants of Californias Wildlands. University of California Press, Berkeley, CA, USA, pp. 101106.
DiTomaso, J.M., Kyser, G.B. and Pirosko, C.B. (2003) Effect
of light and density on yellow starthistle (Centaurea solstitialis) root growth and soil moisture use. Weed Science
51, 334341.
Grevstad, F.S. (1999) Factors influencing the chance of population establishment: implications for release strategies in
biocontrol. Ecological Applications 9, 14391447.
Hansen, R.W. (2004) Handling insects for use as terrestrial
biological control agents. In: Coombs, E.M., Clark, J.K.,
Piper, G.L. and Cofrancesco, A.F., Jr. (eds) Biological
Control of Invasive Plants in the United States. Oregon
State University Press, Corvallis, OR, USA, pp. 5970.
Harley, K.L.S. and Forno, I.W. (1992) Biological Control of
Weeds, A Handbook for Practitioners and Students. Inkata
Press, Melbourne, Australia, 74 pp.
Hall, R.W. and Ehler, L.E. (1979) Rate of establishment of
natural enemies in classical biological control. Bulletin of
the Entomological Society of America 25, 280282.
Kolar, C.S. and Lodge, D.M. (2001) Progress in invasion biology: predicting invaders. Trends in Ecology and Evolution 16, 199204.
Maddox, D.M. (1981) Introduction, Phenology, and Density
of Yellow Starthistle in Coastal, Intercoastal, and Central
Valley Situations in California. Agricultural Research Results, ARR-W-20, Agricultural Research Service (Western
Region), US Department of Agriculture, Oakland, CA,
USA, 33 pp.
McEvoy, P., Grevstad, F.S., Schooler, S., Schat, M. and
Coombs, E.M. (2000) Weed biocontrol as an invasion
process. In: Spencer, N.R. (ed) Proceedings of the X International Symposium of Biological Control of Weeds.
Montana State University, Bozeman, MT, USA, p. 596.
Pitcairn, M.J., Schoenig, S., Yacoub, R. and Gendron, J.
(2006) Yellow starthistle continues its spread in California. California Agriculture 60, 8390.
Simberloff, D. (1989) Which insect introductions succeed
and which fail? In: Drake, J.A., Mooney, H.A., diCastri,
F., Groves, R.H., Kruger, F.J., Rejmanek, M. and Williamson, M. (eds) Biological Invasions, A Global Perspective.
Scientific Committee on Problems of the Environment
(SCOPE), pp. 6175.
612
Evaluating implementation success for seven seed head insects on Centaurea solstitialis in California, USA
Turner, C.E., Johnson, J.B. and McCaffrey, J.P. (1994) Yellow starthistle, Centaurea solstitialis L. (Asteraceae). In:
Nechols, J. (ed) Biological Control in the U. S. Western
Region: Accomplishments and Benefits of Regional Research Project W-84 (19641989). University of California, Division of Agriculture and Natural Resources,
Berkeley, CA, USA, pp. 274279.
Turner, C.E., Piper, G.L. and Coombs, E.M. (1996) Chaetorellia australis (Diptera: Tephritidae) for biological control
of yellow starthistle, Centaurea solstitialis (Compositae),
in the western USA: establishment and seed destruction.
Bulletin of Entomological Research 86, 177182.
Villegas, B. (1998) Implementation status of biological
control of weeds in California. In: Woods, D.M. (ed) Biological Control Program Annual Summary, 1997. California Department of Food and Agriculture, Plant Health
and Pest Prevention Services, Sacramento, CA, USA,
pp. 3538.
613
The ragweed leaf beetle Zygogramma
suturalis F. (Coleoptera: Chrysomelidae)
in Russia: current distribution, abundance
and implication for biological control of
common ragweed, Ambrosia artemisiifolia L.
S.Ya. Reznik, I.A. Spasskaya, M.Yu. Dolgovskaya,
M.G. Volkovitsh and V.F. Zaitzev
Summary
The ragweed leaf beetle, Zygogramma suturalis F. (Coleoptera: Chrysomelidae), was introduced to
Russia in 1978 against the common ragweed, Ambrosia artemisiifolia L. By 1985, it successfully
acclimated and suppressed ragweed in the original release site and several neighbouring fields. However, because of crop rotation, its population density drastically decreased. In 2005 and 2006, we
conducted selective quantitative sampling in Southern Russia. The results showed that the ragweed
leaf beetle was distributed over an area of about 50,000 km2 in Krasnodar territory, west of Stavropol
territory and south of Rostov province, i.e. most of the area heavily infested by ragweed in Russia.
However, the average Z. suturalis population density was very low: approximately 0.001 adults per
square metre (m2) in crop rotations and approximately 0.1 adults/m2 in more stable habitats, although
in a few of the studied plots, up to 23 adults/m2 were recorded. As for common ragweed, average
percent cover in crop rotations and in stable habitats was approximately 1% and 40%, correspondingly. A detectable level of host plant damage (5%) was recorded only in a few plots. The observed
spatial variation of the Z. suturalis population density was mostly determined by the host plant abundance. The last was strongly dependent on the stability of the habitat, being much higher in stable
habitats. Thus, it is still possible that stable protected field nurseries could be a promising method of
Z. suturalis propagation for biological control of ragweed in surrounding locations.
Keywords: weeds, biological control, post-release evaluation.
Introduction
Common ragweed, Ambrosia artemisiifolia L., is one
of the most noxious invasive weeds in Russia infesting
agricultural fields, ruderal habitats etc. over more than
60,000 km2 (Ulyanova, 2003). In an attempt to control this weed, the ragweed leaf beetle, Zygogramma
suturalis F., was introduced to Russia from the United
States and Canada (Kovalev et al., 1983). In 1978,
Zoological Institute, Universitetskaya nab., 1, St. Petersburg 199034,
Russia.
Corresponding author: S.Ya. Reznik <[email protected]>.
CAB International 2008
about 1500 specimens were released in the vicinity of
Stavropol. By 1981, a significant increase in the Z. suturalis population density was recorded. In 1983, ragweed at the experimental release site was eliminated,
and the ragweed leaf beetle began to spread over surrounding fields (Kovalev et al., 1983). Thus, the initial
phase of this introduction was a population explosion
with more than a 30-fold yearly increase in number,
up to 100,000,000 adults/km2 in fields and up to 5000
adults/m2 in aggregations (Kovalev, 1988). In some
cases, a solitary population wave, i.e. a moving zone
of ultra-high population density, was observed (Kovalev, 1988). Following these waves, common ragweed
was exterminated locally very quickly, which allows
one to expect highly efficient biological control of
614
The ragweed leaf beetle Zygogramma suturalis F. (Coleoptera: Chrysomelidae) in Russia
the weed (Kovalev, 1988; Goeden and Andres, 1999).
From 1984 to 1986, a number of further releases were
made in the south of European Russia and Ukraine, but
the population wave was not subsequently observed,
possibly because the insects were released in regularly
exploited agricultural landscapes but not in special protected sites (Reznik et al., 1994).
Later studies conducted in the area of the first release site have demonstrated that, in crop rotations,
Z. suturalis population density drastically decreased
despite a few patches with insect overcrowding and local weed control (Reznik et al., 1994; Reznik, 1996).
In these studies, regular estimations of the ragweed leaf
beetle population density were restricted to a limited
area in the environs of the initial release site. Now, almost 30 years after the initial release, data on Z. suturalis distribution and abundance are still fragmental
(Ugryumov et al., 1994; Polovinkina and Yaroshenko,
1999; Oskin, 2002; Esipenko and Belikova, 2004).
At present, common ragweed invading Central and
Western Europe poses a threat to the economy and human health by producing highly allergenic pollen in
Austria, France, Italy, Hungary and Switzerland (Wittenberg, 2005). Analysis of the results of unsuccessful
projects as well as of successful ones is important in
improving the theory of biological control of weeds
(Harris, 1993; Stiling, 1993; McFadyen, 1998; MllerSchrer et al., 2000; Raghu et al., 2006). In addition,
the planned invasion of the ragweed leaf beetle can be
considered as a model for predicting consequences of
real invasions (Ehler, 1998; Fagan et al., 2002).
In 20052006, we conducted selective quantitative
sampling over the whole area infested by A. artemisiifolia in Southern Russia. The aims of this study were
to estimate Z. suturalis population densities in relation
to environmental factors and to evaluate the impact of
the phytophage on the targeted weed.
etc.) were considered as sampling units. The size of the
plot varied from 10 to 15 m2 (isolated ragweed patches)
to many hectares (agricultural fields). Usually, all plots
along a randomly selected route were inspected.
For each plot, its square (m2) and habitat type: agricultural field, field margin or ruderal (mostly roadsides)
were recorded. The ragweed characteristics (percent
cover and average height) were measured. Z. suturalis
population density was measured by two methods. First,
adults over a given square (one or several randomly
selected transects of 1 m width and of total length depended on the size of the plot) were counted, and the
mean number of adults per square metre was calculated.
Second, sweeping with a standard net (along other
randomly selected transects) was made, and the mean
number of adults per ten sweeps was recorded. Impact
on the targeted weed was evaluated visually by considering the relation (percentage) of the area eaten to primary leaf area. In addition, geographical coordinates of
each plot, height above sea level, date of sampling and
certain additional data were recorded.
Material
Sampling was conducted from 2005 to 2006 in three
main areas of Southern Russia heavily infested with
common ragweed: Krasnodar territory, Stavropol territory and Rostov province (Fig. 1). At nine locations
(191 plots, total approximately 6.3 km2) where the ragweed leaf beetle was relatively abundant, 760 adults
were visually counted within 22,434 m2 of transects,
and 885 adults were collected by 6480 sweeps. In addition, a number of occasional records of Z. suturalis
adults were made and used to determine a distribution
range (Fig. 1). At six locations, the ragweed leaf beetle
was not found, although 1020 ragweed plots (total,
10003000 m2) per location were carefully inspected.
Statistical treatments
Methods and material
Methods
All studies were conducted between 15 July and
15 August, when Z. suturalis adults usually reach peak
density (Kovalev et al., 1983). In our earlier investigations (Reznik, 1993), 0.1m2 plots were used for exact
determination of population densities by counting the
beetles and measuring the plants. However, for the
broad scale investigations, this method was found to
be too time consuming. Later (Reznik et al., 1994), the
plant and the insect population densities were visually
estimated with a scale of 0 to 5. But for the current
study, fast visual estimation was not suitable, considering the low population density of the beetle. Thus, a
new method was developed.
Plots with more or less uniform vegetation (particularly, ragweed abundance) separated from other plots by
some natural borders (field boundary, road, forest belt
As could be expected, estimations of Z. suturalis
population density made by counting and by sweeping
closely correlated (r = 0.78, n = 183, Spearman rank correlation). Moreover, interrelations between data obtained
by counting and by sweeping fit well (r = 0.49, Pearson
correlation) with the linear regression Dc = 0.14Ds, where
Dc is the population density estimated by counting, and
Ds is that estimated by sweeping. Thus, the data obtained
by two methods were combined with the formula Da =
1/2(Dc + 0.14Ds), where Da is the averaged Z. suturalis
population density (adults per square metre). This integral
parameter correlated well both with the results of counting and those of sweeping (r = 0.95 and r = 0.91, correspondingly, Spearman rank correlation). In eight cases,
when sweeping was not made, Dc was taken as Da.
The ragweed population density was described not
only by the height and percent cover but also by an integral characteristic, the ragweed abundance (the product of multiplication of the cover by the height). As the
615
XII International Symposium on Biological Control of Weeds
Figure 1.
The geographical range of Ambrosia artemisiifolia and Zygogramma suturalis in Russia. Squares:
Quantitatively studied locations; triangles: occasional records of Z. suturalis adults; circles: locations
where Z. suturalis was not found; ring: the environs of the first release site, where the earlier studies
(Reznik, 1993; Reznik et al., 1994) were conducted; dash line: the approximate north-eastern boundary of the ragweed invasion.
distributions of A. artemisiifolia and Z. suturalis population densities were far from normal (Fig. 2), these data
were transformed to decimal logarithms and then treated by analysis of variance (ANOVA), the Tukey test
and linear regression analysis. Medians and quartiles of
untransformed data were used as descriptive statistics.
All calculations were made with SYSTAT 10.2.
Results
Wide-scale sampling shows that at present, the ragweed leaf beetle is widely distributed in Krasnodar territory, west of Stavropol territory and south of Rostov
province (Fig. 1). Z. suturalis population density was
significantly (F = 3.4, n = 191, p = 0.001) different
among nine quantitatively studied locations and positively correlated with the average ragweed abundance
(r = 0.91, n = 9, p < 0.001). Two-way ANOVA of the
transformed data on six locations where all main types
of habitats (fields, field margins and ruderal sites) were
inspected (n = 164) showed that the ragweed leaf beetle population density was more dependent on type of
habitat (F = 8.6, p < 0.001) than on location (F = 3.1,
p = 0.012). As for the ragweed, its height was slightly
dependent on location (F = 2.5, p = 0.03) and strongly
dependent on type of habitat (F = 11.7, p < 0.001),
while percent cover was dependent only on type of habitat (F = 57.3, p < 0.001) but not on location (F = 1.4,
p = 0.23). The Tukey test showed that ragweed height,
percent cover and Z. suturalis population density in
agricultural fields were significantly (p < 0.001) lower
that that in field margins and in ruderal sites, while the
difference between field margins and ruderal sites was
not significant (p > 0.3). Thus, for further data analysis,
field margins and ruderal sites were pooled and considered as stable habitats, in contrast to crop rotation.
The average (hereafter, medians, quartiles and sam
ple size are given) population density of the ragweed
leaf beetle was very low, 0.001 (00.03) adults per
square metre in agricultural fields subjected to crop rotation (n = 40) and 0.1 (0.040.27) adults per square
metre in stable habitats (n = 151). However, in few of
the studied plots, Z. suturalis population density ranges
up to two to three adults per square metre (Fig. 2).
Considering rather high level of ragweed infestation:
percent cover 1% (02) and 40% (2060%), height 10
(030) and 50 (3080) cm in crop rotation and in stable
habitats, correspondingly, it is not surprising that a detectable level of host plant damage (5%) was recorded
only in one agricultural field (sunflower) and in few
stable plots with relatively high density of the phytophage (Fig. 3).
616
The ragweed leaf beetle Zygogramma suturalis F. (Coleoptera: Chrysomelidae) in Russia
Figure 2.
Distributions of the percent of studied plots according to the Ambrosia artemisiifolia and Zygogramma suturalis population densities which correspond to mean numbers of Z. suturalis adults per square metre (n =
191), ragweed percent cover (n = 191) and mean ragweed heights (n = 175).
It is noteworthy that, in spite of the drastic difference between medians, general linear model analysis
showed that Z. suturalis population density significantly (p < 0.001) depended on the host plant abundance (cover multiplied by height) but not on the type
of habitat (p = 0.95). Parameters of linear regression
of log-transformed Z. suturalis population density on
log-transformed ragweed abundance almost coincided:
Y = 0.62X 3.2 (r = 0.60, n = 40, p < 0.001) for crop rotations and Y = 0.59X 3.1 (r = 0.40, n = 151, p < 0.001)
for stable habitats (see Fig. 3 for untransformed data).
Discussion
First, our data suggest that Z. suturalis spreads practically over the whole area heavily infested by A. artemisiifolia in Russia (Ulyanova, 2003), although it
was not found in the less infested boundary zone.
Our observations suggest that in this zone, ragweed
grows mostly in settlements and city environs, while
in agricultural landscapes (including field margins),
it is practically absent (unpublished data). It is well
known (Maryushkina, 1986; Ulyanova, 2003) that the
northern limit of A. artemisiifolia geographical distribution in Russia is determined by the day length and
temperature (ragweed is a typical short-day plant) and
the eastern limit by precipitation. Urban environments
were shown to favor the ragweed establishment in the
United States (Ziska et al., 2006). Thus, it is possible
that Russian settlements could also provide better conditions for the ragweed reproduction, i.e. more warm
habitats in the northern boundary and more humid soil
in the eastern boundary. Possibly, such a dispersed synanthropic distribution of ragweed in the boundary zone
prevents or delays its colonization by the ragweed leaf
beetle.
In the heavily infested zone, significant difference
among studied plots and locations is most probably
connected with uneven ragweed abundance but not with
delayed colonization by the ragweed leaf beetle. An increase of the Z. suturalis population density with the A.
artemisiifolia abundance was earlier recorded at smaller spatial scales (Reznik, 1993; Reznik et al., 1994). As
for the two components of the ragweed abundance, i.e.
617
XII International Symposium on Biological Control of Weeds
Z. suturalis population desnity
10
0.1
0.01
1
10
100
1000
10000
100000
Ragweed abundance (percent cover x height)
Agricultural fields
Figure 3.
Stable habitats
Relationship between Ambrosia artemisiifolia abundance (cover
height) and Zygogramma suturalis population density. Each symbol indicates one plot, an agricultural field subjected to crop rotation (n = 40)
or more stable habitat: field margin, roadside, ruderal site, etc. (n = 151).
Regression line (Y = 0.61X 3.1, r = 0.62, n = 191, p < 0.001) is based
on the log-transformed pooled data. Dashed ellipse indicates plots with
a detectable rate of the ragweed damage (5% of the total leaf surface).
percent cover and average height, it is interesting that
relatively stable habitats mostly differed from crop rotations not in ragweed percent cover but in its height. This
could be explained by the fact that mowed ragweed can
regrow quickly, producing many leaves on short stems.
Hence, at the same percent cover in stable habitats, ragweed is much higher.
Note that Z. suturalis population densities recorded
in 2005 to 2006 were lower than those observed during
our previous investigations (Reznik, 1993; Reznik et al.,
1994; Reznik, 1996). As for the data obtained by other
authors, the comparison is hampered by the fact that,
very often, not average values but upper limits were
published. Maximal Z. suturalis population density recorded in Stavropol and Krasnodar territories at the end
of the 1990s was six and 5070 adults per square metre, respectively (Polovinkina and Yaroshenko, 1999;
Oskin, 2002), which is similar to our data obtained
in 19881989 (Reznik et al., 1994) but much higher
than the results of the present. One of the recent papers
(Esipenko and Belikova, 2004) also pointed out a significant decrease in the Z. suturalis population density.
This assertion is indirectly supported by the fact that in
several regions of Krasnodar territory, where in 1993
the ragweed leaf beetle population density ranged up to
400 adults/m2 (Ugryumov et al., 1994), in 2003, only
166 adults were collected for phenetic analysis (Esip-
enko and Savva, 2004). It is not clear if this decline in
population density is connected with recent climatic
changes, with increasing impact of natural enemies, or
with some other factors (Esipenko and Belikova, 2004).
Thus, our results suggest that, by now, the ragweed
leaf beetle had spread throughout most of its potential
area, and further expansion is unlikely. The observed
spatial variation of Z. suturalis population density
seems to be mostly determined by the host plant abundance. The last is strongly dependent on the stability of
the habitat, being much lower in crop rotation fields. In
overwhelming majority of inspected populations, the
impact on the targeted weed is negligible. However,
having regard to the spectacular success achieved in
the permanent experimental plot from 1983 to 1985
(Kovalev, 1988), it is still possible that stable protected
field nurseries could be a promising method of Z. suturalis propagation for biological control of ragweed in
surrounding areas.
As already highlighted by the biological invasion
theory, we conclude that, in spite of certain prerequisites for invasion (abundant food resources, absence of
competitors, specific parasitoids and predators), very
strong population explosion recorded during first years
after introduction finally resulted in wide geographic
distribution but low mean population density and negligible ecological role of an alien insect species.
618
The ragweed leaf beetle Zygogramma suturalis F. (Coleoptera: Chrysomelidae) in Russia
Acknowledgements
This work was partly funded by the grant Scientific
bases of the conservation of biodiversity in Russia
from the Presidium of the Russian Academy of Sciences. We are grateful to Neal R. Spencer for a critical
review of the manuscript.
References
Ehler, L.E. (1998) Invasion biology and biological control.
Biological Control 13, 127133.
Esipenko, L.P. and Belikova, N.V. (2004) Preliminary results
of the studies on Zygogramma suturalis (F.) (Coleoptera,
Chrysomelidae) biology in Krasnodar territory. In: Nadykta, V.D., Ismailov, V.Ya. and Sugonyaev, E.S. (eds)
Biological Plant Protection as a Basis for Agroecosystem
Stability, vol. 1. VNIIBZR, Krasnodar, Russia (in Russian), pp. 122124.
Esipenko, L.P. and Savva, A.P. (2004) Phenotypic variations
in the ragweed leaf beetle Zygogramma suturalis (F.) (Coleoptera, Chrysomelidae). In: Nadykta, V.D., Ismailov,
V.Ya. and Sugonyaev, E.S. (eds) Biological Plant Protection as a Basis for Agroecosystem Stability, vol. 2. VNIIBZR, Krasnodar, Russia (in Russian), pp. 236241.
Fagan, W.F., Lewis, M.A., Neubert, M.G. and van den
Driessche, P. (2002) Invasion theory and biological control. Ecology Letters 5, 148151.
Goeden, R.D. and Andres, L.A. (1999) Biological control of
weeds in terrestrial and aquatic environments. In: Bellows, T.S. and Fisher, T.W (eds) Handbook of biological
control. Academic, New York, USA, pp. 871890.
Harris, P. (1993) Effects, constraints and the future of weed
biocontrol. Agriculture, Ecosystems and Environment 46,
289303.
Kovalev, O.V. (1988) A new biological phenomenon: the solitary population wave, and its role in the biological control
of pests, weeds in particular. Entomophaga 33, 259260.
Kovalev, O.V., Reznik, S.Ya. and Cherkashin, V.N. (1983)
Specific features of the methods of using Zygogramma
Chevr. (Coleoptera, Chrysomelidae) in biological control
of ragweeds (Ambrosia artemisiifolia L., A. psilostachya
D.C.). Entomologicheskoe Obozrenije 62, 402408 (in
Russian).
Maryushkina, V.Ya. (1986) Common ragweed and the principles of its biological control. Naukova Dumka, Kiev,
USSR (in Russian), 119 pp.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
Mller-Schrer, H., Scheepens, P.C. and Greaves, M.P.
(2000) Biological control of weeds in European crops:
recent achievements and future work. Weed Research 40,
8398.
Oskin, A.A. (2002) Control of common ragweed in Stavropol territory. Zashchita rastenii 12, 3334 (in Russian).
Polovinkina, O.A. and Yaroshenko, V.A. (1999). Investigations on the results of the introduction and biocenotic relationships of the ragweed leaf beetle. In: Molochnikov,
N.R. (ed) Man and Noosphere. Proceedings of the AllRussian Conference of the Academy of Natural Sciences.
KGU, Krasnodar, Russia, pp. 7879 (in Russian).
Raghu, S., Wilson, J.R. and Dhileepan, K. (2006) Refining
the process of agent selection through understanding plant
demography and plant response to herbivory. Australian
Journal of Entomology 45, 308316.
Reznik, S.Ya. (1993) Influence of target plant density on herbivorous insect oviposition choice: Ambrosia artemisiifolia L. (Asteraceae) and Zygogramma suturalis F. (Coleoptera, Chrysomelidae). Biocontrol Science and Technology
3, 105113.
Reznik, S.Ya. (1996) Classical biocontrol of weeds in crop
rotation: a story of failure and prospects for success. In:
Moran, V.C. and Hoffman, J.H. (eds) Proceedings of the IX
International Symposium on Biological Control of Weeds.
University of Cape Town, South Africa, pp. 503506.
Reznik, S.Ya., Belokobylskiy, S.A. and Lobanov, A.L.
(1994) Weed and herbivorous insect population densities
at the broad spatial scale: Ambrosia artemisiifolia L. and
Zygogramma suturalis F. (Col., Chrysomelidae). Journal
of Applied Entomology 118, 19.
Stiling, P. (1993) Why do natural enemies fail in classical
biological control programs? American Entomologist 39,
3137.
Ugryumov, E.M., Samus, V.I., Savva, A.P. and Vyalykh
A.K. (1994) Ecologically safe methods of control of common ragweed. In: Sokolova, N.K. and Tishchenko, Z.A.
(eds) Ecologically Safe and Pesticide-Free Technologies
of Plant Production, vol. 2. VNIIBZR, RASKHN, Pushchino, Russia, pp. 251252 (in Russian).
Ulyanova, T.N. (2003) Invasion and introduction of vascular
plants in the flora of Russia and adjacent countries during the last 50 years. In: Pavlov, D.S. Dgebuadze, Y.Y.,
Korneva, L.G. and Slynko Y.V. (eds) Invasion of Alien
Species in Holarctic. IBFW, Borok, Yaroslavl prov., Russia, pp. 133139 (in Russian).
Wittenberg, R. (ed) (2005) An Inventory of Alien Species and
their Threat to Biodiversity and Economy in Switzerland.
Federal office for the Environment, Bern, 155 pp.
Ziska L .H., George, K. and Frenz, D.A. (2006) Establishment
and persistence of common ragweed (Ambrosia artemisiifolia L.) in disturbed soil as a function of an urbanrural
macro-environment. Global Change Biology 12, 19.
619
Long-term field evaluation of Mecinus
janthinus releases against Dalmatian
toadflax in Montana (USA)
S.E. Sing,1 D.K. Weaver,1 R.M. Nowierski2 and G.P. Markin3
Summary
The toadflax stem mining weevil, Mecinus janthinus Germar, was first released in the United States
in Montana, in 1996. This agent has now become established to varying degrees after subsequent releases made at sites throughout the state. Multiple releases of M. janthinus have presented researchers
with a unique opportunity to evaluate the efficacy of this agent in diverse habitats and under a variety
of environmental conditions. The results presented in this paper summarize findings from long-term
field data, illustrating not only the impact of M. janthinus on the target weed, Dalmatian toadflax,
Linaria dalmatica (L.) P. Mill., but also on correlated plant community dynamics. These results additionally provide a valuable means to compare and contrast the biotic response and control efficacy
of this agent at both a regional and sub-continental scale.
Keywords: Linaria, efficacy, plant community response.
Introduction
Dalmatian toadflax, Linaria dalmatica (L.) P. Mill.
(Scrophulariaceae) (USDA, NRCS 2007), is an invasive short-lived perennial forb of Mediterranean origin
(Alex, 1962). Intentionally introduced to North America as an ornamental plant, L. dalmatica is now widespread and has effectively become naturalized through
multiple introductions over time (Lajeunesse, 1999).
Aspects of the species life history and morphology, including a root system characterized by a long, well-developed taproot and extensive lateral roots, dual modes
of reproduction through seed and vegetative root buds,
coupled with a high rate of seed production and long
term seed viability undoubtedly contribute to its dominance in disturbed range and forested lands (Robocker,
1974; Vujnovic and Wein, 1997).
Herbicide treatment of Dalmatian toadflax is hampered by two factors: (1) the species deep root system
Montana State University, Department of Land Resources and Environmental Sciences, P.O. Box 173120, Bozeman, MT 59717-3120, USA.
2
USDA-CSREES, 1400 Independence Avenue SW, Stop 2220, Washington, DC 202500-2220, USA.
3
USDA Forest ServiceRocky Mountain Research Station, 1648 S. 7th
Avenue, Bozeman MT 59717, USA.
Corresponding author: S.E Sing <[email protected]>.
CAB International 2008
1
necessitates precise timing of herbicide application
when root carbohydrate reserves are low and the plant
is therefore more susceptible to chemical translocation
and impact (Robocker et al., 1972) and (2) the protective waxy leaf coating resists herbicide penetration (De
Clerck-Floate and Miller, 2001). Chemical control of
Dalmatian toadflax is expensive due to the typically
large acreages affected. Additionally, repeated herbicide applications are frequently necessary in western
US water-limited habitats because each precipitation
event has the potential to stimulate Dalmatian toadflax
regeneration from fire-resistant rootstocks and characteristically large seedbanks (Zouhar, 2003).
Classical biological control of Dalmatian toadflax in North America was initiated in the late 1960s.
To date, eight exotic agent species targeting the flowers, stems, foliage or roots of Dalmatian toadflax
have been released or, in the case of adventitiously
introduced agents, redistributed, and have established
to varying degrees in North America (Smith, 1956;
Harris and Carder, 1971; Harris, 1984; De Clerck-Floate
and Harris, 2002; McClay and De Clerck-Floate,
2002).
Ten years of research results indicate that biological
control of Dalmatian toadflax is feasible, particularly
with the stem-boring weevil, Mecinus janthinus Germar, the most recently approved agent for control of
Dalmatian toadflax. M. janthinus has established and
620
Long-term field evaluation of Mecinus janthinus releases against Dalmatian toadflax in Montana (USA)
proliferated on many sites throughout western Canada
and the United States and is probably the best agent
currently available for managing Dalmatian toadflax
(Harris et al., 2000; Nowierski, 2004), although population growth is impeded by high levels of overwintering mortality (De Clerck-Floate and Miller, 2002)
and site-specific climatic factors (McClay and Hughes,
2007).
Advocates of weed biological control tout this approach for its specificity. Biological control is, in
general, a significantly more gradual process than alternative control approaches such as herbicide application. Both of these characteristics facilitate control
of the target weed without selectively or intensively
influencing immediate changes in the wider vegetation community. Reductions in invasive species such
as Dalmatian toadflax should result in the restoration
of desirable species, especially those within the target
weeds functional group, in the vegetation community.
Unfortunately, herbicide treatments frequently result in
the desired decrease in Dalmatian toadflax, followed by
either an increase in bare ground or replacement with
an undesirable species that poses an even greater environmental risk. For instance, cheatgrass (Bromus tectorum L.) and exotic, invasive knapweeds commonly
invade areas where Dalmatian toadflax has been treated
with herbicide; cheatgrass is known to significantly
alter vegetation community dynamics and fire cycles
(Whisenant, 1990; Billings, 1994), while spotted knapweed (Centaurea maculosa Lam.) is regarded as an allelopathic species (Bais et al., 2003). The purpose of
this long-term evaluation was to determine if indicators
of improved vegetation community dynamics, specifically increased cover of desirable vegetation, can be
correlated with the release of the biocontrol agent M.
janthinus.
Methods and materials
Vegetation data were recorded as early as 4 years before the release of M. janthinus (1992) and continued
through 2007 at seven release sites located throughout
Montana (see Table 1 for site-specific details). The
Bison Range site (US Fish and Wildlife Service and
Bureau of Indian Affairs) is located in the northwest
near Pablo on Flathead Indian Reservation tribal lands
adjoining the National Bison Range; three sites, Canyon Ferry (Bureau of Reclamation), Elkhorns (Bureau of Land Management) and Mount Helena (City
of Helena Parks and Recreation) are located in the
southwest, near the city of Helena, MT; the Crow site
(Bureau of Indian Affairs) is located south-centrally on
the Crow Reservation, near Lodgegrass, MT; Hardy
Bridge (Bureau of Land Management) is located south
of Great Falls, MT; and the eastern-most Melstone
site (Bureau of Land Management) is located on private ranch land bordering public lands, near the town
of the same name.
The initial releases of M. janthinus in Montana were
made on the Crow and Elkhorns sites in 1996 with limited numbers of adult weevils. M. janthinus was first
released on the Canyon Ferry and Bison Range sites
in 1997, on the Mount Helena site in 1999, and at the
Melstone and Hardy Bridge sites in 2000. Permanent,
paired 20-m vegetation monitoring transects were initially established to run through the densest local infestations of Dalmatian toadflax at each site. Vegetation data were recorded annually from fixed points at
1-m intervals along each transect. Vegetation attributes
were reported from 0.10 m2 Daubenmire frames placed
on fixed sample points, with a total of 40 samples taken
annually at each site.
Data collected from each sample frame included
counts of Dalmatian toadflax plants and individual
stems; the stems were categorized and counted as mature or immature stems. The criterion used to identify
mature stems was obvious evidence of flower buds
or actual flowering. Stem counts were taken in addition to plant densities because Dalmatian toadflax is
not easily or accurately censused on a per-plant basis
without destructive excavation. In addition, percent
cover of Dalmatian toadflax only and of all forbs other
than Dalmatian toadflax was recorded. The remaining area within the sample quadrat not accounted for
by one of the vegetation life forms was categorized
as non-vegetation cover and included the total area
covered by litter, rock and soil. For each plant parameter or vegetation category, mean values for data col
lected for the year of permanent transect establishment
were compared with those for the most recent year of
sampling (2006). Mean comparisons were made for
each site using a Students t test (Dixon and Massey,
1969).
As part of a non-destructive indicator sampling
strategy, 50 randomly selected dead stems from the
previous year were collected at each site near, but external to, the monitoring transects. Collecting dead
Dalmatian toadflax stems provides a means for evaluating M. janthinus establishment and overwintering
mortality without influencing weed or insect population trajectories within the monitoring transects. Stems
were then split with a scalpel to estimate the number
of live or emerged adult weevils. Empty M. janthinus
chambers are a reliable 1:1 indicator of successful adult
M. janthinus emergence (R. DeClerck-Floate, personal
communication). Weevil mortality was also included as
count data and encompassed non-emerged adults, larvae and pupae. Mean comparisons, with = 0.05, were
made between emerged adults and dead individuals at
each site using a Students t test (Dixon and Massey,
1969).
Results
Overwintering mortality continues to be a concern for
Montana M. janthinus populations. Estimates of dead
621
XII International Symposium on Biological Control of Weeds
Table 1.
omparison of mean pre- and post-release vegetation attributes (SE) for transect samples taken at multiple
C
Mecinus janthinus release sites.
Site
Year
Transect
Bison Range
Bison Range
P values
Bison Range
Bison Range
P values
Canyon Ferry
Canyon Ferry
P values
Canyon Ferry
Canyon Ferry
P values
Crow
Crow
P values
Crow
Crow
P values
Elkhorns
Elkhorns
P values
Elkhorns
Elkhorns
P values
Hardy Bridge
Hardy Bridge
P alues
Hardy Bridge
Hardy Bridge
P values
Melstone
Melstone
P values
Melstone
Melstone
P values
Mount Helena
Mount Helena
P values
Mount Helena
Mount Helena
P values
1992
2006
1
1
1992
2006
2
2
1996
2006
1
1
1996
2006
2
2
1993
2006
1
1
1992
2006
2
2
1992
2006
1
1
1992
2006
2
2
2002
2006
1
1
2002
2006
2
2
2000
2006
1
1
2000
2006
2
2
1992
2006
1
1
1992
2006
2
2
No. Dalmatian
toadflax stems
4.10 0.41
0.10 0.07
0.0027
5.00 0.45
0.35 0.17
0.0001
2.90 0.49
6.15 0.88
0.0100
5.10 0.69
2.45 0.41
0.0164
5.35 1.09
0.05 0.05
0.0001
7.15 1.25
0.00 0.00
0.0001
4.40 0.60
0.55 0.20
0.0283
2.80 0.59
1.35 0.44
0.4108
2.80 0.73
0.55 0.28
0.0123
3.10 0.50
0.20 0.16
0.0001
2.30 0.57
2.00 0.70
0.6923
2.40 0.39
2.40 0.54
1.0000
7.60 1.36
2.60 0.63
0.0013
9.90 1.23
1.65 0.52
0.0399
individuals per stem generally outnumbered those of
live adults that successfully emerged (Table 2). M. janthinus mortality was statistically greater than emergence
for at least 1 year of the study at all sites except Bison
Range and Hardy Bridge, while mortality was significantly lower than emergence only at Bison Range in
2003. Mortality was as much as fourfold higher than
adult emergence at certain sites in some years.
Dalmatian
toadflax % cover
Other forbs %
cover
8.80 1.11
0.50 0.34
0.0001
15.75 1.71
1.00 0.46
0.0001
8.30 2.08
12.00 1.79
0.2288
16.50 2.21
4.25 0.83
0.0001
9.80 2.25
0.25 0.25
0.0001
18.45 2.89
0.00 0.00
0.0001
12.15 2.21
2.00 0.67
0.0001
10.75 2.30
2.75 0.92
0.0001
9.50 2.35
1.00 0.46
0.0005
12.25 1.56
0.50 0.34
0.0001
11.00 2.39
5.50 1.95
0.0206
12.25 2.22
5.50 1.25
0.0013
14.95 2.09
8.75 1.58
0.0093
16.35 1.74
5.00 1.70
0.0001
5.50 1.05
14.00 1.29
0.0001
0.25 0.25
9.00 1.39
0.0001
6.25 1.73
4.25 0.41
0.1654
4.25 1.27
11.50 2.15
0.0001
2.751.11
7.00 0.76
0.1071
2.05 0.74
9.75 1.12
0.0113
0.95 0.54
5.75 0.66
0.0001
2.15 0.86
4.25 0.98
0.0435
3.75 1.02
11.50 2.57
0.0014
8.50 1.31
4.50 0.88
0.0028
12.25 1.72
7.00 0.84
0.0013
8.50 1.50
7.75 1.12
0.6038
6.35 1.21
8.75 1.49
0.2483
3.70 1.61
10.75 1.42
0.0014
Bare substrate %
cover
74.30 2.92
72.50 1.83
0.5596
49.75 3.95
81.25 1.58
0.0001
70.95 3.26
70.75 2.21
0.9629
61.25 3.70
72.75 2.42
0.0073
65.60 2.83
82.25 1.38
0.0001
61.50 2.70
73.25 1.82
0.0039
53.55 3.90
72.50 1.68
0.0001
65.00 2.55
76.25 2.20
0.0001
60.75 2.95
64.50 2.56
0.3363
58.50 1.59
72.00 1.37
0.0001
74.00 2.66
63.50 3.29
0.0145
73.25 2.67
74.50 2.35
0.7054
46.70 5.32
68.50 3.12
0.0001
43.10 4.03
76.75 2.44
0.0001
At five of seven study sites, data from both transects showed reductions in toadflax density compared
to pre-release levels (Table 1). No significant change
was recorded at the Melstone site, while transect 1 at
Canyon Ferry showed a significant increase in toadflax
density. Percent cover of Dalmatian toadflax was significantly lower at all sites with the exception of transect 1 at Canyon Ferry. We found that percent cover of
622
Long-term field evaluation of Mecinus janthinus releases against Dalmatian toadflax in Montana (USA)
Table 2.
stimates of mean (SE) number of alive and dead Mecinus janthinus per Linaria dalmatica stem, based on a
E
random sample of 50 stems collected adjacent to release monitoring transects. Alive individuals refer to empty
pupal cells, dead individuals refer to the sum of non-emerged adults, pupae and larvae. Means marked with an
asterisk are significantly different from the other in the same row at P < 0.05.
Site
Year
Bison Range
Number of individuals
2003
2004
2005
2006
2003
2004
2005
2006
2003
2004
2005
2006
2003
2004
2005
2006
2003
2004
2005
2006
2004
2005
2006
2003
2004
2005
2006
Canyon Ferry
Crow
Elkhorns
Hardy Bridge
Melstone
Mount Helena
forbs other than Dalmatian toadflax increased on 10 of
the 14 total transects, and the increase was significant
in eight cases. Non-vegetation cover increased significantly on nine transects, remained unchanged on four
transects and decreased significantly only on transect 1
at Melstone.
Discussion
The survival of M. janthinus was quite low at these
sites over the years studied, and this may have reduced
potential population growth. It has been reported that
extreme temperatures (De Clerck-Floate and Miller,
2002; McClay and Hughes, 2007) and reduced snow
cover (De Clerck-Floate and Miller, 2002) reduce survival and population growth of M. janthinus. Persistent
drought conditions in Montana from 1997 through 2004
undoubtedly affected vegetation dynamics on all sites.
The increase in unvegetated area would limit snow retention, which could increase overwintering mortality
Alive
Dead
0.96 0.18*
0.26 0.10
0.00 0.00
0.02 0.02
0.08 0.08
0.06 0.04
0.02 0.02
0.04 0.03
0.66 0.13
1.68 0.26
2.08 0.41
1.80 0.25
0.02 0.02
0.00 0.00
0.00 0.00
0.20 0.09
0.08 0.03
0.00 0.00
0.00 0.00
0.00 0.00
1.18 0.25
0.78 0.14
0.32 0.08
0.48 0.10
0.88 0.19
0.94 0.17
0.70 0.16
0.50 0.12
0.28 0.16
0.00 0.00
0.08 0.05
0.06 0.03
0.16 0.07
0.12 0.07
0.36 0.11*
1.20 0.18*
2.50 0.44*
3.64 0.52*
1.70 0.26
0.10 0.04*
0.02 0.02
0.00 0.00
0.80 0.25*
0.10 0.05
0.02 0.02
0.00 0.00
0.00 0.00
3.18 0.44*
1.06 0.18
1.44 0.23*
0.78 0.15
2.14 0.47*
1.50 0.25*
2.94 0.40*
due to lower temperatures. Although exact environmental conditions have not been determined at these
study sites, the locations do experience temperature
and precipitation patterns similar to those described for
Canada (De Clerck-Floate and Miller 2002), but percent cover of Dalmatian toadflax was lower in Montana, even before biological control was initiated.
However, the general trend towards increased cover
of forbs at these sites indicates that climatic conditions
alone were not driving the decrease in Dalmatian toadflax density and cover observed at many of our study
sites. This may be attributable to biological control altering the overall dynamics of the entire forb community by reducing competition from Dalmatian toadflax.
In addition, the increased proportion of forbs that was
correlated with an increase in unvegetated area at most
sites does suggest that, under comparative environmental extremes, biological control may offer an alternative
to protracted non-target impacts from non-selective
herbicide.
623
XII International Symposium on Biological Control of Weeds
In addition, we also observed that increased grazing pressure on toadflax appeared to be correlated with
drought conditions, especially in late summer when
the weed was one of few remaining undesiccated species available. Cattle and wildlife grazing at all sites
may have significantly impeded population buildup
of M. janthinus. Grazing of succulent stems removes
developing immature weevils from the population, resulting in a likely reduction in the following seasons
reproductive population. Weevil populations rebounded
when drought conditions eased, and adult weevils became numerous enough to be collected for redistribution on three sites in 2006, suggesting that desiccation
before and during winter has also probably been a major impediment to population increase of M. janthinus
in Montana.
Acknowledgements
This research would not have been possible without the
support of the US Forest Service, the Bureau of Land
Management, the Bureau of Indian Affairs, US Fish
and Wildlife Service, the Bureau of Reclamation and
the Crow and the Confederated Kootenai and Salish
Indian Tribes. The authors wish to acknowledge with
great appreciation Bryan FitzGerald for diligently collecting most of the data associated with this study.
References
Alex, J.F. (1962) The taxonomy, history and distribution
of Linaria dalmatica. Canadian Journal of Botany 40,
295307.
Bais, H.P., Vepachedu, R., Gilroy, S., Callaway, R.M. and
Vivanco, J.M. (2003) Allelopathy and exotic plant invasion: from molecules and genes to species interactions.
Science 301, 13771380.
Billings, W.D. (1994) Ecological impacts of cheatgrass and
resultant fire on ecosystems in the western Great Basin.
In: Monsen, S.B. and Kitchen, S.G. (eds) Proceedings,
Ecology and Management of Annual Rangelands, May
1822 1992, Boise ID, General Technical Report INTGRT-313, US Department of Agriculture, Forest Service,
Ogden UT, USA, pp. 2230.
De Clerck-Floate, R. and Miller, V. (2001) Biological control of Dalmatian toadflax in British Columbia. Agriculture and Agri-Food Canada, Lethbridge Research Centre,
Lethbridge, Canada (pamphlet, 4 pp).
De Clerck-Floate, R.A. and Harris, P. (2002) 72. Linaria dalmatica (L.) Miller, Dalmatian toadflax (Scrophulariaceae).
In: Mason, P.G. and Huber, J.T. (eds) Biological Control
Programmes in Canada, 19812000. CABI, New York,
pp. 368374.
De Clerck-Floate, R. and Miller, V. (2002) Overwintering
mortality of and host attack by the stem-boring weevil,
Mecinus janthinus Germar, on Dalmatian toadflax (Linaria dalmatica (L.) Mill.) in western Canada. Biological
Control 24, 6574.
Dixon, W.J. and Massey, Jr., F.J. (1969) Introduction to Statistical Analysis, 3rd edn. McGraw Hill, New York, 488 pp.
Harris, P. (1984) Linaria vulgaris Miller, yellow toadflax
and L. dalmatica (L.) Mill., broad-leaved toadflax. In:
Kelleher, J.S. and Hulme, M.A. (eds) Biological Control Programmes Against Insects and Weeds in Canada,
19691980, CAB, Farnham Royal, UK, pp. 179182.
Harris, P. and Carder, A.C. (1971) Linaria vulgaris Miller,
yellow toadflax and L. dalmatica (L.) Mill., broad-leaved
toadflax. In: Kelleher, J.S. and Hulme, M.A. (eds) Biological Control Programmes Against Insects and Weeds in
Canada, 19591968. Commonwealth Institute of Biological Control, Technical Communication No. 4, pp. 9497.
Harris, P., De Clerck-Floate, R. and McClay, A. (2000) Mecinus
janthinus Germar. Stem boring weevil. Available at: http://
res2.agr.ca/lethbridge/weedbio/agents/amecinus.htm.
Lajeunesse, S.E. (1999) Dalmatian and yellow toadflax. In:
Sheley, R.L. and Petroff, J.K. (eds) Biology and Management of Noxious Rangeland Weeds. Oregon State University Press, Corvallis, OR, USA, pp. 202216.
McClay, A.S. and Hughes, R.B. (2007) Temperature and
host-plant effects on development and population growth
of Mecinus janthinus (Coleoptera: Curculionidae), a biological control agent for invasive Linaria spp. Biological
Control 40, 405410.
McClay, A.S. and De Clerck-Floate, R.A. (2002) Linaria vulgaris Miller, yellow toadflax (Scrophulariaceae). In: Mason,
P.G. and Huber, J.T. (eds) Biological Control Programmes
in Canada, 19812000. CABI, New York, pp. 375382.
Nowierski, R.M. (2004) Mecinus janthinus. In: Coombs, E.M.,
Clark, J.K., Piper, G.L. and Confracesco, Jr., A.F. (eds)
Biological control of invasive plants in the United States.
Oregon State University Press, Corvallis, OR, USA, pp.
392395.
Robocker, W.C. (1974) Life history, ecology, and control of
Dalmatian toadflax. Technical Bulletin 79. Washington
Agricultural Experiment Station, Washington State University, Pullman, WA, USA, 20 pp.
Robocker, W.C., Schirman, R. and Zamora, B.A. (1972) Carbohydrate reserves in roots of Dalmatian toadflax. Weed
Science 20, 212214.
Smith, J.M. (1956) Biological control of weeds. In: Vinoth, J.
(ed) 1955 Proc. 9th National Weed Committee (Eastern
sect.). National Weed Committee, Canada Department of
Agriculture, Ottawa, ON, Canada, pp. 102103.
USDA, NRCS. (2007) The PLANTS Database. National
Plant Data Center, Baton Rouge, LA, USA. Available at:
http://plants.usda.gov (accessed 8 February 2007).
Vujnovic, K. and Wein, R.W. (1997) The biology of Canadian weeds. 106. Linaria dalmatica (L.) Mill. Canadian
Journal of Plant Science 77, 483491.
Whisenant, S.G. (1990) Changing fire frequencies on Idahos
Snake River Plains: ecological and management implications. In: McArthur, E.D., Romney, E.M., Smith, S.D.
and P.T. Tueller (compilers) Proceedings Symposium on
Cheatgrass Invasion, Shrub Die-Off and Other Aspects
of Shrub Biology and Management, April 57 1989, Las
Vegas, NV. Technical Report INT-276, Department of Agriculture, Forest Service, Intermountain Research Station,
Ogden, UT, USA, pp. 410.
Zouhar, K. (2003) Linaria spp. In: Fire Effects Information
System [Online]. US Department of Agriculture, Forest
Service, Rocky Mountain Research Station, Fire Sciences
Laboratory. Available at: http://www.fs.fed.us/database/
feis/plants/forb/linspp (accessed 27 February 2007).
624
Post-release evaluation of invasive plant
biological control agents in BC using IAPP,
a novel database management platform
S.C. Turner
Summary
When faced with the invasion of alien plant species, the government of British Columbia (BC) uses
the principals of Integrated Pest Management: prevention, early detection rapid response, inventory
surveys and mechanical, chemical and biological control treatments. The BC governments Invasive
Alien Plant Program (IAPP) application houses the data for management of invasive alien plants,
including planning, inventory, mechanical, chemical and biological control, the monitoring of each
of these activities and biological control agent dispersal. The IAPP Application is structured to track
sites and their characteristics as geographic locations. Multiple invasive species with multiple surveys
can be entered for a single site. This allows recording of the change in the invasive plant community
over time and the level of success of treatment efforts. A compilation of this data allows assessment
of the current set of biological control agents for target plant species. By comparing the spread of
Dalmatian toadflax, Linaria dalmatica L. Miller, to the habitat requirements of its biological control
agents, it is possible, using IAPP, to determine whether sufficient, suitable agents exist in the province
or whether additional screening of agents must be pursued.
Keywords: Dalmatian toadflax, biological control agent habitat requirements, spatial
mapping.
Introduction
The Crown public lands of British Columbia (BC)
consist of approximately 885,600 km2 or 93% of the
province (BC Ministry of Sustainable Resource Management, 1997). When faced with the invasion of alien
plant species, BC employs the principals of Integrated
Pest Management: prevention, early detection rapid
response (EDRR), inventory surveys and a multitude
of treatment tools including mechanical, chemical and
biological control. For example, upon initial sightings
of an invasive alien species, mechanical treatment may
be used when the infestation is of a size to be managed
by this method or when the plants reside in a herbiciderestricted area. Herbicide treatment is a necessary and
effective tool that is used judiciously.
British Columbias legislation for herbicide use,
administered and enforced by the Ministry of Envi Ministry of Forests and Range, 515 Columbia Street, Kamloops, BC,
Canada V1S 1E6.
Corresponding author: S. Turner <[email protected]>.
CAB International 2008
ronment, is protective of the natural environment. For
example, the pesticide-free zone required around or
along bodies of water, dry streams and classified wetlands is 10 m, and BC has no herbicides registered for
use against invasive plants in water. When an invasive alien species has spread to an area where repeated
spraying is not economical, a containment strategy is
employed. Under this strategy, a perimeter boundary
around an infestation is established, and no mechanical or chemical treatment is carried out within the
perimeter except in special circumstances. Any invasive plant found beyond the boundary is aggressively
treated.
Once an invasive species has spread beyond the
ability to manage it with mechanical or chemical treatment methods, the area is considered to require restoration. Restoration of infested areas is pursued through
the use of good resource management practices and
the encouragement of healthy biological control agent
populations, where available.
BC Ministry of Forests and Range (MFR) staff
interact with scientists in the Canadian and international community pursuing biological control agents
625
XII International Symposium on Biological Control of Weeds
for invasive alien plant species of concern. Consortia
are formed, comprised of funding agencies, national
researchers and influential stakeholders that plan and
pursue biological control research. The invasive plants
country of origin is investigated for insects and pathogens that are found to attack the target plant species.
Through cooperative efforts, BC funds research to ensure that biological control agents are specific to the
target plant, that is, they will not inflict major damage
to other Canadian plants of economic or environmental
importance.
When new biological control agents have passed
the screening process and approval for importation has
been obtained, they are shipped into the country, passed
through the Agriculture and Agri-Food Canada quarantine facility at Lethbridge, Alberta and received in
British Columbia. Once in British Columbia, the agents
are released at pre-determined sites for optimum establishment. These sites are determined from information
on the agents native habitat, investigation into the
provincial governments Invasive Alien Plant Program
(IAPP) application and field investigation. The release
sites are in turn tracked in the IAPP Application.
The IAPP Application (http://www.for.gov.bc.ca/
hfp/invasive/index.htm), launched in fall 2005, is a
web-based Oracle application that houses data associated with the integrated weed management processes.
Information is entered into the IAPP application for
individual sites including their geographical characteristics and invasive plant species. Data on multiple
invasions from surveys that assess the management
strategies can be included for each site. This creates a
record of the change in the invasive plant populations
over time and allows evaluation of the level of success
of management practices.
The IAPP application consists of two components:
a password-protected Data Entry module and a Map
Display module that is accessible to the public. Collaborating organizations may be given access to IAPP
to record information. The information is visible to all
participants but is protected from manipulation. Efforts
to manage invasive alien plants can, therefore, be coordinated among all participants.
For applied biological control, sites are chosen where
there is a minimal chance of conflicting treatments occurring and after considering habitat-requirement information. The latter information is obtained from
screening and petition documents, field staff and from
the Biogeoclimatic Ecosystem Classification (BEC) system, which is viewed as a layer in the IAPP application.
The BEC (http://www.for.gov.bc.ca/hre/becweb/) system groups ecosystems into hierarchical classifications.
A unit within the BEC system is defined as a particular
plant community and its associated physiography, soil
and climate (Meidinger and Pojar, 1991).
Biological control agent releases are recorded into
the IAPPs Data Entry module. The sites are viewed
spatially in IAPPs Map Display as polygons of vari-
ous sizes to represent, from small to large: a Universal
Transverse Mercator (UTM) co-ordinate; a protected
location for biological control agents that have yet to
establish; and the size and shape of an invasive plant
infestation. Thereafter, a site is monitored to determine
if the agent has established and whether any change has
occurred in the density, area or distribution of the target
invasive plant.
Additionally, the IAPP application can house biological control agent dispersal information. Dispersal
information is used to determine locations for the collection and redistribution of biological control agents.
Altogether, the spatial information displayed in the
IAPP application allows for an increased understanding of the habitat requirements of particular biological
control agents. Predictions can be made of the agents
ability to colonize different habitat types and locations.
This can give invasive plant managers and scientists direction for pursuing new biological control agents and
the habitat types they are required to fill. An example
of how the IAPP application is being used can be seen
in the case of the invasive Dalmatian toadflax, Linaria
dalmatica L. Miller, and its complement of biological
control agents, in particular the stem weevil, Mecinus
janthinus Germar, and the seed weevil, Rhinusa antirrhini (Paykull).
Dalmatian toadflax is a short-lived perennial herb
that was introduced as an ornamental in the United
States in 1894. It originates from the Mediterranean region, from Yugoslavia to Iran (Robocker, 1974). Dalmatian toadflax spreads by creeping root stock and seeds,
each plant potentially producing up to 500,000 seeds
(Robocker, 1974) that are dispersed mainly by wind
and browsing animals. Mature plants are 60 to 120 cm
tall. The stems, several per plant, are smooth and lightgreen, and the flowers are snapdragon shaped (Powell
et al., 1994). The plant is toxic to livestock, and cattle
tend to avoid grazing on infested pastures (Mason and
Huber, 2002).
Dalmatian toadflax is a stress-tolerant plant able to
grow in conditions of low temperature, coarse soils
and summer drought. It exists in the BC BEC zones
of Bunchgrass, Ponderosa pine, Interior Douglas-fir,
Interior cedar-hemlock, Coastal Douglas-fir, Coastal
Western hemlock, Montane spruce and the Sub-boreal
spruce.
Efforts to acquire biological control agents for BC
Dalmatian toadflax began in the 1960s (Mason and
Huber, 2002). Since then, seven agents have been
screened, and these have been released or have arrived
adventively in the province. In particular, M. janthinus (stem-boring weevil) was introduced in 1991, and
R. antirrhini (seed-feeding weevil) was introduced in
1993. These two species are actively being redistributed across the province.
In its native (European) distribution, M. janthinus
was recorded residing in a wide range of habitat types.
It has been recorded from just below the subalpine zone
626
Post-release evaluation of invasive plant biological control agents in BC using IAPP
in the Alps to the maritime lowlands in western and
northern France and northern Germany to the Mediterranean climate of the Rome area in Italy and to the
subcontinental, summer-dry regions of eastern and
southern Yugoslavica and south-western Russia (Jeanneret and Schroeder, 1992). It is also believed to exist in
other parts of southern Germany, in Austria, Hungary
and the Balkans (Jeanneret and Schroeder, 1992).
Based on its native distribution, M. janthinus was
expected to establish in all habitats where yellow toadflax, Linaria vulgaris L., and Dalmatian toadflax exist in
North America between the latitudes of 40 and 52. In
Canada, this would include south-central BC, southern
Alberta and Saskatchewan, as well as the maritime areas
in eastern Canada. In the United States, this would include Washington, Oregon, Montana, northern California
and the maritime areas of the eastern USA (Jeanneret and
Schroeder, 1992). It has also been stated by Powell et al.
(1994) that M. janthinus prefers hot, dry conditions usually found in grasslands or open forest with grasslands.
Typically, M. janthinus attacks Dalmatian toadflax by
ovipositing single eggs into holes that the female chews
in the stem. The holes are sealed and, in turn, covered by
callouses that appear as tiny round blemishes on the stem
(Jeanneret and Schroeder, 1992). Multiple eggs may be
laid into a single stem, and the larvae feed in the stem,
disrupting the flow of nutrients and causing the cessation
of flowering. More than 100 weevils have been found
in a single large stem (Mason and Huber, 2002). This
activity takes place from spring until mid-summer. Infestations of Dalmatian toadflax are noticeably affected
by large populations of M. janthinus, for example in the
Lac du Bois grassland park near Kamloops, BC, where
the weevil was released in 1997 (Figs. 1 and 2).
The native distribution of R. antirrhini is recorded as
central and southern Europe, the Mediterranean region
and the Caucasus (Groppe, 1992). Without in-depth
knowledge of the insects habitats in these regions, it
Figure 1.
is difficult to determine which habitats in BC might be
conducive to either or both agents.
Females of R. antirrhini oviposit into the flower carpel in June (Groppe, 1992). Hence, it is necessary to
locate areas for release of R. antirrhini that have not
had flowering eliminated due to damage caused by M.
janthinus. This has become difficult because M. janthinus has widely established and is spreading naturally,
as well as being manually redistributed.
Methods
The initial introduction of M. janthinus involved placement of the agent into propagation tents. Its redistribution to open field sites began in 1994. Adults of
R. antirrhini were placed in the propagation tents in
1993, and open releases began in 1999. As little was
known about the habitat requirements of R. antirrhini,
efforts have been made to release R. antirrhini into
temperate to mild habitats, with the hope of ensuring
establishment. Both agents were subsequently redistributed across the province, and data describing host
plant populations, biological control agent releases,
and monitoring for weevil establishment, habitat preference and dispersal are being recorded.
Extensive field work has been and continues to be
carried out to gather data about Dalmatian toadflax and
its management. Historical data were incorporated into
IAPP upon its launch (2005). Since then, data has been
added as management activities and evaluations were
carried out.
Site data in IAPP is displayed spatially using dots
and polygons to represent sites where the biological
control agents have been released. With a baseline of
habitat information, additional environments can be
tested for habitat preferences by over-laying the spacial information with habitat information, such as the
BEC zone data.
Figure 2.
Release site for Mecinus janthinus
on Dalmatian toadflax, 2004, before
control.
627
Release site for Mecinus janthinus
on Dalmatian toadflax, 2006, after
control.
XII International Symposium on Biological Control of Weeds
Figure 3.
Release sites (black dots) between 1991 and 1999 and redistribution sites (open dots) to
2006 for Mecinus janthinus released onto Dalmatian toadflax in British Columbia.
628
Post-release evaluation of invasive plant biological control agents in BC using IAPP
Figure 4.
Release sites (yellow dots) between 1993 and 1999 and redistribution sites (grey dots) to
2006 for Rhinusa antirrhini on Dalmatian toadflax overlaid on the release and redistribution sites for Mecinus janthinus (Fig. 3).
629
XII International Symposium on Biological Control of Weeds
IAPP has been used to select new redistribution
sites, track dispersal of biological control agents and
evaluate suitable habitats for establishment. Upon evaluation of IAPP data, sites have been selected that are
free of (or have low populations of) M. janthinus for
R. antirrhini redistribution.
Results
The two weevils, M. janthinus and R. antirrhini, have
been redistributed to 719 and 34 sites, respectively, by
autumn 2006. In the field, M. janthinus has exceeded its
predicted distribution in British Columbia. For example, a pair of M. janthinus weevils was found at Terrace
1 year after the agents release at this location, and 2
years after the release evidence of characteristic stemmining was observed. The northern city of Terrace is
5430N and approximately 280 km further north than
the predicted 52N limit for this insect. Terraces climate is tempered by its proximity to the Pacific Ocean,
and its climate is described as the BEC zone Coastal
Western Hemlock, sub-montane, wet sub-maritime. It
has a rather mild climate compared to the surrounding
areas.
Another city, Williams Lake, 5208N, is at the
northern edge of the predicted limit for M. janthinus.
This interior city is influenced by the cold climate of the
coastal mountain range and the open Cariboo Plateau.
However, as it is located next to a large lake its climate
is also tempered. It is described as the BEC zone Interior Douglas-fir, very dry, mild. A thriving M. janthinus
population at Williams Lake has yielded thousands of
weevils for redistribution. Additionally, M. janthinus
has had a significant impact on its host plant at many of
the release sites there.
Populations of R. antirrhini have been found to establish in all BEC zones in which this species has been
released, including the BEC zones Bunchgrass, Ponderosa pine, Interior Douglas-fir, Interior cedar-hemlock
and Montane spruce. The lowest and highest recorded
elevations were 290 and 1205 m, respectively.
Discussion
The process of redistribution and monitoring of biological control agents continues in the attempt to control Dalmatian toadflax. This objective is complicated
by the number of different agencies and organizations
conducting field work and the natural dispersal of the
agents. The use of IAPP to record release and dispersal
information is a recent development; thus, the data are
just beginning to accumulate (Fig. 3), and preliminary
results have been generated (Fig. 4).
When initial releases were monitored and the agents
were found to have established, new, more extreme
environments were tested, and dispersal information
was recorded. The habitat requirements for the individual biological control agents then became more clear.
When the goal of a biological control program is to
have the invasive alien plant species under control by
a long-term, self-sustaining system, it is critical to understand the habitat requirements and natural dispersal
of the biological control agents.
Eventually, the compilation and evaluation of the
data contained within the IAPP application should allow a comprehensive assessment of the complex of
biological control agents for individual target plant
species. The spread of Dalmatian toadflax in BC is
likely to continue until the plant reaches its ecological
limits. Using IAPP, comparison of the data describing
this spread with the habitat-requirement data of its biological control agents will enable weed managers to determine whether sufficient, suitable agents exist in the
province, where further redistribution of such agents
is required and whether introduction of new agents
should be pursued.
References
Groppe, K. (1992) Final Report Gymnetron antirrhini Paykull
(Col.: Curculionidae). A Candidate for Biological Control
of Dalmatian Toadflax in North America. CABI European
Station, Delemont, Switzerland, 22 pp.
Jeanneret, P. & Schroeder, D. (1992) Biology and Host Specificity of Mecinus janthinus Germar (Col.: Curculionidae),
a Candidate for the Biological Control of Yellow and Dalmatian Toadflax, Linaria vulgaris (L.) Mill. and Linaria
dalmatica (L.) Mill. (Scrophulariaceae) in North America.
CABI, European Station, Delemont, Switzerland, 34 pp.
Mason, P.G. & Huber, J.T. (eds) (2002) Biological Control
Programmes in Canada, 19812000. CAB International,
Oxon, UK, 583 pp.
Meidinger, D. & Pojar J. (1991) Ecosystems of British Columbia. British Columbia Ministry for Forests, Research
Branch, 330 pp.
Powell G., Sturko, A., Wikeem, B. & Harris, P. (1994) Field
guide to the biological control of weeds in British Columbia. Land Management Handbook Number 27. British
Columbia Ministry for Forests, Research Branch, 163 pp.
Robocker, W.C. (1974) Life History, Ecology, and Control of
Dalmatian Toadflax. Washington Agricultural Experiment
Station, Washington State University, USA, 20 pp.
630
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Monitoring of ground cover post release of
Aphthona nigriscutis near Lander, Wyoming
J.L. Baker and N.A.P. Webber
Fremont County Weed and Pest Control District, Room 315, 450 North 2nd Street, Lander, WY 82520, USA
In 1990, Aphthona nigriscutis was released for the control of leafy spurge (Euphorbia esula) at
20 locations near Lander, WY. Establishment at several locations by 1993 prompted a post-release
monitoring effort. Permanent transects were established at many locations. Eight sites are still usable,
the rest being lost to development or herbicide use. At four of these sites, five 50-ft (15.25-m) transects
radiate out from the initial point of release at approximately 70 intervals. Ground cover was measured
at 5-ft (1.52-m) intervals along each transect by using a ten-pin point frame, recording the first contact
with each pin. At the other four sites, there were eight radial transects, 100 ft (30.5 m) long, with data
collected at 10-ft (3.04-m) intervals. The sites differ by slope, aspect, soil type and moisture. Data has
been collected annually at the peak of vegetative production in late July or early August. Since 1993,
leafy spurge has declined from 52% of the ground cover to 11% across all eight sites. Perennial grasses
and forbs have increased, while bare ground has declined.
Benefits to New Zealands native flora from the successful
biological control of mistflower (Ageratina riparia)
J. Barton1 and S.V. Fowler2
Landcare Research, Private Bag 92170, Auckland, New Zealand
2
467 Rotowaro Road, RD 1, Huntly 3771, New Zealand
A white smut fungus (Entyloma ageratinae) and a gall fly (Procecidochares alani) were released in
New Zealand to suppress the invasive weed mistflower (Ageratina riparia). The impacts of the agents
on the target weed and surrounding vegetation were monitored over 5 years. The fly was released 3
years after the fungus and spread much more slowly, so most of the impacts observed are attributable
to the fungus alone. Nonetheless, the number of stem galls produced by the insect increased exponentially to reach damaging levels at release sites. Annual monitoring of study plots revealed that the mean
percentage of leaves infected by the fungus reached 58%, and there was a significant decrease in the
maximum height of mistflower plants. In heavy infestations, the mean percentage cover of mistflower
declined from 81% to 1.5%. As the weed declined, the mean species richness and percentage cover of
native plants increased to approach that of areas without mistflower. In contrast, there was no significant change in the species richness or percentage cover of exotic plants (excluding mistflower). There
was a weak replacement weed effect from the invasive African club-moss (Selaginella kraussiana),
but mostly, the decline in mistflower benefited indigenous plants, including two rare endemic Hebe
species.
CAB International 2008
631
XII International Symposium on Biological Control of Weeds
Tracking population outbreaks: impact and quality of
Aphthona flea beetles on leafy spurge at two spatial scales
R.S. Bourchier
Agriculture and Agrifood Canada, Lethbridge Research Centre, 5403 1st Avenue South,
Lethbridge, AB, Canada T1J 4B1
As classical biocontrol agents go through their first outbreak cycle, the impact, natural dispersal and
quality of insects available for redistribution become important issues. Aphthona lacertosa populations were monitored at 21 sites for 6 years in Alberta, Canada. Impact and weed population changes
were assessed initially at a release stake-scale (10-m circle) and then at the scale of the release patch.
Adult beetles were sampled between three and four times per year, and stem counts and release patch
perimeters were assessed annually using a global positioning system. Releases that occurred in 1997
resulted in 100% establishment in 1998 and visible damage (halos of dead spurge) at 81% of the release
sites. First outbreak populations of beetles occurred in 2000, with maximum reductions in spurge density approaching 100% at some locations. Spurge populations recovered in the drought year of 2001.
In subsequent years, many sites experienced a re-infestation with spurge seedlings, although beetles
were still present. Beetle quality attributes, including sex ratios, size and egg loads, were measured at
each site during population outbreaks. Sex ratios were even and relatively consistent, whereas beetle
size and egg loads varied within the season, which may influence the spread and establishment of new
sites.
Are nutrients limiting the successful biological control of
water hyacinth, Eichhornia crassipes, in South Africa?
R. Brudvig,1 M.P. Hill,2 M. Robertson3 and M.J. Byrne1
University of the Witwatersrand, School of Animal, Plant and Environmental Sciences,
Johannesburg 2050, South Africa
2
Rhodes University, Department of Entomology, Grahamstown, South Africa
3
University of Pretoria, Department of Zoology, Pretoria, South Africa
Water hyacinth remains a problem aquatic weed in South Africa and throughout the world due to eutrophic waterbodies. Previous work correlated water hyacinth growth with nitrogen and phosphorus
in particular. Our field sampling tested the trade off between nutrient levels and biological control
by country-wide monitoring of 14 water-hyacinth-infested sites. Water quality, insect population parameters and growth of water hyacinth plants was measured monthly for 24 months. High levels of
inorganic nitrogen (7.5 mg/l) and phosphates (1.37 mg/l) caused a significant increase in petiole length,
ramet production and biomass. Sites with limiting levels of these nutrients resulted in no or a reduced
amount of growth. Higher levels of insect damage were recorded at eutrophic sites due to higher nitrogen content in the leaf tissue (5% N). Insect damage was still significant at nutrient limited sites where
biological control has been successful. We conclude that eutrophication decreases the successful biological control of water hyacinth. Nutrients override the effects of temperature, and biological control
has achieved more success on water bodies that are nutrient limited due to reduced plant growth.
632
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Spatial evaluation of weed infestation and bioagent
efficacy: an evolution in monitoring technique
V.A. Carney, G.J. Michels Jr and D. Jurovich
Texas A&M System, Texas Agricultural Experiment Station, 2301 Experiment Station Road,
Bushland, TX 79012, USA
Traditional invasive weed and insect biological control agent monitoring can be time consuming, costly and labour-intensive. Practices involving linear transect sampling provide relatively low return on
a monitoring investment, particularly when trying to understand target plant and bioagent spread and
establishment. Global positioning system and geographical information system (GIS) are tools that are
rapidly replacing tape measures and topographic maps. Over the past decade, electronic data collection
and analysis techniques have evolved along with the widespread availability of such technologies. Information is now gathered in a spatially relevant manner to identify key environmental attributes contributing to the efficacy of our biocontrol efforts. This presentation will contrast information typically
obtained from traditional transect vegetation sampling methods against data collected and evaluated
using GIS. Weed and insect population patterns that are readily apparent using a spatial approach, such
as hotspots and patchiness, edge effects and directional trends, are generally unseen using traditional
data collection techniques. GIS analysis allows for an improved statistically sound method of analysing biocontrol efficacy. It also facilitates a much more predictive approach to biocontrol agent release
by utilizing spatial modelling techniques. These concepts will be discussed using field-collected data
on the biological control of knapweeds, leafy spurge and Dalmatian toadflax.
Influence of release size on the establishment and
impact of a biocontrol root weevil
R. De Clerck-Floate
Agriculture and Agri-Food Canada, Lethbridge Research Centre, P.O. Box 3000,
Lethbridge, AB, Canada T1J 4B1
A strategic approach to classical weed biocontrol implementation is emerging through recent efforts to
predict success in agent population establishment and persistence based on initial release size. However, there has been little formal investigation of whether release size can also predict agent efficacy.
A root-boring weevil, Mogulones cruciger, which was introduced in Canada to control houndstongue
(Cynoglossum officinale), was used to test this relationship. Numbers of 0, 100, 200, 300 or 400 weevils were field-released within discrete, isolated houndstongue patches (five replicates per treatment).
Patches were monitored over 2 years for weevil establishment, host attack and changes in houndstongue populations. The weevil established successfully in all treatment patches, regardless of release
size. Release size was positively correlated with weevil numbers and damage to host plants. In contrast, the different release sizes reduced houndstongue populations by the same amount and at the same
rate relative to control patches during the 2-year study. Thus, all release sizes tested could predictably
achieve patch-level control of houndstongue despite differences in the level of feeding. The measurable and predictable impact of the agent on weed populations in this system are mainly attributed to
reliable agent establishment and rapid kill of host individuals.
633
XII International Symposium on Biological Control of Weeds
Development of Mycoleptodiscus terrestris as a
biological control agent of Hydrilla
C.A. Dunlap and M. Jackson
USDA-ARS, National Center for Agricultural Utilization Research, Crop Bioprotection Research Unit,
1815 North University Street, Peoria, IL 61604, USA
In aquarium studies, Mycoleptodiscus terrestris has previously been shown to be effective in controlling the invasive aquatic weed, Hydrilla verticillata. The development of M. terrestris into a commercially viable control option requires methods to produce, formulate and disperse infective propagules.
Our current research addresses these problems at the pilot plant scale. Microsclerotia of M. terrestris
are produced in liquid culture fermentation, processed and dried for enhanced survival. Methods and
handling parameters have been developed to produce uniform-sized particles, which are amenable to
the drying process. The drying process has been optimized to provide stabile propagules with a long
shelf-life. In addition, analysis of the chemical composition of the microsclerotia suggests that the
membrane stabilizers, trehalose and mannitol, are important to its drying tolerance.
Molecular characterization of Striga mycoherbicides
Fusarium oxysporum strains: evidence for
a new forma specialis
A. Elzein1, M. Thines,2 F. Brndle,2 J. Kroschel,3 G. Cadisch1 and P. Marley4
University of Hohenheim, Institute for Plant production and Agroecology in the Tropics and Subtropics (380),
70593 Stuttgart, Germany
2
University of Hohenheim, Institute of Botany (210), 70593 Stuttgart, Germany
3
International Potato Center (CIP), Integrated Crop Management Division, Av. La Molina 1895,
Apartado 1558, Lima 12, Peru
4
Ahmadu Bello University, Faculty of Agriculture/Institute for Agricultural Research,
Department of Crop Protection, Samaru, Zaria, Nigeria
Fusarium oxysporum isolates (Foxy 2 and PSM197) are potential, highly host-specific mycoherbicides
for the control of the parasitic weeds Striga hermonthica and Striga asiatica, which are major biotic
constraints in cereal in semi-arid tropical Africa. The fungal isolates were cultivated on potato dextrose
agar medium and characterized based on partial DNA sequence of the internal transcribed spacer (ITS)
regions of the nuclear ribosomal RNA gene. The ITS sequence was obtained using the universal primers ITS1 and ITS4. Both isolates were identical in ITS sequence. The ITS sequence obtained was not
identical to any ITS sequence deposited in GenBank. By sequencing a non-coding, non-genic region
of the samples, it was possible to differentiate between the strains Foxy2 and PSM197. In addition,
primers for the detection of both isolates were developed. These primers were able to amplify sequence
stretches from both strains with high sensitivity and without false-positive reactions to other Ascomycetes tested. These findings are helpful in monitoring the establishment, spread and persistence of Striga
mycoherbicides in the soil, for quality control of mycoherbicide products and for proving their environmental biosafety by being able to distinguish them from wild, crop-pathogenic strains. The unique
ITS sequence of the two isolates provides strong evidence to consider these pathogens of Striga as a
new forma specialis (f.sp. strigae) that will facilitate and encourage the introduction and acceptance by
regulatory authorities and farmers of Striga mycoherbicides for practical field application.
634
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Prioritizing candidate biocontrol agents for garlic mustard
based on their potential effect on weed demography
E. Gerber,1 H. Hinz,1 D.A. Landis,2 A.S. Davis,3 B. Blossey4 and V. Nuzzo5
CABI Bioscience Switzerland Centre, 2800 Delmont, Switzerland
Michigan State University, Department of Entomology, East Lansing, MI 48824, USA
3
USDA-ARS Invasive Weed Management Unit, Urbana, IL 61801, USA
4
Cornell University, Department of Natural Resources, Ithaca, NY 14853, USA
5
Natural Area Consultants, Richford, NY 13835, USA
1
To reduce the possibility for non-target effects, biological weed control programs should select and
introduce the minimum number of host-specific natural enemies necessary to control an invasive
non-indigenous plant. However, selection of the best agent or agent combination is no easy task and
depends on the ability to forecast the anticipated impact of each herbivore species on host-plant demography. In a project on the biological control of garlic mustard [Alliaria petiolata (M. Bieb.) Cavara
and Grande] in North America, we experimentally investigated the impact of candidate agents on survival and reproductive output of the target plant. Results were combined and fed into a demographic
model to explore the potential impact of each agent at the plant population level. Using these a priori
analyses, we propose potential release strategies for the candidate agents in North America.
The accidentally introduced Canada thistle mite
Aceria anthocoptes in the western USA:
utilization of native Cirsium thistles?
R.W. Hansen
USDAAPHISPPQ, National Weed Management Lab, Suite 108, 2301 Research Boulevard,
Fort Collins, CO 80526, USA
Aceria anthocoptes (Acari: Eriophyidae) is a Eurasian mite that feeds on leaves and stems of Canada
thistle, Cirsium arvense. A. anthocoptes was introduced in North America and now appears to be
widely established, at least in the northern USA, where Canada thistle is a widespread exotic weed.
A. anthocoptes is one of two Aceria mites recorded from C. arvense in Europe; no eriophyid mites had
previously been collected from Canada thistle in the United States. In 2005 and 2006, we studied the
biology of A. anthocoptes at several Canada thistle sites in northern Colorado, USA. We also sampled
populations of four native thistles (Cirsium canescens, Cirsium scariosum, Cirsium scopulorum and
Cirsium undulatum) to detect mite populations. These four thistles occur in grassland, foothills and
montane habitats in Colorado and often grow in close proximity to Canada thistle. We collected eriophyid mites from all native plants; mite densities were similar to, or even exceeded, densities from adjacent Canada thistle populations. For now, we are presuming that collected mites are A. anthocoptes,
though taxonomists are confirming identifications. If specimens are determined to be A. anthocoptes,
our information should preclude the use of this mite as a C. arvense biocontrol agent in the United
States.
635
XII International Symposium on Biological Control of Weeds
Formulation of Colletotrichum truncatum into
complex coacervate biocontrol of scentless
chamomile, Matricaria perforata
R.K. Hynes, P. Chumala, D. Hupka and G. Peng
Agriculture and Agri-Food Canada, 107 Science Place, Saskatoon, SK, Canada S7N 0X2
Colletotrichum truncatum (schwein.) Andrus and W. D. Moore is a phytopathogenic fungus to scentless chamomile, Matricaria perforate Mrat, a noxious weed in western Canada. High virulence and
host specificity of the fungus towards scentless chamomile allowed considering it as a potential candidate for weed biocontrol. Microencapsulation of C. truncatum conidia in a complex coacervate has
been investigated. Complex coacervates have been widely used as a microencapsulation technique for
oil-dispersible active ingredients in the pharmaceutical and food industries. Conidia of C. truncatum
are hydrophilic and oil-indispersible; therefore, an initial formulation step of suspending them in a
water/oil invert emulsion was required before encapsulating in a complex coacervate. To maximize
percent conidial encapsulation, parameters such as wall materials, i.e. protein-polysaccharide, stirring
speed, surfactants, and conidial suspension to oil ratios, were optimized. Weed control efficacy of the
formulations was determined on scentless chamomile at the six to eight leaf stages both under greenhouse and field conditions. In addition, the synergistic effect between the C. truncatum formulation
and the herbicide Sencor was evaluated. In this paper, our new approach to formulate C. truncatum
in a complex coacervate and the efficacy assay results are presented, and implications for controlling
scentless chamomile are discussed.
Efficacy of the seed feeding bruchid beetle, Sulcobruchus
subsuturalis, in the biological control of Caesalpinia
decapetala in South Africa
F.N. Kalibbala, E.T.F. Witkowski and M.J. Byrne
University of the Witwatersrand, School of Animal Plant and Environmental Sciences, Private Bag 3,
WITS, 2050 Johannesburg, South Africa
The release of the seed beetle, Sulcobruchus subsuturalis, for biological control of Caesalpinia decapetala in South Africa has been ongoing since 1999. We assessed seed dispersal and the effects of
S. subsuturalis on seed banks, seedling recruitment and seed viability within five sites. Seeds were not
dispersed beyond 10 m of the parent plant. This partly reduces the chances of seeds reaching suitable
microsites and partly facilitates high seed mortality near the parent plant given that high beetle densities ensure that many seeds are colonized. On average, no site contained more than 5 seeds/m2 because
the beetle attack destroys seeds. Therefore, the seed bank is very small, possibly not lasting a year. One
seedling established per square metre on average within all sites, suggesting that S. subsuturalis may
have no effect on the population dynamics of C. decapetala despite the great impact on seed viability;
overall, 93.7% of predated seeds were not viable. As seed-feeding agents alone rarely reduce weed
density, additional agents should be released to attain successful control of this weed.
636
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Field studies of the biology of the moth,
Bradyrrhoa gilveolla, as a potential biocontrol
agent for Chondrilla juncea
J. Kashefi,1 G.P. Markin2 and J.L. Littlefield3
USDA-ARS, EBCL Substation, Thessaloniki, Greece
US Forest Service, RMRS, Forest Science Laboratory, Bozeman, MT, USA
3
Montana State University, Department of LRES, Bozeman, MT, USA
1
The root-attacking moth, Bradyrrhoa gilveolla, has been released as a potential biocontrol agent of
Chondrilla juncea in Argentina and Australia; however, both efforts failed. As part of our effort to
establish this insect in North America, we have conducted an in-depth field study of its biology in
northern Greece with the goal of making a more informed release in which we synchronize the phenologies of this insect and its host. Our study population is at 950 m altitude in the mountains of northern
Greece, an area climatically matching our intended release area, the interior mountains of the state of
Idaho (USA). The field population has a moderate incidence of disease and parasitism, but its population has still remained fairly high for the 3 years it has been studied. Besides obtaining a much more
complete picture of its basic field biology, the most interesting discovery is that despite living in an area
with a long, cold, snow-covered and extensive winter, this insect has no distinct over-wintering stage or
synchronized period of emergence of adults during the summer. Another interesting aspect of this study
is the way that the larvae react and protect themselves from severe and cold winters.
Release of additional strains of the rust, Phragmidium
violaceum, to enhance blackberry biocontrol in Australia
L. Morin,1,2 R. Aveyard,2 K.L. Batchelor,3 K.J. Evans,1,4 D. Hartley2 and M. Jourdan5
Cooperative Research Centre for Australian Weed Management, Glen Osmond, SA, Australia
2
CSIRO Entomology, GPO Box 1700, Canberra, ACT, 2601 Australia
3
CSIRO Entomology, Floreat Park, Private Bag 5, P.O. Wembley, WA 6913, Australia
4
University of Tasmania, TIAR, 13 St Johns Avenue, New Town, TAS 7008, Australia
5
CSIRO European Laboratory, Campus International de Baillarguet, 34980 Montferrier-sur-Lez, France
1
Additional strains of the leaf-rust fungus, Phragmidium violaceum, were approved in early 2004 for
release in Australia to improve the prospects of biologically controlling European blackberry (Rubus
fruticosus agg.) across the wide range of taxa and genotypes that exist. In tests before release, the additional strains were found to be pathogenic on all Rubus genotypes tested except the clones of Rubus
laciniatus. Only one strain (G18-TG-00-4-1) was capable of infecting clones of this species. Results
from host-specificity tests concurred with previous findings that P. violaceum does not pose a threat to
commercial blackberry cultivars and Australian native Rubus species. As the additional strains cannot
be distinguished from existing populations of P. violaceum in Australia, several avenues have been
pursued to develop a simple and robust molecular diagnostic tool to monitor their establishment after
release. The challenges in developing such a tool will be presented. The strains are currently being
released on a national scale in partnership with land holders.
637
XII International Symposium on Biological Control of Weeds
Impact of the bridal creeper rust fungus,
Puccinia myrsiphylli
L. Morin,1,2 A. Reid1,2 and A.J. Willis1,3,4
1
Cooperative Research Centre for Australian Weed Management, Glen Osmond, SA, Australia
2
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
3
CSIRO Plant Industry, GPO 1600, Canberra, ACT 2601, Australia
4
Present address: Australian Government Department of Foreign Affairs and Trade,
Canberra, ACT, Australia
The rust fungus, Puccinia myrsiphylli, released in 2000 for the biological control of bridal creeper
(Asparagus asparagoides) in Australia, is the most widespread and effective agent against this environmental weed of national significance. We used a glasshouse experiment to determine how different
levels of artificial defoliation and rust fungus infection affect bridal creeper growth parameters, such as
below-ground biomass and regrowth. Every fortnight for a total of 20 weeks, plants with a standardized number of tubers were manually defoliated (by 25%, 50%, 75% and 100%) or sprayed with a
suspension of rust spores in water (104 and 105 spores per millilitre). Tuber number, relative growth
rate and rhizome length of plants sprayed with the highest density of rust spores were similar to that of
the 75% defoliated treatment. Most of the plants sprayed with the highest spore density and a few from
the 100% defoliated treatment never regrew after the last treatment application. We also conducted a
fungicide-exclusion field experiment at Camden, NSW, using standardized bridal creeper in pots, to
determine the impact of natural rust fungus infection. Similar reductions in tuber number, rhizome
length and regrowth to that observed in the glasshouse experiment were recorded in the field.
Overview of the biological control of the invasive plant
Chromolaena odorata (Asteraceae) in the Old World
R. Muniappan and G.V.P. Reddy
University of Guam, Western Pacific Tropical Research Center, Mangilao, GU 96923, USA
Chromolaena odorata (Asteraceae) is of Neotropical origin, introduced as an ornamental plant to the
Calcutta Botanical Gardens, India, in 1836. This plant is highly allelopathic. It occupies vacant lands,
pastures, disturbed forests, game reserves, roadsides and orchards. In the dry season, the tops dry
up and become a fire hazard. For the biological control of this weed, several natural enemies were
screened, and of these, the arctiid moth, Pareuchaetes pseudoinsulata, and the tephritid gallfly, Cecidochares connexa, have proven effective. P. pseudoinsulata has been established in Ghana, India, Malaysia, Sri Lanka, Indonesia, Philippines, Guam, Saipan, Rota, Tinian, Pohnpei, Chuuk, Kosrae, Yap
and Papua New Guinea. C. connexa has been established in Indonesia, East Timor, Philippines, Guam,
Rota Saipan, Palau Chuuk, Yap, Pohnpei, Kosrae and Papua New Guinea. In 1922, the International
Organization for Biological Control formally approved the establishment of the Global Chromolaena
Working Group. Seven international workshops have been conducted, one in each of Bangkok, Thailand (1988); Bogor, Indonesia (1991); Abidjan, Ivory Coast (1993); Bangalore, India (1996); Durban, South Africa (2000); Cairns, Australia (2003) and Pingtung, Taiwan (2006). Six proceedings and
16 newsletters were published.
638
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Trichopria columbiana a pupal parasite of the Hydrellia
spp. introduced for the management of hydrilla
J.G. Nachtrieb,1 M.J. Grodowitz2 and N. Harms1
University of North Texas, US Army Engineer Research and Development Center,
Environmental Laboratory, Lewisville, TX, USA
2
US Army Engineer Research and Development Center, Environmental Laboratory, Vicksburg, MS, USA
1
Parasitism by Trichopria columbiana has been suggested as a possible factor limiting population size
and associated damage by Hydrellia pakistanae and Hydrellia balciunasi as biocontrol agents for hydrilla management. Parasitism impacts to emergence and population size were examined and hydrilla
stems collected to determine immature numbers and leaf damage. Pupae were also collected and held
individually to quantify actual parasitism. Limited impact is assumed, as Hydrellia spp. populations increased to almost 6000 immatures per kilogram, with leaf damage approaching 36% despite parasitism
levels of almost 30% at the end of the growing season with even higher fly population levels found in
following years. Host selection behaviour of T. columbiana was also quantified with four behavioural
categories identified: grooming/resting, stem examination, searching and ovipositioning. Wasps spent
the greatest percentage of time searching for prey and the least examining stems. Ovipositioning time
increased in wasps ready for egg deposition only. Choice experiments addressing ovipositing preference of T. columbiana to specific life stages and associated wasp survivorship found that the intermediate pupal stage was both preferred for ovipositing and exhibited the highest percent survivorship.
What is responsible for the low establishment of the
bridal creeper leaf beetle in Australia?
M. Neave,1 L. Morin1,2 and A. Reid1,2
2
1
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Cooperative Research Centre for Australian Weed Management, Glen Osmond, SA, Australia
The leaf beetle Crioceris sp. was released in 2002 for the biological control of bridal creeper (Asparagus asparagoides), one of southern Australias worst environmental weeds. The leaf beetle feeds on
young bridal creeper shoots and emerges at the beginning of the growing season before other biological
control agents (rust fungus Puccinia myrsiphylli and leafhopper Zygina sp.) become active. However,
establishment and population build-up of the leaf beetle has so far been disappointing because it only
established at three of 16 sites where it was released in 2002 and 2003. Glasshouse and field experiments were performed to determine possible causes for low establishment rates. Results confirmed that
our rearing colony is producing highly fecund females. Temperature, however, significantly affected
the number of eggs laid, time to hatch and number of larvae produced. The field experiments indicated
that predation may limit the survival and establishment of the leaf beetle in the field, but additional
experiments including more replicates are required to confirm observations. We recommend releases
of large numbers of beetles to compensate for any predation or parasitism that may occur. Alternatively, exclusion cages could be used at the time of release to enhance initial increases in leaf beetle
populations.
639
XII International Symposium on Biological Control of Weeds
Introduction, specificity and establishment of Tetranychus
lintearius for biological control of gorse in Chile
H. Norambuena
INIA-Carillanca, Casilla 58-D, Temuco, Chile
During the late 1990s, the gorse spider mite, Tetranychus lintearius, was introduced into Chile for the
biological control of Ulex europaeus (gorse). To obtain approval for release, the mite was subjected
to host-specificity testing. The target and 11 test plants species were exposed to the mite in no-choice
and choice tests. Mite development, oviposition and frass deposition were assessed. Mite development
through a next generation occurred on gorse only. Gorse was highly preferred compared to the test
plant species in all the assays (P < 0.001). Spider mite frass proved to be a reliable parameter, indicating feeding (acceptability) to evaluate specificity. High specificity shown for the mite lead to its release
between 1997 and 2000 at sites between the 37 and 42S, and since then, it has established at 90% of
the sites. After 9 years, the mite populations increased more than 1000-fold at many sites and spread
up to 20 km in 3 years. Mite colonization, dispersal and plant damage were stronger in dry relative to
humid areas. This is the first time that a spider mite has been introduced, subjected to host specificity
tests and established in Chile and in South America.
Were ineffective agents selected for the biological control
of skeletonweed in North America? A post-release analysis
L.K. Parsons,1 L.M. Collison,2 J.D. Milan,1 B.L. Harmon,1 G. Newcombe,3
J. Gaskin4 and M. Schwarzlnder1
University of Idaho, Department of Plant Soil and Entomological Sciences, Moscow ID 83844-2339, USA
2
Willamette University, Department of Biology, 900 State Street, Salem, OR 97301, USA
3
University of Idaho, Department of Forest Resources, Moscow, ID 83844-2339, USA
4
USDA-ARS, Sidney, MT 59270, USA
Three agents were released for the biological control of rush skeletonweed, Chondrilla juncea L., in
North America between 1975 and 1977. Although all three species are widely established, weed densities are increasing. Aside from anecdotal reports, there is little quantitative information on the efficacy
of the biological control agents or factors that limit population size or impact. Thus, the question of
whether ineffective agents were selected or effective agents are hampered by unpredicted environmental factors remains unanswered. We studied attack rates for the rush skeletonweed gall mite, Aceria
chondrillae Canestrini (Acari: Eriophyidae), and the rush skeletonweed rust, Puccinia chondrillina
Bubak and Sydenham. We found that, despite the potential for rapid population growth of both agents,
winter mortality for the gall mite was higher than 90% in two consecutive years, and the rust was
attacked by the mycoparasite Eudarluca caricis (Fr.) Eriksson. In addition, there was host-plant resistance to at least the rust. Under favourable environmental conditions (mild winter climates and
absence of the mycoparasite), both agents impair rush skeletonweed in North America. We argue that
the gall mite and the rust fungus are effective biocontrol agents but that their efficacy is hampered by
the heterogeneity of habitats invaded and the diversity of rush skeletonweed genotypes introduced in
North America.
640
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Confirming host-specificity predictions for Oxyops vitiosa,
a biological control agent of Melaleuca quinquenervia
P.D. Pratt, M.B. Rayamajhi, T.D. Center and P.W. Tipping
USDA/ARS, Invasive Plant Research Laboratory, 3205 College Avenue, Fort Lauderdale, FL 33314, USA
An underlying assumption of weed biological control asserts that laboratory-based host-specificity
testing accurately predicts the realized host range of herbivores after release. We tested this assumption
for the Australian weevil, Oxyops vitiosa, which was introduced into Florida (USA) for the biological control of Melaleuca quinquenervia. A series of common garden experiments were conducted to
determine if O. vitiosa would exploit species predicted to be non- or suboptimal hosts in host-range
tests. Adult immigration into replicated common gardens was influenced by species, with >90% of
individuals located on Melaleuca hosts. While adults alighted on 78% of the test plants, oviposition
was restricted to Melaleuca species and the exotic Psidium guajava. All stages of O. vitiosa larvae
were observed on the three Melaleuca species while only first instars occurred on P. guajava. Mean
larval densities were greatest for M. quinquenervia, which represented 92% of all larvae observed. The
residency time for marked weevils placed on test species was greatest for Melaleuca congeners, which
also recruited 98% of all recovered weevils. Felling the M. quinquenervia stand that surrounded a common garden resulted in high levels of immigration within the study plots, but feeding and oviposition
on non-target plants were nonexistent. These results support the premise that risk assessments based on
physiological host ranges are conservative when compared to realized ecological host ranges.
Biological control of the ivy gourd, Coccinia grandis
(Cucurbitaceae), in the Mariana Islands
G.V.P. Reddy,1 J. Bamba,2 T.Z. Cruz1 and R. Muniappan1
1
University of Guam, Western Pacific Tropical Research Center, Mangilao, GU 96923, USA
2
University of Guam, Agriculture and Natural Resources, Cooperative Extension Service,
Mangilao, GU 96923, USA
The invasive plant, ivy gourd, Coccinia grandis, is of African origin and was introduced to Guam
and Saipan in the 1980s. It has occupied more than 200 ha in different parts of Guam and about
2000 ha of Saipan. A biocontrol program has been initiated in Guam and Saipan by introducing the
natural enemies Melittia oedipus Oberthor (Lepidoptera: Sessidae), Acythopeus burkhartorum (Coleoptera: Curculionidae) and Acythopeus cocciniae (Coleoptera: Curculionidae). The Animal and Plant
Health Inspection Service of the US Department of Agriculture has accepted the list of species used
for host-specificity tests in Hawaii, and it was decided that the endemic species, Zehnaria guamensis
(Cucurbitaceae), be tested in Guam. Host-specificity tests of these biocontrol agents were conducted at
the quarantine facility in Guam. In May 2003, A. cocciniae was field-released in Guam and Saipan. It
has established in both islands and caused defoliation of C. grandis by the larval mining of the leaves.
Acythopeus burkhortorum was field-released in Guam in October 2004 and on Saipan in February
2005. Its field establishment on Guam has been confirmed, and its establishment in Saipan is yet to be
verified. Host-specificity tests on M. oedipus are being carried at the University of Guam Quarantine
facility.
641
XII International Symposium on Biological Control of Weeds
Quantifying the impact of biological control: what have we
learned from the bridal creeper-rust fungus system?
A. Reid1,2 and L. Morin1,2
1
Cooperative Research Centre for Australian Weed Management, Glen Osmond, SA, Australia
2
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Evaluating the post-release impact of biological control programs is often neglected due to inadequate
resources, time constraints or the perception that any success should be obvious. A quantitative review
of the Australian weed biological control literature revealed that the majority of programs have not
been adequately evaluated. While many studies have only focussed on assessing agent establishment
and damage caused to individual plants, few have quantified impact of agents on weed populations
and consequences for associated plant communities. The advantages and disadvantages of various approaches to measure impact of agents will be discussed using as a case study the rust fungus, Puccinia
myrsiphylli, released for the biological control of bridal creeper (Asparagus asparagoides). This agent
has been extensively evaluated using glasshouse and field-manipulative experiments as well as preand post-release monitoring of impact on field populations. General guidelines for evaluating impact
of agents will be presented.
From invasive to fixed-in-place: the transformation
of Melaleuca quinquenervia in Florida
P.W. Tipping, P.D. Pratt and T.D. Center
USDA-ARS Invasive Plant Research Laboratory, Fort Lauderdale, FL 33314, USA
Melaleuca quinquenervia (melaleuca) once spread unimpeded across the south Florida landscape,
infesting 0.61 million ha at its height. The complete lack of top-down regulation of its growth and
reproduction resulted in its rapid spread into pine flatwoods, cypress domes, sawgrass prairies and hardwood hammocks. The first biological agent, Oxyops vitiosa, was introduced in 1997 and the second,
Borelioglycaspis melaleucae, in 2002. These natural enemies, especially O. vitiosa, have transformed
both the habit and reproductive capacity of Melaleuca. Plants attacked by O. vitiosa grow slower
(9.1 vs 96.1 cm year-1); produce many more tips (4.2 vs 2.8 tips cm height-1) resulting in a shorter,
bushier plant, produce fewer seed capsules (0.006 vs 0.343 capsule clusters centimetre per tree height),
resulting in 97.5% less seed per tree. Existing tree densities have declined 35.4% since 2002 when
not protected from these natural enemies, while protected areas increased by 9.4%. In another study,
Melaleuca was able to recruit only 2.4% of its previous density of seedlings/saplings. It is now clear
that the capacity of Melaleuca to invade and dominate new habitats has been severely constrained by
biological control.
642
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Long-term field evaluation of Mecinus janthinus releases
against Dalmatian toadflax in Montana (USA)
S.E. Sing,1 D.K. Weaver,1 R.M. Nowierski2 and G.P. Markin3
Montana State University, P.O. Box 173120 Bozeman, MT 59717-3120, USA
USDACSREES, 1400 Independence Avenue, SW, Stop 2220, Washington, DC 20250-2220, USA
3
USDA Forest Service, 1648 South 7th Avenue, Bozeman, MT 59717, USA
1
The toadflax stem mining weevil, Mecinus janthinus, was first released in the United States, in Montana, in 1996. This agent has now become established to varying degrees on subsequent releases made
throughout the state. Multiple releases of M. janthinus have presented researchers with a unique opportunity to evaluate the efficacy of this agent in diverse habitats and under a variety of environmental
conditions. The results presented in this paper summarize findings from long-term field data, illustrating not only the impact of M. janthinus on the target weed, Dalmatian toadflax, but also on correlated
plant community dynamics. These results additionally provide a valuable means to compare and contrast the biotic response and control efficacy of this agent at both a regional and sub-continental scale.
Population dynamics and long-term effects of
Galerucella spp. on purple loosestrife, Lythrum salicaria,
and non-target native plant communities in Minnesota
L.C. Skinner1 and D.W. Ragsdale2
Minnesota Department of Natural Resources, 500 Lafayette Road, St. Paul, MN 55155-4025, USA
University of Minnesota, Department of Entomology, 219 Hodson Hall, 1980 Folwell Ave, St. Paul,
MN 55108, USA
1
2
An 11-year field study (19952006) assessed the effects of Galerucella calmariensis and Galerucella
pusilla (Coleoptera: Chrysomelidae) on purple loosestrife, Lythrum salicaria, and non-target native
plant communities in Minnesota. Galerucella spp. populations initially peaked between 3 and 5 years
after establishment. At all sites, purple loosestrife density declined (up to 90%) in response to an
increase in Galerucella spp. Galerucella spp. appear to have a strong numerical response to purple
loosestrife density, which led to multiple boom and bust cycles occurring on many of the sites during
the 11-year period. Declines in Galerucella spp. typically allowed purple loosestrife populations to
rebound. Generally, Galerucella spp. populations rebounded as loosestrife abundance increased. The
number and amplitude of the boom and bust cycles appears to be related, in part, to the density of the
initial purple loosestrife infestation. Sites where purple loosestrife approached 100% cover tended to
cycle more frequently than sites with higher plant diversity and abundance. It appears that, in more
diverse sites, increased plant competition prevented purple loosestrife from attaining pre-release densities. As purple loosestrife populations declined, plant species richness and/or abundance increased
within release sites. We will discuss these results in context of overall success of the program.
643
XII International Symposium on Biological Control of Weeds
Midges and wasps gain tarsus hold successful release
strategies for two Hieracium biocontrol agents
L.A. Smith,1 P. Syrett2 and G. Grosskopf3
Landcare Research, PO Box 40, Lincoln, New Zealand
2
14 Rockview Place, Christchurch, New Zealand
3
CABI Bioscience Centre Switzerland, Rue des Grillons 1, CH-2800 Delmont, Switzerland
1
Hieracium gall midges (Cecidomyiidae: Macrolabis pilosellae) and gall wasps (Cynipidae: Aulacidea
subterminalis) have been released in New Zealand to combat invasive Hieracium species (Hieracium
pilosellae, Hieracium praealtum, Hieracium caespitosum and Hieracium aurantiacum). Gall wasps
were first released in 1999 and gall midges in 2002. A strategy of releasing small numbers of individuals at many dispersed sites was adopted to overcome site specific factors, which may have precluded
insect establishment. Wasp releases comprised either 100 newly emerged adults or 100 over wintered
galls containing ready to emerge adults. Midge releases comprised 2040 galled plants containing
larvae and pupae, planted at each site. To date, 99 wasp and 136 midge releases have been made, with
proven establishment at 32% and 92% of sites, respectively. In many instances, failed establishment
or establishment and subsequent failure was attributed not only to drought but also to changes in land
management, e.g. cultivation and spraying. Wasp gall densities were measured at up to 122 chambers
per square metre six seasons post release. Midge galled plants were measured at up to 1.2 per square
metre two seasons post release. Long-term biocontrol agent and vegetation monitoring has been established to quantify impact of wasps and midges on hawkweeds and other exotic and native plant
species.
Are seedfeeding insects adequately controlling yellow
starthistle (Centaurea soltitialis) in the western USA?
R.L. Winston and M. Schwarzlnder
University of Idaho, Department of Plant, Soils, and Entomological Sciences, Moscow, ID 83844-2339, USA
Yellow starthistle, Centaurea solstitialis L. (Asteraceae), is an exotic plant infesting 7.5 million ha of
land in the western USA. Six seedhead-feeding insects were released against this plant between 1984
and 1992. We conducted insect exclosure experiments at four yellow starthistle field sites in the Hells
Canyon ecosystem (Idaho) to determine whether insects solely targeting seedheads can adequately
control yellow starthistle or whether the release of additional agents targeting different plant parts is
warranted. We compared several plant response variables between plots sprayed with pyrethroid and
imidacloprid insecticides and plots sprayed with equal amounts of water over the course of two consecutive field seasons. More than 80% of seedheads in control plots were attacked in 2005; more than
93% were attacked in 2006. Seed production in insecticide-treated seedheads was reduced by 27.1%
and 58.5% in 2005 and 2006, respectively. However, yellow starthistle plant density increased from
2005 to 2006. Despite the continued high abundance and attack rates of seedhead feeding insects, it
does not appear that the soil seed bank of yellow starthistle is sufficiently impaired to cause reductions
in plant populations. Thus, the planned release of additional agents targeting roots and stems may
greatly benefit the biological control program against yellow starthistle.
644
Abstracts: Theme 8 Release Activities and Post-release Evaluations
Impact of the rust fungus Uromycladium tepperianum
on the invasive tree, Acacia saligna, in South Africa:
15 years of monitoring
A.R. Wood
ARC-Plant Protection Research Institute, P. Bag X5017, Stellenbosch 7599, South Africa
In the early 1980s, Acacia saligna was rated as the invasive alien weed posing the greatest threat to the
Cape Floristic region. The rust fungus, Uromycladium tepperianum, was introduced into South Africa
from Australia in 1987 and established throughout the range of the weed. Five populations of the weed
were monitored annually from 1991 to 2005, and the number of trees recorded in each of four permanent transects at each site. The tree density declined by between 87% and 98% from 1991 to 2005 at
these sites. The average annual mortality rate (SE) of infected trees dying at the sites was 18% (2%).
Trees were destructively sampled once at 19 sites during March to May of 2004 and 2005. The canopy
mass at these sites was 1290% lower than data published before the introduction of U. tepperianum.
The average age of trees in these sites was between 2.2 and 6.1 years, despite the last major disturbance
being at least 10 to 20 years previous at 11 of the sites. Pod and seed production was determined and
seed fall calculated at four sites in 2004 and compared to data gathered in 1989. The average seed fall
in 2004 was reduced by 89% compared to that recorded earlier. It is concluded that U. tepperianum
continues to be a highly effective biocontrol agent against A. saligna in South Africa.
Success at what price? Establishment, spread and
impact of Pareuchaetes insulata on Chromolaena
odorata in South Africa
C. Zachariades,1 L.W. Strathie,1 D. Sharp2 and T. Rambuda3
Plant Protection Research Institute, Agricultural Research Council, Private Bag X6006,
Hilton 3245, South Africa
2
Department of Water Affairs and Forestry, Private Bag X24, Howick 3290, South Africa
3
University of KwaZulu-Natal, School of Biological and Conservation Sciences, Private Bag X01,
Scottsville 3209, South Africa
1
The defoliator, Pareuchaetes insulata (Lepidoptera: Arctiidae), was the first agent established on Chromolaena odorata (Asteraceae) in South Africa. Following the methods used to establish Pareuchaetes
pseudoinsulata elsewhere, P. insulata from Florida was mass-reared and released at 17 sites in KwaZuluNatal province from 2001 to 2003. Establishment was achieved at only one site, in the coastal town of
Umkomaas, which received 380,000 larvae over 21 months, and even here, the insects virtually disappeared for 15 months. Because the rearing and release methods used should have overcome disease and
Allee effects and climatic modelling indicated reasonable matching, the only obvious remaining reason
for apparent non-establishment was poor matching with plant biotype. Therefore, Jamaican and Cuban
populations were released at several sites from 2003 to 2005. Although initial patterns of increase and
spread of these populations were better than with the Florida population, long-term success was poor.
The Umkomaas population underwent a massive increase in 2005 and, by mid-2006, had defoliated
large areas of chromolaena within a 4-km radius of the release site, causing death or decreased competitiveness of plants. Spread was evident 25 km along the coast and 10 km inland. Laboratory trials
indicated that larval defoliation substantially decreased plant growth. Before this dramatic success, the
programme drew criticism for its high cost and poor results. We discuss issues surrounding the appropriate termination point in attempting to establish an agent.
645
This page intentionally left blank
Theme 9:
Management Specifics,
Integration, Restoration
and Implementation
Session Chair: John Hoffmann
647
This page intentionally left blank
Keynote Presenter
Integration of biological control into weed
management strategies
J.M. DiTomaso1
Summary
In addition to biological control, many other management techniques including mechanical, cultural and chemical options can be used to control invasive plants. Integration of these strategies with
biological control can, in some cases, provide more effective long-term management of a particular
weed than the use of a single control option. Herbicides do not generally impact insect populations
or pathogens and therefore can be used without compromising the effectiveness of most biological
control agents (BCAs). In some cases, particularly with aquatic weeds, the combination of sublethal
herbicide concentrations and BCAs can act synergistically. Other studies have used integrated management systems that combine BCAs and prescribed burning. Although burning in spring or summer
can kill exposed BCAs, both insects and pathogens are mobile organisms and have the opportunity to
readily recolonize the treated site. In other situations, timely burning can increase available nitrogen
or remove the soil litter layer, thus benefiting the population growth of BCAs by increasing access
to the soil surface or improving the nutrient levels in the target plant. Biological control in concert
with competitive, desirable non-target plants can also improve the control of some invasive plant
species and prevent subsequent establishment of the weed. Although there are some studies that have
demonstrated the benefits of incorporating biological control strategies with other management options, examples of this approach are few. However, with an increased research effort in this area,
the potential for successfully using BCAs in an integrated weed management programme is very
promising.
Keywords: IPM, invasive plant, biological control, integrated management.
Introduction
The goal of any weed management plan should be not
only to control the noxious plant but also improve the
degraded community, enhance the utility of that ecosystem, and prevent reinvasion or invasion by other
noxious weed species. To accomplish this often requires a long-term integrated management plan. Development of a management programme and selection
of the proper tool(s) depends on many factors, including the specific weed species requiring management,
its associated vegetative community, initial infestation
density, effectiveness of the control techniques, time
necessary to achieve control, environmental consider-
Department of Plant Sciences, Mail Stop 4, University of California,
One Shields Avenue, Davis, CA 95616, USA <jmditomaso@ucdavis.
edu>.
CAB International 2008
1
ations, chemical use restrictions, topography, climatic
conditions and relative cost of the control techniques
(DiTomaso et al., 2006c). A number of considerations
can influence the choice of options, most important being the primary land-use objective, which can include
forage production, preservation of native or endangered
plant species, wildlife habitat development, water management and recreational land maintenance. A successful long-term management programme often includes
combinations of mechanical, cultural, biological and
chemical control techniques (DiTomaso et al., 2006c).
Probably the least studied integrated strategy involves the combination of biological control with
other management options. This is due in part to the
complexity of such a system compared to other control options, the limited number of weedy species with
well-established biological control agents (BCAs) and
the interdisciplinary nature of such studies. However,
the opportunity to integrate BCAs, whether insects or
pathogens, into long-term integrated weed manage-
649
XII International Symposium on Biological Control of Weeds
ment strategies is numerous and have great potential
for success.
Integrated Approaches
In some cases, a combination of control options may
be necessary to facilitate establishment of desirable
vegetation or to prevent catastrophic wildfires. For example, in some locations in the western United States,
the saltcedar leaf beetle (Diorhabda elongata Brull)
has proven to be a very effective defoliator of saltcedar
(Tamarix spp.). Four to five years of continuous defoliation has begun to result in mortality of saltcedar (Milbrath et al., 2003). Saltcedar in some of these areas can
occupy large expanses and the resultant dried vegetative biomass will inevitably increase the fuel load that
can potentially lead to large-scale wildfires. As a result,
other control options, including prescribed burning and
mechanical removal, are being considered as a subsequent option following the success of the insect agent.
Although this is an example of an integrated management strategy, saltcedar control is primarily restricted
to the use of the BCA. In contrast, most examples of
integrated weed management using biological control
rely on additional management tools to additively or
synergistically contribute to the control of the weed.
This review will describe a variety of integrated strategies that have been, or potentially may be, used successfully. Of the integrated approaches combined with
biological control, most also involve herbicides, prescribed burning, and/or the use of competitive desirable vegetation.
Herbicide and biological control agents
Terrestrial invasive plants: Many researchers have
evaluated the direct impact of herbicides on weed
BCAs (Table 1). In most cases, herbicides do not cause
direct damage to insects, and thus can have excellent
Table 1.
potential in integrated strategies with biological control
organisms.
In some cases, however, herbicides can temporarily
disrupt the establishment of a weed biological control
programme. In leafy spurge (Euphorbia esula L.) stands
treated in autumn with 2,4-D and picloram, Aphthona
spp. were shown to be temporarily less abundant compared to untreated plots, with no significant long-term
benefit to weed control (Larson et al., 2007). The authors felt that the flea beetles may have abandoned the
herbicide-treated patches for greater resources available in untreated plots. They concluded that there was
little advantage to combining herbicides with biological control in areas where biological control is already
considered successful.
However, in a yellow starthistle (Centaurea solstitialis L.)-infested area in California, the attack rates of
the hairy weevil (Eustenopus villosus Boheman) and
the false peacock fly (Chaetorellia succinea Costa) did
not differ in the year after a clopyralid treatment (Pitcairn et al., 2000). At this site, the insects did not avoid
the treated areas, even though yellow starthistle plant
density was considerably lower than the untreated adjacent sites. Lym and Carlson (1994) noted that the
combination of herbicides and biological control was
most effective for leafy spurge control if between 15%
and 25% of the area was left untreated to sustain insect
populations.
One of the key aspects in the proper use of herbicides in combination with BCAs is the timing of the
chemical application. Although insect biological control and herbicides represent a very promising integrated management approach for purple loosestrife
(Lythrum salicaria L.) in North America, applications
of the herbicide glyphosate too early in the season can
destroy the food source for the leaf beetle Galerucella
calmariensis L. (Lindgren et al., 1999). This is of particular concern where the beetles are well established.
A late-season application of the herbicide, when plants
are in bloom, is more compatible with G. calmariensis
List of biological control insects, target weed species and herbicides that can be applied without damage to the
insect.
Insect biocontrol agent
Rhinocyllus conicus Froehlich
Ceuthorhynchidius horridus Panzer
Sphenoptera jugoslavica Obenberger
Cyphocleonus achates Fahraeus
Urophora affinis Frfld. and
U. quadrifasciata Meig.
Eustenopus villosus Boheman and
Chaetorellia succinea Costa
Hyles euphorbiae L.
Spurgia esulae Gagne
Galerucella calmariensis L.
Target weed species
Herbicide
Reference
Carduus spp.
Carduus spp.
Centaurea diffusa Lam.
2,4-D
2,4-D
Picloram, Clopyralid
Trumble and Kok, 1980
Trumble and Kok, 1980
Wilson et al., 2004
Centaurea maculosa Lam.
Centaurea maculosa
Picloram
2,4-D, Picloram
Centaurea solstitialis L.
Clopyralid
Jacobs et al., 2000
McCaffrey and Callihan,
1988; Story et al., 1988
DiTomaso et al. 2006
Euphorbia esula L.
Euphorbia esula
Lythrum salicaria L.
2,4-D, Picloram
2,4-D, Picloram
Glyphosate, Triclopyr
Rees and Fay, 1989
Lym and Carlson, 1994
Lindgren et al., 1999
650
Integration of biological control into weed management strategies
than an early season application because most adults
will have already entered winter diapause. In this situation, G. calmariensis would potentially control young
purple loosestrife seedlings in the following season
(Lindgren et al., 1999).
Autumn applications of 2,4-D and clopyralid were
not compatible with two root-feeding insects, Cyphocleonus achates Fahraeus and Agapeta zoegana L., for
the control of spotted knapweed (Centaurea maculosa
Lam.), whereas spring treatments were compatible
(Story and Stougaard, 2006). This was due to the low
level of survival of the larvae following autumn application timing. By comparison, spring application of
2,4-D plus picloram for leafy spurge control was detrimental to Aphthona spp. establishment, as it eliminated
the adult food source. However, autumn applications
did not negatively affect establishment or reproduction
of the flea beetles, and when used in combination, leafy
spurge control was better and more economical than
either method used alone (Lym et al., 1996; Lym, 1998;
Lym and Nelson, 2002). In the year after the herbicide
treatment, leafy spurge stem density decreased by 85%
to 95% when applications were made to plants infested
with the BCAs (Lym et al., 1996; Lym and Nelson,
2002). In contrast, when only the insects were present,
it took 3 years longer to reduce the infestation to the
same level achieved in 1 year with the integrated strategy. Once leafy spurge density was reduced, the Aphthona flea beetles maintained acceptable control for at
least an additional 7 years (Lym and Nelson, 2002).
This resulted in a threefold to fivefold cost savings to
land managers who typically reapply herbicides annually for a number of years. Lym (1998) concluded that
the benefit of this combination was due to the decrease
in seed production through the activity of the insects and
the inhibition in vegetative spread by creeping roots
through the action of the herbicide. The response was
even better when Aphthona spp. populations were already established at the time of the herbicide treatment
(Lym and Nelson, 2002).
Another study integrated the root-feeding beetle
Sphenoptera jugoslavica Obenberger with low rates of
picloram or clopyralid for the control of diffuse knapweed (Centaurea diffusa Lam.). The results indicated
that, compared to other months, a June herbicide application in Colorado (USA) was the most compatible
with the insect and improved control compared to either technique used alone (Wilson et al., 2004). The
June treatment timing mimicked a midsummer arrest
of diffuse knapweed growth that is normally triggered
by late-summer drought. Insect damage is greater in
drought-stressed plants than in other periods of the season when plants have adequate moisture. Herbicides
applied in June increased infestation of diffuse knapweed plants by the root feeder by 25% the year after
treatment compared to untreated areas. This response
was suggested to be due to the herbicide-induced arrest
in summer growth (Wilson et al., 2004).
Pitcairn et al. (2000) hypothesized that combining
clopyralid application with insect BCAs could provide
more effective long-term control of yellow starthistle.
While an initial clopyralid application would reduce
plant density and new recruitment into the seed bank,
subsequent activity of the BCA on seed production
would slow the rate of reinfestation. In a field study to
test this hypothesis, they found that 1 year following
a clopyralid treatment, there were no significant differences in attack rates between the treated and untreated
populations, despite the low density of yellow starthistle
in the treated area (Pitcairn et al., 2000). In the year after
treatment, they also found that BCAs suppressed seed
production by 76% compared to controls. If successful,
the combination of the two control methods would reduce the need for continuous herbicide treatments.
Aquatic invasive plants: A number of studies have
also demonstrated the value or potential value of using
biological control in an integrated approach with herbicides in aquatic systems. For example, the efficacy
of two biological control weevils, Neochetina eichhorniae Warner and N. bruchi Hustache, was compared for
control of water hyacinth, Eichhornia crassipes (Mart.)
Solms, in both a managed aquatic system, where the
herbicide 2,4-D was routinely applied as a maintenance
treatment, and an unmanaged site with no herbicide
treatment. The maintenance treatment did not eliminate
water hyacinth. Rather, it resulted in lower insect populations and fewer but healthier, more vigorous plants
with higher nutritional quality (Center et al., 1999). Despite the lower insect populations over time, the weevils benefited from the herbicide treatment compared
to the untreated area. Intraspecific competition in the
unmanaged site led to smaller, more stressed individual
plants that were less suitable for weevil populations.
The larger plants of the managed areas enhanced weevil reproduction and, in the long-term, provided more
sustainable control of water hyacinth. The authors concluded that such an integrated approach can exploit the
benefits of both methods while minimizing the negative aspects of each (Center et al., 1999).
Combinations of sublethal herbicide rates and
pathogens have been used synergistically to increase
the susceptibility of the target species, while reducing the damage to non-target plants. This approach
has been used for the control of Eurasian watermilfoil
Myriophyllum spicatum L. (Sorsa et al., 1988), coontail Ceratophyllum demersum L. (Smit et al., 1990)
and particularly for hydrilla Hydrilla verticillata (L.f.)
Royle (Netherland and Shearer, 1996; Nelson et al.,
1998; Shearer and Nelson, 2002).
Applying a sublethal concentration of either fluridone or endothall in combination with a hydrilla-specific
fungal pathogen, Mycoleptodiscus terrestris (Gerd.)
Ostazeski, reduced hydrilla biomass by >90% and was
considerably more effective than using the herbicide or
pathogen alone (Netherland and Shearer, 1996; Nelson
et al., 1998; Shearer and Nelson, 2002). It was hypoth-
651
XII International Symposium on Biological Control of Weeds
esized that the sublethal rate of either herbicide stressed
the hydrilla plants by inhibiting growth and weakened
their natural plant defence system, thus increasing
their susceptibility to pathogen attack (Netherland and
Shearer, 1996; Nelson et al., 1998). For fluridone, the
combination also reduced exposure time. Using either
endothall or fluridone alone required higher rates that
also caused injury to desirable native aquatic plants, including vallisneria (Vallisneria americana Michx.) and
native pondweeds (Potamogeton spp.) (Nelson et al.,
1998; Shearer and Nelson, 2002). By using a lower rate
of the herbicide with the pathogen, excellent selectivity
was achieved and the damage to these non-target species was minimal.
Prescribed burning and biological
control agents
In many areas, prescription burning can be used to
reduce the incidence of catastrophic wildfires (Briese,
1996) or to control invasive plants (DiTomaso et al.,
2006a,b,c). When used in combination with biological
control, the timing of prescribed burning needs to take
into consideration the reproductive capacity and life
history of the BCAs. As with the combination of herbicides and biological control, some unburned patches
should be preserved to maintain a refuge population of
the BCAs (Briese, 1996).
For the integrated control of leafy spurge using burning, Fellows and Newton (1999) showed that burning
conducted between mid-May and October did not have
a detrimental effect on larval survival in Aphthona nigriscutis Foudras and increased the insects establishment by more than twofold compared to unburned
areas. During this timing interval, the adults were not
active and the juveniles were below ground. The enhanced establishment of A. nigriscutis was due to an
increase in colonization in the bare ground created by
the burn. It was postulated that the litter layer interferes with reproduction of the BCAs. A. nigriscutis is
considered difficult to establish in dense, mixed stands
of leafy spurge and grass. Burning, however, increased
the initial density of leafy spurge in the first growing
season after the burn, which increased colonization
through recruitment. However, the increase in the insect populations gave about seven times better control
of leafy spurge than in the unburned site by the end of
the first year (Fellows and Newton, 1999). Based on
these results, the authors concluded that periodic burning of leafy spurge patches at the proper time of year
would lead to expansion of the established colonies and
provide earlier control of leafy spurge (Fellows and
Newton, 1999).
A similar situation was also reported for common
St. Johnswort (Hypericum perforatum L.). Prescription
burning stimulated an increase in the crown density
of the species through vegetative regrowth from rootstocks and lateral roots (Briese, 1996). Concomitant
with this, St. Johnswort beetle (Chrysolina quadrigemina Suffrian) populations dramatically declined as a
result of the fire. Interestingly, insect numbers quickly
rebounded, primarily through influx of the beetle from
adjacent non-burned sites. This increase in the beetle population in the burned site was associated with
increased recruitment and fecundity due to higher
available nitrogen and perhaps other nutrients that
stimulated plant growth in the burn site. High foliar nitrogen levels in these St. Johnswort plants were associated with a population buildup of the beetle, which may
have recognized the burn sites as favourable feeding
areas (Briese, 1996).
Prescribed burning has been shown to be an effective tool for the management of yellow starthistle in
California (DiTomaso et al., 2006b). At the same
time, several biological control insects are widespread
throughout the state and there is some concern that
prescribed burning may negatively impact the presence or activity of these organisms. A study conducted
by DiTomaso et al. (2006b) showed no significant
reduction in the attack rates of false peacock fly (C.
succinea) in a burn site 1 year later. For hairy weevil
(E. villosus), attack rates were high in both burned and
adjacent unburned areas but were highest in the burned
areas. Thus, despite the likely death of weevil larvae
within the seed heads of yellow starthistle in the burned
site the previous year, new recruitment of BCAs the
following year was rapid.
Plant competition and biological
control agents
The competitive ability of a plant can be significantly compromised by the activity of biological control organisms. Insects that bore into roots, shoots and
stems, defoliate the plant, destroy seeds or extract plant
fluids can reduce the competitive ability of that plant
with regard to its neighbouring vegetation (DiTomaso
et al., 2006c). Similarly, pathogens that infect the vegetation or underground parts can reduce the photosynthetic ability, water-mining capacity or vegetative
growth of a species.
The density and cover of spotted knapweed in the
western United States is generally lower in areas with
higher grass competition (Mller-Schrer, 1991). Story
et al. (2000) released the root-mining moth A. zoegana
into two adjacent areas, one with high grass cover
(~50%) and the other with low grass cover (~10%). After monitoring the buildup of moth populations, as well
as the effect on the number of bolting spotted knapweed plants, they found that by the third year, the percentage of knapweed plants infested with A. zoegana
in the high grass-cover plots was nearly twice that of
the low grass-cover plots. This corresponded to a 50%
652
Integration of biological control into weed management strategies
reduction in the number of bolted plants per unit area in
the high-grass site compared to the low-grass site.
Conclusion
Although there are few studies focused on combining
biological control efforts with other weed management
techniques, there are great opportunities to increase the
efficiency of weed management strategies using integrated combinations. The potential exists for BCAs
that may currently be considered ineffective when used
alone to be successful when combined with other control techniques. The complexity of these systems, however, requires a comprehensive understanding of the biology and ecology of the organisms, appropriate timing
of the treatments and the long-term effects. Over time,
integrated management approaches may provide more
effective and economical options compared to the use
of single management techniques.
References
Briese, D.T. (1996) Biological control of weeds and fire management in protected natural areas: are they compatible
strategies? Biological Conservation 77, 135141.
Center, T.D., Dray, F.A. Jr., Jubinsky, G.P. and Grodowitz,
M.J. (1999) Biological control of water hyacinth under
conditions of maintenance management: can herbicides
and insects be integrated? Environmental Management
23, 241256.
DiTomaso, J.M., Brooks, M.L., Allen, E.B., Minnich, R.,
Rice, P.M. and Kyser, G.B. (2006a) Control of invasive
weeds with prescribed burning. Weed Technology 20,
535548.
DiTomaso, J.M., Kyser, G.B., Miller, J.R., Garcia, S., Smith,
R.F., Nader, G., Connor, J.M. and Orloff, S.B. (2006b)
Integrating prescribed burning and clopyralid for the
management of yellow starthistle (Centaurea solstitialis).
Weed Science 54, 757767.
DiTomaso, J.M., Kyser, G.B. and Pitcairn, M.J. (2006c) Yellow Starthistle Management Guide, Publication #2006-03.
California Invasive Plant Council, Berkeley, CA, 74 pp.
Fellows, D.P. and Newton, W.E. (1999) Prescribed fire effects on biological control of leafy spurge. Journal of
Range Management 52, 489493.
Jacobs, J.S., Sheley, R.L. and Story, J.M. (2000) Use of picloram to enhance establishment of Cyphocleonus achates
(Coleoptera: Curculionidae). Environmental Entomology
29, 349354.
Larson, D.L., Grace, J.B., Rabie, P.A. and Anderson, P. (2007)
Short-term disruption of a leafy spurge (Euphorbia esula)
biocontrol program following herbicide application. Biological Control 40, 18.
Lindgren, C.J., Gabor, T.S. and Murkin, H.R. (1999) Compatibility of glyphosate with Galerucella calmariensis; a
biological control agent for purple loosestrife (Lythrum
salicaria). Journal of Aquatic Plant Management 37,
4448.
Lym, R.G. (1998) The biology and integrated management of
leafy spurge (Euphorbia esula) on North Dakota rangeland. Weed Technology 12, 367373.
Lym, R.G. and Carlson, R.B. (1994) Effect of herbicide treatment on leafy spurge gall midge (Spurgia esulae) population. Weed Technology 8, 285288.
Lym, R.G. and Nelson, J.A. (2002) Integration of Aphthona
spp. flea beetles and herbicides for leafy spurge (Euphorbia esula) control. Weed Science 50, 812819.
Lym, R.G., Carlson, R.B., Messersmith, C.G. and Mundal,
D.A. (1996) Integration of herbicides with flea beetles,
Aphthona nigriscutis, for leafy spurge control. In: Moran,
V.C. and Hoffmann, J.H. (eds) Proceedings of the IX International Symposium on Biological Control of Weeds.
University of Capetown, South Africa, pp. 480481.
McCaffrey, J.P. and Callihan, R.H. (1988) Compatibility of
picloram and 2,4-D with Urophora affinis and U. quadrifasciata (Diptera: Tephritidae) for spotted knapweed control. Environmental Entomology 17, 785788.
Milbrath, L.R., DeLoach, C.J. and Knutson, A.E. (2003)
Initial results of biological control of saltcedar (Tamarix
spp.) in the United States. Proceedings of the Symposium,
Saltcedar and Water Resources in the West, pp. 135141.
Mller-Schrer, H. (1991) The impact of root herbivory as
a function of plant density and competition: survival,
growth, and fecundity of Centaurea maculosa in field
plots. Journal of Applied Ecology 28, 353362.
Nelson, L.S., Shearer, J.F. and Netherland, M.D. (1998)
Mesocosm evaluation of integrated fluridone-fungal
pathogen treatment on four submersed plants. Journal of
Aquatic Plant Management 36, 7377.
Netherland, M.D. and Shearer, J.F. (1996) Integrated use of
fluridone and a fungal pathogen for control of hydrilla.
Journal of Aquatic Plant Management 34, 48.
Pitcairn, M.J., DiTomaso, J.M. and Popescu, V. (2000) Integrating chemical and biological control methods for control of yellow starthistle. In: Woods, D.M. (ed) Biological
Control Program Annual Summary, 1999. California Department of Food and Agriculture, Plant Health and Pest
Prevention Service, Sacramento, CA, pp. 5861.
Rees, N.E. and Fay, P.K. (1989) Survival of leafy spurge
hawk moths (Hyles euphorbiae) when larvae are exposed
to 2,4-D or picloram. Weed Technology 3, 429431.
Shearer, J.F. and Nelson, L.S. (2002) Integrated use of endothall and a fungal pathogen for management of the submersed aquatic macrophyte Hydrilla verticillata. Weed
Technology 16, 224230.
Smit, Z.K., Arsenovic, M., Sovljanski, R., Charudattan, R.
and Dukie, N. (1990) Integrated control of Ceratophyllum
demersum by fungal pathogens and fluridone. Proceedings EWRS 8th Symposium on Aquatic Weeds, p. 3.
Sorsa, K.K., Nordheim, E.V. and Andrews, J.H. (1988) Integrated control of Eurasian watermilfoil, Myriophyllum
spicatum, by a fungal pathogen and a herbicide. Journal
of Aquatic Plant Management 26, 1217.
Story, J.M. and Stougaard, R.N. (2006) Compatibility of
two herbicides with Cyphocleonus achates (Coleoptera:
Curculionidae) and Agapeta zoegana (Lepidoptera: Tortricidae), two root insects introduced for biological control of spotted knapweed. Environmental Entomology 35,
373378.
Story, J.M., Boggs, K.W. and Good, W.R. (1988) Optimal
timing of 2,4-D applications for compatibility with Urophora affinis and U. quadrifasciata (Diptera: Tephritidae)
for control of spotted knapweed. Environmental Entomology 17, 911914.
653
XII International Symposium on Biological Control of Weeds
Story, J.M., Good, W.R., White, L.J. and Smith, L. (2000)
Effects of the interaction of the biocontrol agent Agapeta
zoegana L. (Lepidoptera: Cochylidae) and grass competition on spotted knapweed. Biological Control 17, 182190.
Trumble, J.T. and Kok, L.T. (1980) Impact of 2,4-D on Ceuthorhynchidius horridus (Coleoptera: Curculionidae) and
their compatibility for integrated control of Carduus thistles. Weed Research 20, 7375.
Wilson, R., Beck, K.G. and Westra, P. (2004) Combined effects of herbicides and Sphenoptera jugoslavica on diffuse knapweed (Centaurea diffusa) population dynamics.
Weed Science 52, 418423.
654
Biological control of Melaleuca
quinquenervia: goal-based assessment
of success
T.D. Center,1 P.D. Pratt,1 P.W. Tipping,1 M.B. Rayamajhi,1
S.A. Wineriter2 and M.F. Purcell3
Summary
Success means different things to different people. Unfortunately, the success or failure of weed biological control projects is often evaluated by nonparticipants lacking knowledge of the original goals
set by project architects. Criteria for success should match objectives and goals clearly articulated so
that success can be properly archived for future synthesis. The Australian tree Melaleuca quinquenervia (Cav.) S.T. Blake, an aggressive invader of the Florida Everglades, may be the largest plant
ever targeted for biological control. We realized early on that biological control agents would not
remove the many tons of woody biomass that comprised these infestations and so would be unlikely
to reduce the infested acreage. Control of this plant by other means, however, was complicated by the
billions of canopy-held seeds that are released following injury to the tree. A plan was developed in
coordination with land management agencies wherein the goal of biological control was to curtail melaleuca expansion and suppress regeneration while using other means to remove mature trees. Three
insect species have been released and others are under consideration. These agents, supplemented by
the impacts of an adventive rust fungus and a scale insect, have met established goals and this project
shows signs of an emerging success based on the established goals.
Keywords: Everglades, invasive plants, habitat restoration.
Introduction
Success has many fathers while failure dies an orphan. This oft-quoted aphorism illustrates the political
necessity of highlighting successes when they occur so
that ones endeavors continue to be supported in the
future. Unfortunately, weed biological control projects are rarely undertaken based on the likelihood of a
successful outcome (Peschken and McClay, 1995). InUS Department of Agriculture, Agricultural Research Service, Invasive
Plant Research Laboratory, 3225 College Avenue, Fort Lauderdale,
FL 33312, USA.
2
US Department of Agriculture, Agricultural Research Service, Invasive Plant Research Laboratory, 1911 SW 34th Street, Gainesville,
FL 32608, USA.
3
US Department of Agriculture, Agricultural Research Service, Australian Biological Control Laboratory, and Commonwealth Scientific and
Industrial Research Organization, Entomology, Long Pocket Laboratories, 120 Meiers Road, Indooroopilly, QLD 4068, Australia.
Corresponding author: T.D. Center <[email protected], paul.
[email protected], [email protected], min.rayamajhi@ars.
usda.gov, [email protected], [email protected]>.
CAB International 2008
1
stead, biological control is often the method of last resort after other methods against recalcitrant weeds have
failed. This is not conducive to improving the overall
statistical success rate but is often the most responsible
or economic option. Biological control of many serious weed problems would likely never be attempted if
target choice was based primarily on maximizing the
probability of success as advocated by Peschken and
McClay (1995).
Many investigators have focused on the performance of individual agents to gauge success, primarily
with an aim toward predicting which taxonomic groups
make the best biological control agents. Such post-hoc
analyses suggest a low success rate for weed biological control, with only a small proportion of successfully established agents producing effective control
(Crawley, 1989a,b). Critics have used these statistics to
advise against the use of biological control as a weed
management tool (Louda and Stiling, 2004). McFadyen (1998, 2000), however, strongly disagreed with
this advice and emphasized the need for project-based
assessments.
655
XII International Symposium on Biological Control of Weeds
Most authors use the term success to refer only
to complete success, wherein no other measures are
needed to reduce the weed populations to acceptable levels. However, this neglects the importance of partially
successful projects that have value when less effort
is subsequently required to control the weed, because
the density or extent of weed populations is reduced,
or the weed is less able to reinvade cleared areas or
is slower to disperse (Hoffmann, 1995). Success and
failure are at the extreme ends of a continuum of many
possible outcomes and even moderate amounts of
stress can reduce the competitive ability of a weed and
render it less invasive (Center et al., 2005; Coetzee et
al., 2005).
Successful biological control agents often act by
preventing continued expansion of a weed population,
rather than by reducing population densities (Hoffmann,
these proceedings). Hoffmann also noted that it may be
necessary to model weed outbreaks that never happen
to perceive biological control effects. Documentation
of such effects is difficult, at best, which explains why
so many projects are incompletely evaluated and even
successful projects may be undervalued or forgotten.
Thus, statistical success rates should be viewed with
circumspection, inasmuch as only obvious successes
are reported. Furthermore, weed declines may occur incrementally over many years or even decades and may
not be easily observed, especially when observational
baselines shift over time, project funding terminates, or
personnel changes interrupt collection of critical data.
Success of projects should be assessed in terms of
the projects original goals and objectives. Hence, a
project can and should be deemed successful whether
or not the density of the weed is reduced so long as the
goals set out by the project architects are met. In this
sense, it is possible to have complete success without
complete control so long as the project goals are clearly
stated, understood and documented. A recent project aimed at the control of Melaleuca quinquenervia
(Cav.) S.T. Blake (melaleuca) in South Florida as part
of a broader Everglades restoration effort is used herein
to illustrate this concept.
The target
Melaleuca is a large tree (up to 30 m tall) of Australian
origin that was introduced into southern Florida during
the latter part of the 19th century (Dray et al., 2006). It
has invaded wetland habitats, especially fire-maintained
Everglades ecosystems, where the burning regime now
favours melaleuca over less fire-tolerant native species.
As a result, vast areas of these heterogeneous marshes
have been transformed into swamp forests consisting of
melaleuca monocultures. Melaleuca rapidly dominates
infested areas after its initial colonization (Laroche and
Ferriter, 1992) and at its peak was estimated to occupy
at least 607,000 ha of conservation lands in the southern part of Florida (Bodle et al., 1994).
Control of melaleuca is complicated by the fact that
it grows in areas that are hazardous and strenuous to
work in and where access is difficult. These difficulties are exacerbated by the trees reproductive biology.
Melaleuca flowers numerous times each year, often
several times on the same stem axis due to indeterminant growth, forming spike-like clusters composed of
multiple, dichasial groups of three flowers each (Tomlinson, 1980). Each cluster contains up to 75 individual
flowers. Fruits arising from these flowers are persistent
serotinous capsules that each contains 200350 minute
seeds (Meskimen, 1962). These generally remain in the
fruits until disruption of the vascular connection causes
the capsules to desiccate and open, often en masse,
after a fire, freeze, drought, or herbicide treatment
(Meskimen, 1962). A few (about 12% per year) open
continuously, as radial growth of the stem separates the
vascular connection, producing a light, perpetual seed
rain of about 3 billion seeds/ha/year (M. Rayamajhi
et al., unpublished data). Seeds that fall to the ground
form a rather short-lived soil seed bank with a half-life
of less than 1 year (Van et al., 2005). A single large
tree located within a dense stand retains about 50 million seeds in its canopy with stands holding as many
as 25 billion seeds/ha (M. Rayamajhi, unpublished
data). An isolated tree may hold twice as many seeds
as one of similar size in a dense stand. Surprisingly, a
large proportion (8590%) of these are actually hollow
seed coats (Rayachhetry et al., 1998; Rayamajhi et al.,
2002). Nonetheless, the remaining 1015% of embryonic seeds create an enormous regenerative capacity
capable of producing seedling densities of up to 2256
seedlings/m2 (Franks et al., 2006) following a massive
simultaneous seed shed induced by fire, drought, or
herbicide application. These may grow into thickets of
up to 130,000 small trees/ha (Van et al., 2000). As the
stand matures, it thins to about 800015,000 trees/ha
comprised mostly of mature trees with an understory
of suppressed saplings (Rayachhetry et al., 2001). The
standing biomass in these forests has been estimated at
129263 metric tonnes (t)/ha (Van et al., 2000).
Isolated individual trees constitute a seed source for
further encroachment. The seeds, when released, generally fall within 15 m of the parent tree (Meskimen, 1962).
They often grow into dome-shaped clumps or heads
with the parent trees in the centre and progressively
younger trees toward the periphery. These eventually
coalesce with others blanketing vast acreages of wetlands with dense swamp forests. The isolated outliers
therefore are regarded as potential new infestations and, as part of a quarantine strategy, are first
priority for control operations (Woodall, 1981).
The trees within these stands produce multiple adventitious roots that form an intertwined skirt at the
waterline or on saturated soil (Meskimen, 1962). These
contribute biomass to the forest floor and trap large
amounts of litterfall as well as organic debris causing
soil accretion (White, 1994), thus increasing the local
656
Biological control of Melaleuca quinquenervia: goal-based assessment of success
elevation (T. Center, personal observation). Altering
the elevation of the Everglades even by a few centimeters dramatically shifts plant community composition
(Ogden, 2005), thus these newly created melaleuca islands forever change the physiography and ecology of
the area. There is also evidence that essential oils in
melaleuca litter may be allelopathic (Di Stefano and
Fisher, 1983). These changes render infested habitats
unsuitable for many native species making restoration
difficult if not impossible.
The melaleuca management plan
The South Florida Water Management District in conjunction with the Exotic Pest Plant Council convened
a meeting of the major agencies that were managing
the melaleuca problem. They developed a Melaleuca
Management Plan for Florida, published during 1990
and revised in 1994 and 1999.
Two points were evident during the development of
this plan. First, biological control could not eliminate
the huge amounts of woody biomass present; herbicidal
and mechanical control would therefore be needed to
reduce the infestations to a maintenance level. Second,
public agencies could not expend public funds to control melaleuca infestations on private lands that often
abutted cleared tracts of public lands. These unassailable stands provided an invasion front and a potential
seed reservoir to support reinvasion of cleared sites. The
role of biological control in this plan was to neutralize
the reproductive potential of these remaining stands by
reducing seed production, seedling recruitment and regeneration; thereby inhibiting spread, reducing reinvasion of cleared areas and facilitating traditional control
measures. However, implementation of biological con-
Figure 1.
trol would take time, whereas chemical and mechanical
control could be employed rapidly. So the plan relied
on an early deployment of traditional control measures
that would gradually be supplanted by biological control as agents became available (Figure 1).
The biological control agents
Insects associated with melaleuca were enumerated in
Australia during the late 1980s and early 1990s (Balciunas, 1990). These inventories revealed an entomofauna of over 400 species (Balciunas, 1990; Balciunas
et al., 1993a,b, 1994, 1995a,b,c; Burrows et al., 1994,
1996). The most promising species were studied further and three have now been released.
The first insect evaluated was the weevil Oxyops vitiosa Pascoe (Purcell and Balciunas, 1994). This insect,
being a flush feeder on growing stem tips, was desirable because of its ability to disrupt flower production,
which depends on continual growth of the stem axis.
It proved to be host-specific (Balciunas et al., 1994;
G. Buckingham, unpublished report) and was released
during 1997 (Center et al., 2000). Its need to pupate
in dry soil (Purcell and Balciunas, 1994; Center et al.,
2000), however, limited it to habitats that were not permanently under water. Field and laboratory assessments
of a mirid, Eucerocoris suspectus Distant, in Australia (Burrows and Balciunas, 1999) suggested that its
host range was limited to melaleuca and a few close
relatives. Follow-up studies in US quarantine facilities
failed to confirm this so it was dropped from consideration. The host range of the pergid sawfly Lophyrotoma
zonalis (Rohwer) proved sufficiently narrow (Burrows
and Balciunas, 1997; Buckingham, 2001), but after
discovering that larvae synthesize toxic octapeptides
The strategy employed to control melaleuca in south Florida
involved early use of traditional control methods to reduce
biomass while biological controls were being developed and
implemented.
657
XII International Symposium on Biological Control of Weeds
(Oelrichs et al., 1999), we elected not to release it out
of concern over potential negative effects to insectivorous wildlife.
The melaleuca psyllid Boreioglycaspis melaleucae
Moore was found to be host-specific (Purcell et al.,
1997; Wineriter et al., 2003), and was released during
2002 (Center et al., 2006, 2007). It feeds mainly on the
new growth but will also utilize older leaves and the
green stems. Furthermore, it completes its life cycle entirely on the plant so it is less restricted by habitat. The
tube-dwelling pyralid Poliopaschia lithochlora (Lower)
was highly rated because of its ability to damage melaleuca and its preference for low-lying, humid habitats
(Galway and Purcell, 2005), but its use of an ornamental
species, Melaleuca viminalis (Sol. ex Gaertner) Byrnes,
during testing diminished its prospects (M. Purcell, unpublished data). A fergusoninid gall fly, Fergusonina
turneri Taylor, and its mutualistic nematode Fergusobia
melaleucae Davies and Giblin-Davis, also proved to be
highly specific (Giblin-Davis et al., 2001) and were first
released during 2005 (Blackwood et al., 2006). It has
proven difficult to establish but efforts are continuing.
Most recently a stem-galling cecidomyiid, Lophodiplosis trifida Gagn, has proven to be host-specific (S. Wineriter et al., unpublished data) and should gain approval
for release. A bud-feeding weevil Haplonyx multicolor
Lea and a leaf-galling cecidomyiid Lophodiplosis indentata Gagn are currently under consideration.
Two adventive organisms have also recently infested
melaleuca trees in Florida. A pestiferous, undescribed
scale insect (Pemberton, personal communication) was
detected in Florida during 1999. It attacks melaleuca
trees as well as some 300 other plant species (Pemberton, 2003; R. Pemberton, unpublished data). The
guava rust Puccinia psidii G. Winter (Basidiomycetes:
Uredinales), which infects mainly young foliage, appeared during 1997 (Rayachhetry et al., 1997) and is
now widespread.
The effects of the biological
control agents
Numerous studies aimed at determining the impacts
of O. vitiosa and B. melaleucae have been conducted or
are ongoing. However, determinations of the individual
effects of one have been confounded by the presence of
the other, as well as by the presence of the adventive rust
fungus and scale insect. These studies have included
comparisons of melaleuca stands with and without the
agents, caging studies, defoliation experiments, insecticide and fungicide exclusion experiments, and before
and after comparisons of stand dynamics.
Flower and seed production
The effects of herbivory by O. vitiosa on melaleuca
performance were possible early during the release
program when none of the other organisms were present. Pratt et al. (2005) compared flowering frequency in
melaleuca stands where the weevil had been released to
stands without them. They found that the likelihood of
flowering increased with tree size but that undamaged
trees were 36 times more likely to reproduce than damaged trees in similar habitats (Figure 2). Overall, about
45% of the weevil-free trees were flowering compared
to about 2% of infested trees.***
In another study, Pratt et al. (2005) enclosed the
canopies of small (2.9 cm diameter at breast height or
Proportion of Trees Flowering
1.0
Damaged trees
Undamaged trees
0.8
0.6
0.4
0.2
0.0
0
Diameter at Breast Height (cm)
Figure 2.
Release of the weevil Oxyops vitiosa profoundly affected flowering of melaleuca trees. The proportion of the trees that produced flowers was much lower
after being damaged by the weevils regardless of size.
658
Biological control of Melaleuca quinquenervia: goal-based assessment of success
dbh) trees with sleeve cages and introduced weevil larvae into the enclosures, either once or twice, to produce
one or two defoliations of the young foliage. The second defoliation was done about 10 weeks after the first.
These treatments were compared to controls with no
defoliation or to trees artificially defoliated by manually
removing all foliage. Flower production on all trees
was monitored monthly for 1 year. The control trees
flowered normally during this period, whereas trees artificially defoliated failed to produce any flowers. Trees
defoliated once or twice by the weevil larvae produced
a few flowers but numbers were not statistically different from each other or from the artificial defoliation
treatment (Figure 3).
Interestingly, in a comparison of ten herbivoreimpacted trees with ten non-impacted trees at similar, nearby sites at Estero, Florida, Rayamajhi et al.
(unpublished data) found that herbivory by O. vitiosa
resulted in higher rates of capsule abortion when compared to sites without natural enemies. Mean number of
capsules in herbivore-impacted infructescences was reduced by nearly 50% compared to the herbivore-absent
site. This decreased density of capsules was apparent as
gaps in the capsule clusters caused by abortion of the
undeveloped fruits. The herbivore-impacted trees were
very similar to those near Brisbane, Australia where the
average infructescence was 5.7 cm long but contained
only 18 capsules (Rayamajhi et al., 2002). The average
numbers of seeds per capsule were similar in both the
Florida and Australian sites.
Rayamajhi et al. (unpublished data) have also found
that when the trees were subjected to attack by O. vi-
Inflorescences per Tree (no.)
14
12
tiosa, the percentage of embryonic seeds decreased, as
did seed viability and germination ability. Seed viability and germination tests (Van et al., 2005) also revealed
reductions in both measures of seeds from herbivoreattacked trees compared to controls.
Seedling survival
Franks et al. (2006) described the effects of the weevil larvae and the psyllids, alone and in combination,
on growth and survival of melaleuca seedlings by caging the insects on 26 cm-tall seedlings in field plots.
They compared these results to a natural infestation of
the insects on nearby seedlings. O. vitiosa larvae had
no effect on seedling height, leaf number, or survival,
whereas psyllids caused all of these measures to decrease by about 5560% over the 5-month term of the
study. About 95% of seedlings survived when protected
from psyllids as compared to only 40% when exposed
to herbivory (Center et al., 2007).
In another study, Tipping et al. (unpublished data)
found that after becoming infested by both the psyllid
and the weevil, melaleuca trees recruited a much lower
density of seedlings than trees without either herbivore. They also compared densities of saplings in plots
9 m2 that were periodically treated with insecticide to
exclude herbivory to saplings in untreated plots. The
plots were located in an area that had burned during
June 1998, resulting in a massive seed rain and thickets of about 1000 seedlings/m2. By the time the study
was initiated during 2002, these had become saplings
and had grown to about 70 cm in height. Densities in
10
8
6
4
B
2
0
B
B
Control
Herbivory 1
Herbivory 2
Mechanical
Damage Treatment
Figure 3.
Small trees were caged then subjected to herbivory by Oxyops vitiosa either
once (Herbivory 1) or twice (Herbivory 2) or to mechanical defoliation and
compared to undefoliated controls. Defoliated trees, regardless of the manner
of defoliation, produced very few flowers relative to the controls.
659
XII International Symposium on Biological Control of Weeds
the protected plots were virtually unchanged during the
5-year period of the study as compared to those in the
unprotected plots which declined by almost half.
Sapling growth
Tipping et al. (unpublished data) conducted two insecticide exclusion studies on the growth of melaleuca
saplings in common garden experiments over about
a 3-year period. The first experiment investigated the
effect of the melaleuca weevil, O. vitiosa, and supplemental irrigation on the growth of small trees. The second
examined the effects of herbivorous insects (both
the psyllids and the weevils) and plant chemotype
(nerolidol or viridifloral). In both cases, plants treated
with insecticide more than doubled in stature, whereas those not treated grew very little. In the first study,
plants attacked by O. vitiosa grew at a much slower
rate compared to the protected plants (Figure 4). The
unprotected plants produced more stem tips per unit of
height, creating a shorter, bushier habit, which provided
more resource for the tip-feeding insects. Supplemental
irrigation improved the growth of insecticide-treated
trees but had no effect on trees that were not treated
with insecticide. Chemotype had no apparent effect on
the impact of the insects. Seed capsule production was
much lower among unprotected plants in both studies.
140
Change in Height (%)
120
100
Stand dynamics
Rayamajhi et al. (2007) studied the dynamics of melaleuca stands before and after the widespread impacts
of the biological control agents. They found that the
average density of the trees in mature stands declined
by 72% overall from 15,800 trees/ha during 1996 to
4400 trees/ha during 2003. Interestingly, the standing
biomass based on harvesting studies increased somewhat from an initial average of 263 t/ha to 274 t/ha
during the latter harvest. This was because most of the
mortality was among the smaller suppressed trees in
the understory that represented a small proportion of
the biomass. The density of small trees, those with a
dbh of less than 10 cm, decreased 83% from 12,600 to
2200 trees/ha; density of intermediate-sized trees with
a dbh of 1020 cm decreased 46% from 2600 to 1400
trees/ha; density of large trees >20 cm increased from
600 to 800 trees/ha.
Another study (Rayamajhi et al., 2007) showed that
densities decreased between 1997 and 2006, in part due
to self-thinning. The decline accelerated after the effects of biological control became apparent and the rate
of decline was inversely related to the position of the
trees within the stands. Densities of trees at the periphery, which consisted mostly of small individuals, decreased by about 6076 individuals/ha/year before 2001
Insecticide, Rainfall Only
Insecticide, Rainfall + Irrigation
No insecticide, Rainfall Only
No insecticide, Rainfall + Irrigation
80
60
40
20
01 01 02 02 02 02 02 02 02 02 02 02 02 03 03 03 03 03 03 03 03
v- ec- an- eb- ar- pr- ay- un- ul- ug- ct- ov- ec- an- eb- ar- pr- un- ul- ep- cto
J
J
-N -D -J -F -M -A -M -J 0- -A 1-O5-N 7-D 5-J 6-F 5-M 0-A 9-J 31- 4-S 0-O
1
29 18 15 19 13 17 21 25 3 27
1 1 1 2 2 3 1
Sample Date
Figure 4.
Small trees grown in a common garden plot were treated with insecticide to exclude herbivorous insects and given supplemental irrigation and compared to unprotected and unwatered
trees. Protected trees grew vigorously and those receiving supplemental irrigation grew the
most. Unprotected trees grew very little and the supplemental irrigation seemed to have little effect.
660
Biological control of Melaleuca quinquenervia: goal-based assessment of success
as compared to 16,725 individuals/ha/year after 2001.
Densities in the inner portions of the stands, which contained higher proportions of larger trees, decreased at
relatively constant rates. This further demonstrated the
greater effect of herbivory on smaller trees. The average
diameter of the trees increased, not because they grew
but because of selective mortality of smaller individuals. This was corroborated by a decrease in or leveling
off of total basal area coverage during the post-release
period in contrast to a prior increasing trend.
Despite the finding that the surviving larger dominant trees accounted for most of the biomass, biomass
allocation changed due to extensive defoliation of all
of the trees. The foliage of large trees growing in dense
stands was limited to the upper branches at the treetops
and this accounted for only 5.1% of the total biomass
during 1996, before insect-induced defoliation. This decreased from 17 to 8 t/ha to represent only 1.5% of the
total biomass during 2003 (Figure 5). The biomass allocated to seed capsules decreased by 85% from 6.7 t/ha,
or 0.46% of the total biomass to 1.0 t/ha, or 0.29% of
the total biomass.
Litter-traps were placed under mature melaleuca
stands to collect leaf litter in an attempt to measure the
activity of the biological control agents in the canopy of
taller trees. The proportion of fallen leaves that exhibited weevil damage symptoms was analysed. Though
the weevil releases began in 1997, the first weevil-damaged leaves did not appear in the traps until 1999 (represented by 5% of the trapped leaves) and by 2005, the
proportion of damaged leaves reached approximately
45% (Rayamajhi et al., 2007). This increased percentage of damaged leaves reflected the decreasing proportions of leaf biomass (stem to leaf biomass), increasing
tree mortality, and decreasing tree densities.
Figure 5.
Stump regrowth
The ability of melaleuca to sprout from cut stumps
complicates control. This requires follow-up herbicide
treatment to prevent coppicing and stand regeneration.
Several studies have indicated that the flush of foliage
associated with this regrowth is highly attractive to both
psyllids and weevils, as well as to the rust fungus. Pratt
et al. (unpublished data) found that insecticide exclusion of biological control agents led to an increase in
leaf and stem biomass compared to unprotected stumps
(Figure 6). Chronic attack over an 18-month period led
to mortality of almost half the unprotected plants. In
a similar insecticide- and fungicide-exclusion study,
Rayamajhi et al. (unpublished data) found that the rust
fungus, P. psidii, played an important additive role.
Proportion of photosynthetic tissues and the mortality
of stems were higher in treatments involving both insects (O. vitiosa and B. melaleucae) and rust (P. psidii)
together than in treatments using either alone. Death of
the regrowth often led to the death of the stump itself
(M. Rayamajhi et al., unpublished data). These data
indicate that biological control can compliment, and
in some cases replace, the use of herbicides for stump
treatment.
Discussion
Clearly, many melaleuca stands have undergone significant declines and remaining trees are now in poor
condition. However, vast stands of melaleuca still exist
that overtly appear unchanged. Yet after closer scrutiny,
we have revealed that the dynamics of these stands
have changed in very significant ways. Fewer trees
now produce flowers, those that do flower produce
The proportion of the total tree biomass allocated to foliage declined dramatically due to defoliation primarily by Oxyops vitiosa and the psyllid Boreioglycaspis melaleucae.
661
XII International Symposium on Biological Control of Weeds
Figure 6.
Regrowth from stumps, reflected by the biomass of stems and leaves
produced after felling of the original trees, was substantially more when
regrowth was treated with insecticide thus reducing the effects of the weevils and the psyllids.
fewer inflorescences and the inflorescences produced
contain fewer individual blossoms. Many of the fruits
abort and those that do manage to set seed produce a
smaller proportion of viable seeds. The constant defoliation of the stem tips causes the capsules to desiccate
and release seeds during drier periods when conditions
are unfavourable for germination. Those that do fall,
lodge in a favourable site and manage to germinate are
infested by psyllids that kill a large proportion before
they attain a significant size. If they survive, they grow
slowly due to constant defoliation and produce few
flowers. Meanwhile, existing stands have nearly been
removed from publicly held lands and those on private
lands are less invasive. Hence, the goals of the project,
as stated above, have been met so the project should
be considered a success. It is not yet a complete success in that biological control is more effective in some
habitats and during some periods than others but additional agents that are currently under development may
fill these gaps.
Acknowledgements
The research reported herein was supported by funding
from the South Florida Water Management District,
the Florida Department of Environmental Protection,
the US Army Corps of Engineers, the Miami-Dade
County Department of Environmental Resource Management, and Lee County as well as by the USDAAgricultural Research Service Areawide Projects. We
thank all past and present staff of the USDAARS Australian Biological Control Laboratory and the Invasive
Plant Research Laboratory. We are also indebted to
the Student Conservation Service and the AmeriCorps
program for the tremendous support provided by the
many conservation interns that have been involved in
this program.
References
Balciunas, J.K. (1990) Australian insects to control melaleuca.
Aquatics 12, 1519.
Balciunas, J.K., Bowman, G.J. and Edwards, E.D. (1993a)
Herbivorous insects associated with paperbark Melaleuca
quinquenervia and its allies: I. Noctuoidea (Lepidoptera).
Australian Entomologist 20, 1324.
Balciunas, J.K., Burrows, D.W. and Edwards, E.D. (1993b)
Herbivorous insects associated with the paperbark tree
Melaleuca quinquenervia and its allies: II. Geometridae
(Lepidoptera). Australian Entomologist 20, 9198.
Balciunas, J.K., Burrows, D.W. and Purcell, M.F. (1994) Insects to control melaleuca I: Status of research in Australia. Aquatics 16, 1013.
Balciunas, J.K., Burrows, D.W. and Horak, M. (1995a)
Herbivorous insects associated with the paperbark tree
Melaleuca quinquenervia and its allies: IV. Tortricidae
(Lepidoptera). Australian Entomologist 22, 125135.
Balciunas, J.K., Burrows, D.W. and Purcell, M.F. (1995b)
Insects to control melaleuca II: Prospects for additional
agents from Australia. Aquatics 17, 16 and 1821.
Balciunas, J.K., Burrows, D.W. and Purcell, M.F. (1995c)
Australian insects for the biological control of the paperbark tree, Melaleuca quinquenervia, a serious pest of
Florida, USA, wetlands. In: Delfossee, E.S. and Scott,
R.R. (eds) Proceedings of the Eighth International Symposium on Biological Control of Weeds. DSIR/CSIRO,
Melbourne, Australia, pp. 247267.
Blackwood, S., Pratt, P.D. and Giblin-Davis, R. (2006) Budgall fly release for biocontrol of Melaleuca in Florida.
Biocontrol News and Information 26(2), 48N.
662
Biological control of Melaleuca quinquenervia: goal-based assessment of success
Bodle, M.J., Ferriter, A.P. and Thayer, D.D. (1994) The bio
logy, distribution, and ecological consequences of Melaleuca quinquenervia in the Everglades. In: Davis, S.M
and Ogden, J.C. (eds) EvergladesThe Ecosystem and
its Restoration. St. Lucie Press, Delray Beach, FL, USA,
pp. 341355.
Buckingham, G.R. (2001) Quarantine host range studies with
Lophyrotoma zonalis, an Australian sawfly of interest for
biological control of melaleuca, Melaleuca quinquenervia,
in Florida. BioControl 46, 363386.
Burrows, D.W. and Balciunas, J.K. (1997) Biology, distribution and host range of the sawfly, Lophyrotoma zonalis
(Hym., Pergidae), a potential biological control agent for
the paperbark tree, Melaleuca quinquenervia. Entomophaga 42, 299313.
Burrows, D.W. and Balciunas, J.K. (1999) Host-range and
distribution of Eucerocoris suspectus (Hemiptera: Miridae), a potential biological control agents for the paperbark tree Melaleuca quinquenervia (Myrtaceae). Environmental Entomology 28, 290299.
Burrows, D.W., Balciunas, J.K. and Edwards, E.D. (1994)
Herbivorous insects associated with the paperbark tree
Melaleuca quinquenervia and its allies: III. Gelechioidea
(Lepidoptera). Australian Entomologist 21, 137142.
Burrows, D.W., Balciunas, J.K. and Edwards, E.D. (1996)
Herbivorous insects associated with the paperbark tree
Melaleuca quinquenervia and its allies: V. Pyralidae (Lepidoptera). Australian Entomologist 23, 716.
Center, T.D., Van, T.K., Rayachhetry, M., Buckingham, G.R.,
Dray, F.A., Wineriter, S.A., Purcell, M.F. and Pratt, P.D.
(2000) Field colonization of the melaleuca snout beetle
(Oxyops vitiosa) in south Florida. Biological Control 19,
112123.
Center, T.D., Van, T.K., Dray, F.A., Franks, S.J., Rebelo, M.T.,
Pratt, P.D. and Rayamajhi, M.B. (2005) Herbivory alters
competitive interactions between two invasive aquatic
plants. Biological Control 33, 173185.
Center, T.D., Pratt, P.D., Tipping, P.W., Rayamajhi, M.B.,
Van, T.K., Wineriter, S.A., Dray, F.A., Jr. and Purcell,
M. (2006) Field colonization, population growth, and
dispersal of Boreioglycaspis melaleucae Moore, a biological control agent of the invasive tree Melaleuca
quinquenervia (Cav.) Blake. Biological Control 39,
363374.
Center, T.D., Pratt, P.D., Tipping, P.W., Rayamajhi, M.B.,
Van, T.K., Wineriter, S.A. and Dray, F.A., Jr. (2007) Initial impacts and field validation of host range for Boreioglycaspis melaleucae Moore (Hemiptera: Psyllidae), a
biological control agent of the invasive tree Melaleuca
quinquenervia (Cav.) Blake (Myrtales: Myrtaceae: Leptospermoideae). Environmental Entomology 36, 569
576.
Coetzee, J.A., Center, T.D., Byrne, M.J. and Hill, M.P. (2005)
Impact of the biocontrol agent Eccritotarsus catarinensis,
a sap-feeding mirid, on the competitive performance of
waterhyacinth, Eichhornia crassipes. Biological Control
32, 9096.
Crawley, M.J. (1989a) Plant life-history and the success of
weed biological control projects. In: Delfosse, E.S. (ed.)
Proceedings of the VII International Symposium on Biological Control of Weeds. Instituto Sperimentale per la
Patologia Vegetale, Ministero dell Agricoltura e delle Foreste, Rome, Italy, pp. 1726.
Crawley, M.J. (1989b) The successes and failures of weed
biocontrol using insects. Biocontrol News and Information 10, 213223.
Di Stefano, J.F. and Fisher, R.F. (1983) Invasion potential of
Melaleuca quinquenervia in Southern Florida, USA. Forest Ecology and Management 7, 133141.
Dray, F.A., Bennett, B.C. and Center, T.D. (2006) Invasion
history of Melaleuca quinquenervia (Cav.) S.T. Blake in
Florida. Castanea 71, 210225.
Franks, S.J., Kral, A.M. and Pratt, P.D. (2006) Herbivory by introduced insects reduces growth and survival of Melaleuca
quinquenervia seedlings. Environmental Entomology 35,
366372.
Galway, K.E. and Purcell, M.F. (2005) Laboratory life history and field observations of Poliopaschia lithochlora
(Lower) (Lepidoptera: Pyralidae), a potential biological
control agent for Melaleuca quinquenervia (Myrtaceae).
Australian Journal of Entomology 44, 7782.
Giblin-Davis, R.M., Makinson, J., Center, B.J., Davies, K.A.,
Purcell, M., Taylor, G.S., Scheffer, S.J., Goolsby, J. and
Center, T.D. (2001) Fergusobia/Fergusonina-induced
shoot bud gall development on Melaleuca quinquenervia.
Journal of Nematology 33, 239247.
Hoffmann, J.H. (1995) Biological control of weeds: the way
forward, a South African perspective. In: British Crop Protection Council Proceedings No. 64: Weeds in a Changing
World. BCPC, Farnham, Surrey, UK, pp. 7789.
Laroche, F. and Ferriter, A.P. (1992) The rate of expansion
of melaleuca in south Florida. Journal of Aquatic Plant
Management 30, 6265.
Louda, S. M. and Stiling, P. (2004) The double-edged sword
of biological control in conservation and restoration. Conservation Biology 18, 5053.
Meskimen, G.F. (1962) A Silvical Study of the Melaleuca
Tree in South Florida. MS thesis. University of Florida,
Gainesville, FL, USA, 177 pp.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
McFadyen, R.E.C. (2000) Successes in biological control
of weeds. In: Spencer, N.R. (ed.) Proceedings of the
X International Symposium on Biological Control of
Weeds. Montana State University, Bozeman, MT, USA.,
pp. 314.
Oelrichs, P.B., MacLeod, J.K., Seawright, A.A., Moore,
M.R., Ng, J.C., Dutra, F., Riet-Correa, F., Mendez, M.C.
and Thamsborg, S.M. (1999) Unique toxic peptides isolated from sawfly larvae in three continents. Toxicon 37,
537544.
Ogden, J.C. (2005) Everglades ridge and slough conceptual
ecological model. Wetlands 25, 810820.
Pemberton, R.W. (2003) Potential for the biological control
of the lobate lac scale, Paratachardina lobata lobata (Hemiptera: Kerridae). Florida Entomologist 86, 353360.
Peschken, D.P. and McClay, A.S. (1995) Picking the target:
A revision of McClays scoring system to determine the
suitability of a weed for classical biological control. In:
Delfosse, E.S. and Scott, R.R. (eds) Proceedings of the
Eighth International Symposium on Biological Control of
Weeds. CSIRO, Melbourne, Australia, pp. 137143.
Pratt, P.D., Rayamajhi, M.B., Van, T.K., Center, T.D. and Tipping, P.W. (2005) Herbivory alters resource allocation and
compensation in the invasive tree Melaleuca quinquenervia. Ecological Entomology 30, 316326.
663
XII International Symposium on Biological Control of Weeds
Purcell, M.F. and Balciunas, J.K. (1994) Life history and distribution of the Australian weevil Oxyops vitiosa (Coleoptera: Curculionidae), a potential biological control agent
for Melaleuca quinquenervia (Myrtaceae). Annals of the
Entomological Society of America 86, 867873.
Purcell, M.F., Balciunas, J.K. and Jones, P. (1997) Biology
and host-range of Boreioglycaspis melaleucae (Hemiptera: Psyllidae), a potential biological control agent for
Melaleuca quinquenervia (Myrtaceae). Environmental
Entomology 26, 366372.
Rayachhetry, M.B., Elliott, M.L. and Van, T.K. (1997) Natural epiphytotic of the rust Puccinia psidii on Melaleuca
quinquenervia in Florida. Plant Disease. 81, 831.
Rayachhetry, M.B., Elliott, M.L. and Van, T.K. (1998) Regeneration potential of the canopy-held seeds of Melaleuca quinquenervia in south Florida. International Journal
of Plant Science 159, 648654.
Rayachhetry, M.B., Van, T.K., Center, T.D. and Laroche, F.
(2001) Dry weight estimation of the aboveground components of Melaleuca quinquenervia trees in southern
Florida. Forest Ecology and Management 142, 281290.
Rayamajhi, M.B., Van, T.K., Center, T.D., Goolsby, J.A.,
Pratt, P.D. and Racelis, A. (2002) Biological attributes of
the canopy held melaleuca seeds in Australia and Florida,
US. Journal of Aquatic Plant Management 40, 8791.
Rayamajhi, M.B., Van, T.K., Pratt, P.D., Center T.D. and Tipping, P.W. (2007). Melaleuca quinquenervia dominated
forests in Florida: analyses of natural-enemy impacts on
stand dynamics. Plant Ecology 192, 119132.
Tomlinson, P.B. (1980) The Biology of Trees Native to Tropical Florida. Harvard University Printing Office, Allston,
MA, USA, 480 pp.
Van, T.K., Rayachhetry, M.B. and Center, T.D. (2000). Estimating above-ground biomass of Melaleuca quinquenervia
in Florida, USA. Journal of Aquatic Plant Management
38, 6267.
Van, T.K., Rayamajhi, M.B. and Center, T.D. (2005) Seed
longevity of Melaleuca quinquenervia: a burial experiment in south Florida. Journal of Aquatic Plant Management 43, 3942.
White, P. (1994) Synthesis: vegetation pattern and process in
the Everglades ecosystems. In: Davis, S.M. and Ogden,
J.C. (eds) Everglades, The Ecosystem and Its Restoration.
St. Lucie Press, Florida, USA, pp. 445458.
Wineriter, S.A., Buckingham, G.R. and Frank, J.H. (2003)
Host range of Boreioglycaspis melaleucae Moore (Hemiptera: Psyllidae), a potential biological control agent of
Melaleuca quinquenervia (Cav.) S.T. Blake (Myrtaceae),
under quarantine. Biological Control 27, 273292.
Woodall, S.L. (1981) Integrated methods for melaleuca control. In: Geiger, R.K. (ed.) Proceedings of Melaleuca
Symposium, Sept. 2324 1980. Florida Department of
Agriculture and Consumer Services, Division of Forestry,
Gainesville, FL, USA, pp. 135140.
664
Hydrilla verticillata threatens
South African waters
J.A. Coetzee1 and P.T. Madeira2
Summary
South Africas inland water systems are currently under threat from hydrilla, Hydrilla verticillata
L. Royle (Hydrocharitaceae), the worst submerged aquatic weed in the USA. The presence of the
weed was confirmed for the first time in South Africa in February 2006, on Pongolapoort Dam in
KwaZulu-Natal. An aerial survey revealed that the infestation on this dam covers approximately
600 ha, which is far greater than initially thought. Despite reports that it may be present in other water
bodies, surveys have shown that it is restricted to Pongolapoort Dam. We conducted a boater survey
which showed that there is significant potential for this devastating weed to spread beyond Pongolapoort Dam, and containment of hydrilla is of utmost priority. Research into the suitability of the
already established biological control agents, Hydrellia pakistanae Deonier and H. balciunasi Bock
(Diptera: Ephydridae), from the USA, as potential agents in South Africa, is also being conducted.
However, the South African hydrilla biotype is different from the biotypes in the USA, and this needs
to be borne in mind when considering which agents to release.
Keywords: potential spread, management, genetic analysis.
Introduction
The confirmation of Hydrilla verticillata L. Royle
(Hydrocharitaceae) (hydrilla) in South Africa from
Pongolapoort Dam, KwaZulu-Natal province (KZN),
in early 2006 (L. Henderson, personal communication,
2006) prompted immediate action to contain and control this weed, and prevent further spread to other water bodies around South Africa. At present, it appears
that hydrilla is restricted to Pongolapoort Dam, which
is the centre of a multimillion rand tourist industry. It
is imperative that we gain an understanding of the dynamics of the hydrilla invasion in South Africa, and potential for its control, because there is a knowledge gap
in South Africa surrounding submerged aquatic plants,
particularly from a biological control aspect.
ARCPlant Protection Research Institute, P/Bag X134, Queenswood
0121, South Africa. Current address: Department of Zoology and Entomology, Rhodes University, PO Box 94, Grahamstown 6140, South
Africa.
2
USDA-ARS, Invasive Plant Research Laboratory, 3205 College Avenue, Fort Lauderdale, FL 33314, USA.
Corresponding author: J.A. Coetzee <[email protected], paul.
[email protected]>.
CAB International 2008
1
Current distribution of
hydrilla in South Africa
and potential for spread
Hydrilla is one of the most problematic submerged
plants worldwide, invading both tropical and temperate
regions because of its tolerance to a wide range of environmental conditions (Cook and Lnd, 1982). It is not
clear how or when hydrilla entered South Africa, and
so the first step in the hydrilla biocontrol programme in
South Africa was to determine the extent of its distribution. Following reports that hydrillas presence was
suspected in a number of water bodies in KZN, both
aerial and boat surveys were undertaken, which confirmed that hydrilla is currently restricted to an area of
about 600 ha in Pongolapoort Dam. However, heavy
rains in early January 2007 resulted in the flooding
of this area, and the dam increased in capacity from
73% to 92% full in 1 week. The possibility that hydrilla
has spread throughout the dam, and into the Pongola
River below should not be ruled out, and warrants further investigation.
While containment of hydrilla in Pongolapoort Dam
is currently the main control strategy in South Africa,
there is potential for this plant to spread. In the USA,
the main mode of spread of the weed is via recreational
665
XII International Symposium on Biological Control of Weeds
boaters and fishermen (Balciunas et al., 2002) as fragments of the plant get caught in anchors and propellers
and are then transported between water bodies. It is
therefore very likely that hydrilla may spread throughout South Africa in this manner, particularly because
Pongolapoort Dam attracts thousands of tourists annually, and because it is home to the annual Tiger Fishing Festival, the biggest tiger fishing competition in
the southern hemisphere that attracts fishermen from
all over South Africa, and neighbouring Swaziland and
Mozambique.
Fishermen are primarily responsible for hydrillas
spread in the USA. Therefore we conducted a survey at
the annual Tiger Fishing Festival in September 2006 to
determine the potential for hydrilla to spread throughout South Africa by assessing boating behaviour of the
fishermen, and whether they were aware of the presence of hydrilla on the dam. One hundred sixty-three
fishermen were asked questions from a structured questionnaire. The results showed that 51% of the fishermen
interviewed only used their boats once on Pongolapoort
Dam, and that was at the September competition. The
results also showed that 14.1% of the fishermen used
their boats only on Pongolapoort Dam between January 2005 and September 2006, and 20 respondents
(12.2%) never used their boats anywhere except once
at the 2006 fishing competition, between January 2005
and September 2006.
However, analysis of the number of times fishermen used their boats in South Africa highlighted that
dams outside of KZN were visited more frequently
than those in KZN, and the majority of fishermen traveled between 200 and 800 km to reach their fishing
destinations, emphasizing the potential for hydrilla to
spread around South Africa. Even though a containment strategy is in place on Pongolapoort Dam, this
survey stressed that more water bodies in South Africa
need to be assessed for the presence of hydrilla as a
result of boating activities before the fishing competition in September.
Management options
Mechanical and chemical control has been the most
widely used control methods in the USA, although
their success is varied. Typically, mechanical control
is time-consuming and only offers temporary control,
and its use has been dissuaded in South Africa, particularly because new infestations can result from plant
fragments. Until a biocontrol programme can be implemented, chemical control is currently the most favourable option for hydrilla. It should be controlled using
herbicides as soon as possible because it is confined to
only one system. The most effective herbicide to date
against hydrilla in the USA is fluridone, which has been
widely used for large-scale control (Dayan and Netherland, 2005). Trials with this herbicide will commence
as soon as it is imported into South Africa.
Fluridone does have non-target side effects on other
aquatic vegetation and fish, which has led to sublethal
doses being applied as this minimizes these effects, and
is more cost-effective. However, hydrilla has become
resistant to these doses in the USA, which has complicated control programmes (Michel et al., 2004; Dayan
and Netherland, 2005). So it becomes a riskbenefit issue in South Africa is it worth using a lethal dose
that will remove large amounts of the plant, but that
will impact the fauna and flora in the dam, against the
potential of resistance developing if sublethal doses are
used, thereby ruling out the most effective control strategy against it?
The most sustainable long-term strategy to control
hydrilla should be biological control. The option of using the two species of ephydrid flies, Hydrellia pakistanae Deonier and H. balciunasi Bock (both Diptera:
Ephydridae) as potential control agents against hydrilla
in South Africa is being investigated, because these are
the only two agents that have established in the USA
(Center et al., 1997; Bennett and Buckingham, 2000;
Grodowitz et al., 2000; Wheeler and Center, 2001). In
addition, the weevil, Bagous hydrillae OBrien (Coleoptera: Curculionidae), which was tested and released
in the USA to control hydrilla but never established
because it requires periods of drought for pupation
(Grodowitz et al., 2000), is being considered as a
control agent in South Africa because drawdowns are
implemented on Pongolapoort dam. Permits have been
granted to import the flies and the weevil into South
Africa from the USA, so that host specificity testing
may commence.
Expanded surveys are also being conducted by both
the USDA-ARS and CSIRO to find additional control
agents for hydrilla in the USA (Overholt and Wheeler, 2006), following the discovery that infestations of
hydrilla are resistant to fluridone (Michel et al., 2004;
Puri et al., 2007). Surveys in Burundi, Uganda and
other central and east African countries, and in Sumatra and China have found promising agents (Overholt
and Wheeler, 2006), which could be considered as additional control agents in South Africa.
Identification and origin of the
introduced biotype using
chloroplastic markers
Hydrilla is a widely distributed species whose range
extends from New Zealand and Australia, through
Southeast Asia, north through China, into Siberia, and
west into Pakistan, and it also has a local and disjointed
range in Africa and northern Europe (Cook and Lnd,
1982). Studies have identified more than 28 different
hydrilla biotypes, which could have important consequences for biological control of the plant in South
Africa. Four major biotype clusters and one minor outlier cluster have been identified (Madeira et al., 1997,
666
Hydrilla verticillata threatens South African waters
1999). The USA has two hydrilla biotypes a dioecious strain that clusters closely with an Indian strain,
and a monoecious strain that clusters closely with an
accession from Korea (Madeira et al., 1997).
Several biocontrol agents have been released in
the USA, but only the leaf mining fly H. pakistanae
is causing significant damage (Wheeler and Center,
2001). Regional variation in both the host plant and the
control agent populations could potentially affect the
effectiveness of new releases. It is therefore essential
in any biocontrol programme to know which biotype
is being dealt with to maximize the efficacy of control
agents, by selecting agents from the same area as the
plant biotype.
Therefore, samples of South African hydrilla were
analysed using the trnL intron and trnL-F intergenic
spacer of the chloroplast to determine to which major
cluster of worldwide hydrilla the South African hydrilla belongs (Madeira et al., 2007). In the sequencing
it was identical to Malaysian and Indonesian samples.
This biotype is also monoecious, and produces copious numbers of flowers, pollen, seedpods and seeds.
South African hydrilla is therefore very different from
hydrilla in the USA, and the control agents currently in
use in the USA might not be as suitable to the biotype
in South Africa.
Biotype analysis is also interesting from an introduction point of view. By determining to which cluster hydrilla belongs, inferences about how it was introduced
to South Africa can be made. Hydrilla was introduced
into the USA via the aquarium trade (Schmitz et al.,
1991), and it is likely that this was also the mode of
introduction into South Africa. Interestingly, the majority of aquarium plants imported into South Africa come
from Singapore, Malaysia, which is where the South
African hydrilla biotype is most closely related.
Conclusions
Much progress has been made in understanding the biology of hydrilla, the nature of the infestation and the
potential for the weed to spread further in South Africa.
We also know what the biotype is, but little is known
about the flowering and reproductive phenology of this
biotype in South Africa. This calls for further study
in both the laboratory and the field in the upcoming
months. Furthermore, the certainty that the Hydrellia
flies and Bagous hydrillae will be suitable against the
South African biotype cannot be guaranteed. It seems
that the best way forward would be for South Africa
to undertake surveys in Sumatra and China, and other
Southeast Asian countries such as Thailand, Indonesia
and Malaysia, from where the biotype originates.
Acknowledgements
Angela Bownes (ARCPlant Protection Research Institute) is thanked for help in the field. The Invasive
Alien Species Programme of KwaZulu-Natal Department of Agriculture and Environmental Affairs is acknowledged for financial assistance.
References
Balciunas J.K., Grodowitz, M.J., Cofrancesco, A.F. and
Shearer, J.F. (2002) Hydrilla. In: Van Driesche, R., Lyon,
S., Blossey, B., Hoddle, M. and Reardon, R. (eds) Biological Control of Invasive Plants in the Eastern United
States. USDA Forest Service, Morgantown, WV, USA,
pp. 91114.
Bennett, C.A. and Buckingham, G.R. (2000) The herbivorous
insect fauna of a submersed weed, Hydrilla verticillata
(Alismatales: Hydrocharitaceae). In: Spencer, N.R. (ed.)
Proceedings of the X International Symposium on Biological Control of Weeds, July 414, 1999. Montana State
University, Bozeman, MT, USA, pp. 307313.
Center, T.D., Grodowitz, M.J., Cofrancesco, A.F., Jubinsky,
G., Snoddy, E. and Freedman, J.E. (1997) Establishment
of Hydrellia pakistanae (Diptera: Ephydridae) for the
biological control of the submersed aquatic plant Hydrilla
verticillata (Hydrocharitaceae) in the southeastern United
States. Biological Control 8, 6573.
Cook, C.D.K. and Lnd, R. (1982) A revision of the genus
Hydrilla (Hydrocharitaceae). Aquatic Botany 13, 485504.
Dayan, F.E. and Netherland, M.D. (2005) Hydrilla, the perfect aquatic weed, becomes more noxious than ever. Outlooks on Pest Management 16, 277282.
Grodowitz, M.J., Doyle, R. and Smart, R.M. (2000) Potential
use of insect biocontrol agents for reducing the competitive ability of Hydrilla verticillata. ERDC/EL SR-00-1.
US Army Engineer Research and Development Center,
Vicksburg, MS, USA, 27 pp.
Madeira, P.T., Van, T.K., Steward, K.K. and Schnell, R.J.
(1997) Random amplified polymorphic DNA analysis of
the phenetic relationships among world-wide accessions
of Hydrilla verticillata. Aquatic Botany 59, 217236.
Madeira, P.T., Van, T.K. and Center, T.D. (1999) Integration
of five Southeast Asian accessions into the world-wide
phenetic relationships of Hydrilla verticillata as elucidated by random amplified polymorphic DNA analysis.
Aquatic Botany 63, 161167.
Madeira, P.T., Coetzee, J.A., Center, T.D., White, E.E. and
Tipping, P.W. (2007) The origin of Hydrilla verticillata
recently discovered at a South African dam. Aquatic Botany 87, 176180.
Michel, A., Scheffler, B.E., Arias, R.S., Duke, S.O., Netherland,
M. and Dayan F.E. (2004) Somatic mutation-mediated
evolution of herbicide resistance in the non-indigenous
invasive plant hydrilla (Hydrilla verticillata). Molecular
Ecology 13, 32293237.
Overholt, B. and Wheeler, G. (2006) Renewed efforts to identify hydrilla biocontrol agents in Asia and Africa. Biocontrol News and Information 27, 5353.
Puri, A., MacDonald, G.E. and Haller, W.T. (2007) Stability
of fluridone-resistant hydrilla (Hydrilla verticillata) biotypes over time. Weed Science 55, 1215.
Schmitz, D.C., Nelson, B.V., Nall, L.E. and Schardt, J.D.
(1991) Exotic aquatic plants in Florida: A historical perspective and review of present aquatic plant regulation
program. In: Center, T.D., Doren, R.F., Hofstetter, R.L.,
667
XII International Symposium on Biological Control of Weeds
Myers, R.L. and Whiteaker, L.D. (eds) Proceedings of a
Symposium on Exotic Pest Plants, November 24, 1988,
Miami, FL. United States Department of the Interior, National Park Service, Washington, DC, USA, pp. 303336.
Wheeler, G.S. and Center, T.D. (2001) Impact of the biological control agent Hydrellia pakistanae (Diptera: Ephydridae) on the submersed aquatic weed Hydrilla verticillata
(Hydrocharitaceae). Biological Control 21, 168181.
668
Status of the biological control of
banana poka, Passiflora mollissima
(aka P. tarminiana) in Hawaii
R.D. Friesen,1 C.E. Causton2 and G.P. Markin3
Summary
Surveys were conducted between 1982 and 1995 on banana poka, Passiflora mollissima Bailey (also
known as P. tarminiana, subgenus Tacsonia) and related species in the Andes Mountains of South
America. The objective was to identify potential biocontrol agents for control of banana poka in
Hawaii, USA. Host-related insect diversity was greatest in Colombia, Ecuador, Peru and Venezuela,
and poorest in Bolivia and Chile. Insect species observed represented eight orders, 35 families and
approximately 67 species. Fifteen species were evaluated as potential biocontrol agents, of which
five received in-depth testing. Two moths, Cyanotrica necyria Felder and Rogenhofer (Lepidoptera;
Notodontidae) and Pyrausta perelegans Hampson (Lepidoptera; Pyralidae), were approved and released in Hawaii in 1988 and 1991, respectively; however, C. necyria did not establish and Pyr.
perelegans established but has had negligible impact. A third moth, Josia fluonia Druce, J. ligata
group (Lepidoptera; Notodontidae), had been approved for release and two flies, Dasiops caustonae
Norrbom and McAlpine (Diptera; Lonchaeidae) and Zapriothrica nr. salebrosa Wheeler (Diptera;
Drosophilidae), were undergoing final evaluation when the programme was terminated. A pathogen,
Septoria passiflorae Syd., was released in 1996, with mixed results. Banana poka remains a serious
weed pest in Hawaii.
Keywords: foreign exploration, South America, Passiflora tripartita, Passiflora tarminiana, weed biological control.
Introduction
Banana poka, Passiflora mollissima Bailey (also referred to in literature as P. tripartita and P. tarminiana),
was introduced into Hawaii from South America as an
ornamental around the year of 1900, but escaped domestication to become a major forest weed in the upper elevation mountain rain forests (La Rosa, 1984).
P. mollissima vines form dense mats of vegetation that
cover understory plants and climb into overstory cano-
Southern Kansas Cotton Growers Cooperative, Inc., PO Box 321, Winfield, KS 67156, USA.
2
Department of Terrestrial Invertebrates, Charles Darwin Research Station, A.P. 17-01-3891, Ecuador.
3
USDA Forest Service, Rocky Mountain Research Station, Bozeman,
MT 59717-2780, USA.
Corresponding author: R.D. Friesen <[email protected],
[email protected], [email protected]>.
CAB International 2008
1
pies of Koa, Acacia koa A. Gray, and Ohia, Metrosideros polymorpha Gaud., causing higher incidence of
tree limb breakage from their weight and wind blowdown (Warshauer et al., 1983; La Rosa, 1984). Feral
pigs feed on the fallen fruit, severely disturbing the soil
and surrounding plants by their rooting, and disseminating the seeds in their excrement.
By the early 1980s, banana poka was considered
the most serious threat to upper elevation forests of
Hawaii, severely infesting 520 km2 (Warshauer et al.,
1983). In 1982, the State of Hawaii appropriated funds
to the Department of Forestry and Wildlife (DOFAW)
for the evaluation of the potential for biological control
of P. mollissima, marking the beginning of a concerted
effort to control it.
This paper summarizes the results of foreign exploratory work from 1982 to 1996 in South America,
describes the most promising insects that were studied,
and identifies those having potential as biological control agents.
669
XII International Symposium on Biological Control of Weeds
Materials and methods
Passiflora species of the subgenus Tacsonia, which
include P. mollissima, occur throughout the Andes of
South America, from 2000 to 3200 m, where they are
cultivated for their fruits (Killip, 1938). P. mollissima
and its related forms are typically associated with humans. Surveys included commercial stands, single
plants or small stands around homes, and feral plants in
natural or disturbed areas. Commercial sites frequently
were exposed to regular applications of insecticide,
which reduced their utility for insect collection.
Sampling of plants included visual inspection of foliage and dissection of flower buds, open flowers, fruit,
roots, crowns and stems. Rearing and biological studies
were conducted in laboratories in Venezuela (Causton
et al., 2000), Colombia and Hawaii. Host specificity
testing was carried out in the Hawaii Volcanoes National Park (HVNP) insect quarantine facility. Voucher
specimens were submitted to the USDA ARS Systematics Entomology Laboratory (SEL) in Beltsville, MD,
as the definitive taxonomic authority. Many specimens
came back as unidentifiable species, but a few new
species descriptions were obtained. Specimens were
retained at the USDA FS Institute of Pacific Island Forestry in Hilo, HI.
Results
Species of the subgenus Tacsonia most commonly
observed above 2200 m were P. mollissima (curuba,
tumbo, taxo) and P. mixta L. Other species were only
rarely observed, for example, P. manicata L., P. pinnatistipula Cav. Below 2200 m, Tacsonia species were
replaced primarily by P. edulis Sims (i.e. maracuy or
passionfruit), P. ligularis Juss (granadilla amarilla) and
P. quadrangularis L. (badea). We included P. edulis in
our surveys due to concern over agent crossover to Hawaiian passionfruit, although its acreage and economic
importance is small (Martin, 1994).
Table 1 summarizes the species of insects encountered. Where possible, the findings reported by Pemberton (1989), Causton et al. (2000) and unpublished
reports by Markin and Friesen are included. The primary insect species studied during the course of the
Hawaiian programme are briefly described below.
Lepidoptera
Cyanotricha necyria Felder (Notodontidae): This
was the first species to be pursued as a biological
control agent due to its severe impact on commercial
production in western Colombia and central Ecuador,
where severe outbreaks frequently defoliated the plants
(Casaas-Arango et al., 1990). C. necyria was tested
and released in 1988 (Markin and Nagata, 1989; Markin et al., 1989), but failed to become established in
Hawaii (Markin et al., 1989; Campbell et al., 1993).
PyraustaperelegansHampson(Pyralidae=Crambidae):
This bud-feeder was widespread in Colombia, Ecuador,
Peru and Venezuela. In Colombia, Pyr. perelegans was
considered to be of major economic impact (Rojas and
Chacn, 1983). The biology of Pyr. perelegans is discussed by Rojas and Chacn (1982). Pyr. perelegans was
released in 1990 and is now established in Hawaii. However, populations have remained low and impact is negligible (Campbell et al., 1993; Markin and Nagata, 2000).
Acrocercops nr. pylonias Meyrick (Gracillariidae):
This leaf-miner was commonly observed attacking P.
mollissima and P. mixta in Colombia and northern Ecuador and possibly in Venezuela. In Colombia, damage
due to A. nr. pylonias was incidental at all sites except
for one near Pasto, where the leaf-miner was the dominant pest (Hugo Calvache, personal communication,
December 1988). Late instar larvae form a distinct
blotch that can cover several square centimeters of the
upper leaf surface. Impact of the blotches was perceived
small (Pemberton, 1989). This insect appeared promising and was scheduled for further field biology studies
at the time the programme was cancelled in 1996.
Odonna passiflorae Clarke (Oecophoridae): Larvae
of this species bore the vines and root crowns of mature
P. mollissima, that is, stems of 2550 mm diameter.
Multiple attacks in the root crown can kill the entire
plant (Chacon and de Hernandez, 1981). Unfortunately, Colombia was the only country where we found
O. passiflorae, and collecting the insect was difficult.
One stand of P. mollissima was discovered near Lake
La Cocha that suffered consistent losses of vines attributed to this pest species, but the planting was destroyed
before extensive studies could be concluded.
Heliconiidae: Agraulis vanillae L., Dione glycera
C&R Felder and D. juno Cramer: All three species
were commonly observed at many of the survey sites
and were polyphagous among Passiflora species, including P. edulis (Table 1). D. juno, a gregarious species of typically 2050 larvae per cluster, was capable
of denuding entire plants, while foliar damage due to
Agr. vanillae was only sometimes significant, primarily on very young vines. D. glycera was widely distributed, but larval density was always very low and
feeding damage negligible. Agr. vanillae, D. juno and
D. glycera were particularly susceptible to nuclear
polyhedrosis viruses (NPVs). Agr. vanillae is already
present in Hawaii, but has had no discernible impact
on P. mollissima.
Josia fluonia Druce (Notodontidae): Larvae of this
moth were observed feeding on foliage of P. mollissima only in central and northern Ecuador. The moths
are day-flying. Larval density was low, that is, usually
several larvae per plant, and feeding damage was typically light. J. fluonia was tested and cleared for release
in 1996, but the programme was terminated before releases were made.
Josia ligata group (Notodontidae): Larvae were observed feeding on foliage of P. mollissima and P. mani-
670
Status of the biological control of banana poka, Passiflora mollissima (aka P. tarminiana) in Hawaii
Table 1.
List of insects observed feeding on Passiflora species in the Andes of Bolivia, Chile (Ch), Colombia, Ecuador,
Peru, and Venezuela, 19821994. Information includes their order, family, insect species, Passiflora host species, stage host tissue, country. Table includes unpublished and previously published findings, e.g., Causton et al.
(2000), Pemberton (1989), from the Hawaiian biological control project.
Order
Coleoptera
Family
Buprestidae?
Cerambycidae
Chrysomelidae
Insect Species
Unidentified sp.
Unidentified sp.
near Hebestola sp.
near Lepturges sp.
Trachyderes sp.
Cassidinae (unident. sp.)
Diabrotica sp.
Epitrix sp.
Lactica brevicolis Jacoby
Lactica sp.(?)
Paralactica sp.(?)
Curculionidae
Elateridae
Lucanidae
Scarabidae
Scolytidae
Diptera
Drosophilidae
Lonchaeidae
Hemiptera
Brachyomus sp.
Compsus sp.
Cryptorhyncus cerdo
Fiedler
Exorides ?lajoyei Bovie
E. ?corrugatus Marshall
Pandeletius ?andeanus
Howden
Unidentified sp.
Unidentified sp.
Unidentified sp.
Unidentified sp.
Unidentified sp.
Zapriothrica nr salebrosa
Wheeler
Zapriothrica nr nudiseta
Wheeler
Dasiops caustonae Norrbom and McAlpine
Mycetophilidae
Dasiops curubae Steyskal
Dasiops gracilis Norrbom
and McAlpine
Dasiops inedulis Steyskal
Dasiops spp. (specimens
unident.)
Neosilba sp. (poss N. certa
Walker)
Mycetophila spp.
Coreidae
Leptoglossus sp.
Passiflora host
species
Stage-Host tissue
Countrya
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mixta
P. mollissima
P. pinnatistipula
P. mollissima
P. mollissima
P. mixta
P. manicata
P. mollissima
P. mollissima
P. mollissima
P. mollissima
larvaelive stem
larvaedead crown
larvaedead stems
larvaedead stems
larvaedead stems
adultsleaves
adultsleaves
adultsleaves
adultsleaves
adultsleaves
adultsleaves, flowers
adultsleaves, flowers
adultsleaves, flowers
adultsleaves, flowers
adultsleaves, flowers
adultsleaves, terminals
adultsleaves, terminals
adultsleaves, terminals
V
C
V
V
V
V
C, V
E
P, V
P
V
P
E
E
E, V
V
V
V
P. mollissima
P. mollissima
P. mollissima
adultsleaves, terminals
adultsleaves, terminals
adultsleaves, terminals
V
V
V
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
V
V
B
V
V
P. mollissima
adults?
adults?
larvaeroots
adults?
larvae, adultsstem,
branches
larvaeflower buds
E, V, C
P. mollissima
larvaeflower buds
P. mixta
P. mollissima x
P. exoniensis
P. manicata
larvaeflower buds
larvaeimm. fruit
E, V
V
larvaeimm. fruit
P. mixta
P. mollissima
P. mollissima x
P. exoniensis
P. mollissima
P. edulis
larvaeimm. fruit
larvaeimm. fruit
larvaeimm. fruit
V
B, E, V
V
larvaeflower buds
larvaeimm. fruit
B, E
V
P. edulis
P. mollissima
larvaeflower buds
larvaeflower buds
B, V
C
P. mollissima
larvaeimm. fruit
P. mollissima
P. mixta
P. mollissima
larvaeflower buds
larvaeflower buds
adults, nymphsfruit,
stems
C, E, V
E
V
(continued on next page)
671
XII International Symposium on Biological Control of Weeds
Table 1.
(Continued) List of insects observed feeding on Passiflora species in the Andes of Bolivia, Chile (Ch), Colombia,
Ecuador, Peru, and Venezuela, 19821994. Information includes their order, family, insect species, Passiflora
host species, stage host tissue, country. Table includes unpublished and previously published findings, e.g., Causton et al. (2000), Pemberton (1989), from the Hawaiian biological control project.
Order
Hymenoptera
Homoptera
Lepidoptera
Family
Insect Species
Pentatomidae
Pyrrhocoridae
Tingidae
Apidae
Cicadellidae
Cercopidae
Pseudococcidae
Unidentified sp.
Unidentified sp.
Unidentified sp.
Unidentified sp.
Trigona sp.
Unidentified spp. (various)
Unidentified sp.
Unidentified sp.
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
P. mollissima
Coccidae
Unidentified sp.
P. mollissima
Arctiidae
Lophocampa sp.?
Turuptiana neurophylla
Turuptiana
sanguinipectus(?) Seitz
Unidentified sp.
Unidentified sp.
Unidentified sp.
Acrocercops sp. near
pylonias
Coleophoridae
Geometridae
Gracillariidae
Heliconiidae
Passiflora host
species
Unidentified sp.
Agraulis vanillae L.
Dione glycera C&R
Felder
Dione juno Cramer
Noctuidae
Notodontidaeb
Euptoieta hegesia
Comstock
Copitarsia sp. (?)
Copitarsia consueta
Cyanotricha necyria
Felder
Josia fluonia Druce
[not released]
Josia ligata group
Oecophoridae
Psychidae
Pyralidae
(= Crambidae)
Odonna passiflorae Clarke
Unidentified sp.
Pyrausta perelegans
Hampson
672
Stage-Host tissue
Countrya
E
V
V
B
V
V
V
V
P. mollissima
P. mollissima
P. mollissima
adultsunknown
adultsfruit?
adultsfruit?
adults, nymphsleaves
adultsflowers
adultsleaves
adultsleaves
adults, immatures(?)
leaves
adults, immatures(?)
leaves
larvaeleaves
larvaeleaves
larvaeleaves
P. mollissima
P. mollissima
P. mollissima
P. mollissima
larvaeleaves
larvaeleaves
larvaeleaves, flowers
larvaeleaves
V
V
B, E, V
C, E, P
P. mixta
P. mollissima
P. mollissima
larvaeleaves
larvaeleaves, fruits
larvaeleaves
P. edulis
P. ligularis
P. manicata
P. alata
larvaeleaves
larvaeleaves
larvaeleaves
larvaeleaves
E
V
B, C, E,
P, V
E, V
E, P
E, P
V
P. edulis
P. ligularis
P. mollissima
larvaeleaves
larvaeleaves
larvaeleaves
P. edulis
P. ligularis
P. manicata
P. mollissima
P. mollissima
larvaeleaves
larvaeleaves
larvaeleaves
larvaeleaves
larvaeleaves
P. mollissima
E, V, P
P. mollissima
P. manicata
larvaeleaves,
flower buds,
larvaeflowers
larvaeleaves
P. mollissima
P. mollissima
larvaeleaves
larvaeleaves
C, E, P
E
P. mollissima
P. manicata
P. mollissima
P. mollissima
P. mixta
larvaeleaves
larvaeleaves
larvaestems
larvaeleaves
larvaeflower
buds, fruit
stem tips
E
E
C
B, V
E, P, V
V
V
C
E
V
V
B, C,
Ch, E,
P, V
E, P, V
E
E
V, E
V
C
E
Status of the biological control of banana poka, Passiflora mollissima (aka P. tarminiana) in Hawaii
Table 1.
(Continued) List of insects observed feeding on Passiflora species in the Andes of Bolivia, Chile (Ch), Colombia,
Ecuador, Peru, and Venezuela, 19821994. Information includes their order, family, insect species, Passiflora
host species, stage host tissue, country. Table includes unpublished and previously published findings, e.g., Causton et al. (2000), Pemberton (1989), from the Hawaiian biological control project.
Order
Family
Insect Species
Passiflora host
species
P. mollissima
Lepidoptera
Orthoptera
Thysanoptera
a
b
Saturniidae
Tortricidae
Unidentified sp.
Unidentified sp.
P. mollissima
P. mollissima
Unidentified
Acrididae
unidentified
Unidentified sp.
Unidentified sp.
Meridacris subaptera
Unidentified sp.
P. mollissima
P. mollissima
P. mollissima
P. mollissima
Stage-Host tissue
larvaeleaves,
flower buds,
stem tips
larvaeleaves
larvaeleaves,
stems, stem tips
larvaeleaves
larvaestems
adultsleaves
adults, nymphs
unknown
Countrya
C, E,
V, P
V
V
V
V
V
V
B = Bolivia; Ch=Chile; C=Colombia; E=Ecuador; P=Peru; V=Venezuela
For the family Notodontidae, we used the classification of Miller (1996).
cata only in central and northern Ecuador. Larvae of
J. ligata were very similar to J. fluonia larvae in appearance, behaviour and feeding damage. However, J.
ligata was found to be able to complete development
on several Passiflora species, including P. edulis, dropping it from further consideration.
Diptera
Zapriothrica nr. salebrosa Wheeler (Drosophilidae):
This flower bud-attacking fly was probably the most
common and widely distributed insect found during
our surveys and has long been recognized as a pest of
P. mollissima (Chacon and Rojas, 1984; A.D. Casaas,
1984, unpublished results); it was observed in Colombia, Ecuador, Peru and Venezuela. However, CasaasArango et al. (1996) and Causton et al. (2000) mention
another similar species, Z. nr. nudiseta, attacking only
P. mollissima in Colombia, suggesting that taxonomic
review of this group may be necessary. Pemberton
(1989) identified Z. nr. salebrosa as a candidate agent
and preliminary studies were conducted in the field
in Colombia. The biology of Zapriothrica sp. is discussed by Casaas-Arango et al. (1996). The release
of Pyr. perelegans, another bud-feeder, and studies of
a fruit-attacking fly Dasiops caustonae Norrbom and
McAlpine (Diptera; Lonchaeidae) lead to putting studies of Z. nr. salebrosa on hold.
Dasiops species (Lonchaeidae): Species of the genus
Dasiops have long been recognized as major pests of
cultivated species of Passiflora (Posada et al., 1976;
Chacon and Rojas, 1984). Lonchaeid larvae were regularly encountered in flower buds or in the developing fruit in all of the countries surveyed except Chile
(Table 1), often causing significant losses of fruit
bodies. Field identification of the species proved to
be impossible. A taxonomic review of Dasiops species associated with Passiflora described three of
the five species we collected from P. mollissima as
capable of attacking P. edulis, that is D. curubae, D.
gracilis and D. inedulis (Norrbom and McAlpine,
1997). A newly described species, D. caustonae Norrbom and McAlpine, appeared to be confined to Passiflora species in the subgenus Tacsonia, except for
P. manicata, and was the only insect found attacking
the developing fruit. Its biology is discussed by Causton and Rangel (2002). Attempts to colonize Dasiops
species in quarantine in Hawaii failed due to our inability to induce mating, although oviposition of sterile
eggs readily occurred.
Mycetophila (Mycetophilidae): At least two species
of these fungus gnats were observed attacking flower
buds in Northern Ecuador , Colombia and Western Venezuela; bud loss was often very significant. Specimens
of adults could only be determined to the genus level
(Gagn, personal communication, 1995). Multiple larvae were found in each infested bud, with up to 17 in
one bud (Causton et al., 2000). The larvae were very
sensitive to disturbance and/or dehydration, as healthy
larvae within dissected buds usually died shortly after
inspection.
Coleoptera
Longhorned Beetles (Cerambycidae): Cerambycids
were collected from mature vines of P. mollissima, that
is, older than 8 years, at two locations. Larvae of an
unidentified species, 12 cm long, were recovered from
root crowns of dead vines in the Lake La Cocha area in
Colombia. At least three species of cerambycids, near
Hebestola sp., near Lepturges sp. and Trachyderes sp.,
were recovered from dead or dying branches of vines
673
XII International Symposium on Biological Control of Weeds
in Venezuela (Table 1). Further investigation will be
necessary to confirm their identities and feeding type,
that is, as primary or secondary.
Leaf Beetles (Chrysomelidae): Adults of several species of leaf beetles were observed feeding on foliage
and/or blooms of several Passiflora species in Colombia, Ecuador, Peru and Venezuela, but no eggs or larvae
were recovered in the field and breeding colonies could
not be established in the laboratory. Adults of Diabrotica sp. and Epitrix sp. were quite common in Venezuela,
but were observed to be polyphagous (Causton et al.,
2000).
Discussion
Except for the fungal pathogen, Septoria passiflorae
Syd. released in 1996, no new insect agents have been
released since Pyr. perelegans in 1990. However, progress was made in identifying several promising candidates and eliminating others. Below is a ranking and
brief justification of the agents we consider most promising for future work: (1) Z. nr. salebrosa/Z. nr nudiseta
because of their potential impact on reproduction of P.
mollissima. Their biology and colonization techniques
have already been determined, although species identities and distinctions need to be clarified; (2) O. passiflorae was capable of killing mature plants. Aspects
of its biology and a local collaborator for collection
of the species are known, which could facilitate initiation of biological studies; (3) D. caustonae and (4)
Mycetophila sp. due to their potential to significantly
impact reproduction of P. mollissima. Before studies
may progress, the problem of establishing reproducing
colonies in captivity needs to be solved. Host testing in
the insects country of origin could bypass this problem; and (5) the unidentified lepidopteran stem borer
from Venezuela (Causton et al., 2000). Like O. passiflorae, this species is recognized as having potential
for significantly increasing plant mortality in Hawaii.
However, this insect species was only observed once
throughout the duration of the 4-year Venezuelan study.
Confirmation of its identity remains first priority. Other
candidates of lower priority but of potential interest include the chrysomelid beetles from northern Ecuador
and cerambycid beetles from Colombia and Venezuela. As a note, although C. necyria and Pyr. perelegans
were not successful in Hawaii, they may be well suited
for different environments.
A third biological control agent, the fungal pathogen
S. passiflorae, was also identified during the Hawaiian programme, tested and released in 1996 (Trujillo,
2001). The pathogen is now established through parts
of the range of P. mollissima in Hawaii and is credited
with giving substantial reduction in biomass by causing
early defoliation in certain areas (Trujillo, 2001, 2004).
However, in parts of the range, the pathogen is ineffective or is no longer giving adequate control (Markin,
personal communication, May 2006).
P. mollissima remains an invasive exotic species in
Hawaii and in other parts of the world. If, in the future
others should decide to attempt biocontrol of P. mollissima, we hope that this summary of our survey results
will be of value.
Acknowledgements
This project was sponsored by the USDA Forest Service Institute of Pacific Island Forestry (USDA FS
IPIF) and the State of Hawaii through funding provided to the Hawaii Department of Forestry and Wildlife
(DOFAW) from 1982 to 1993. We especially thank
project leaders Gene Conrad (USDA FS IPIF) and
Victor Tanimoto (DOFAW). In South America, we are
particularly indebted to A.M. Rojas de Hernandez and
Anadelfa D. Casaas (Univ. del Valle, Cali, Colombia); Armando Briceo (Univ. de Los Andes, Mrida,
Venezuela); Danilo Silva (Inst. Para la Invest. de una
Produccin Tropical, Mrida, Venezuela); Jaime Jaramillo and Giovanni Onore (Pont. Univ. Catlica del
Ecuador, Quito, Ecuador); Eugenio Dussoulin Escovar
(Univ. de Tarapaca, Quito, Ecuador); Jaime Sarmiento
and Vivianna Baptista (Nacional Coleccin de Fauna
Boliviana, La Paz, Bolivia); Jorge Caballero and Edwin Butron (Caritas Boliviana, La Paz, Bolivia); Carlos Alarcon and Juan Villarroel (Univ. Mayor de San
Simn, Cochabamba, Bolivia); Jose Luis Isurza and
Concepcin Paredes of CARE (Sucre, Bolivia). Very
special thanks to D.C. Ferguson, R. Gagn, C. Ville
Lelah, J.F. Macklepine, James Miller, Allan Norrbom,
R.W. Poole and R.E. White for their invaluable assistance with insect identifications.
References
Campbell, C.L., Markin, G.P. and Johnson, M.W. (1993) Fate
of Cyanotricha necyria (Lepidoptera:Notodontidae) and
Pyrausta perelegans (Lepidoptera:Pyraustida), released
for the biological control of banana poka (P. mollissima)
on the Island of Hawaii. Proceedings of the Hawaiian Entomological Society 32, 123130.
Casaas-Arango, A.D., Trujillo, E.E., Rojas de Hernandez,
A.M. and Taniguchi, G. (1990) Field biology of Cyanotricha necyria Felder (Leptidopter; Dioptidae), a pest of
Passiflora species in southern Colombia and Ecuadors
Andean Region. Journal of Applied Entomology 109,
9397.
Casaas-Arango, A.D., Trujillo, E.E., Friesen, R.D. and Rojas
de Hernandez, A.M. (1996) Field biology of Zapriothrica
sp. Wheeler (Diptera; Drosophilidae), a pest of Passiflora
spp. of high elevation possessing long tubular flowers.
Zeitschrift fur Angewandt Entomologie 120, 111114.
Causton, C.E. and Rangel, A.P. (2002) Field observations on
the biology and behaviour of Dasiops caustonae Norrbom
and McAlpine (Dipt., Lonchaeidae), as a candidate biocontrol agent of Passiflora mollissima in Hawaii. Journal
of Applied Entomology 126, 169174.
674
Status of the biological control of banana poka, Passiflora mollissima (aka P. tarminiana) in Hawaii
Causton, C.E., Markin, G.P. and Friesen, R. (2000) Exploratory survey in Venezuela for biological control agents of
Passiflora mollissima in Hawaii. Biological Control 18,
110119.
Chacon, P. and de Hernandez, M. (1981) Immature stages of
Odonna passiflorae Clarke (Lepidoptera: Oecophoridae):
Biology and morphology. Journal of Research on Lepidoptera 20, 4345.
Chacon, P. and Rojas, A.M. (1984) Entomofauna asociada a
P. mollissima, P. edulis f flavicarpa, y P. quadrangularis
en El Departamento Del Valle del Cauca. Turrialba 34(3),
297311.
Killip, E.P. (1938) The American species of Passifloraceae.
Field Museum of Natural History (Chicago) Botanical
Series 19.
La Rosa, A.M. (1984) The Biology and Ecology of Passiflora
mollissima in Hawaii. Cooperative National Park Studies
Unit, University of Hawaii at Manoa, Department of Botany, Technical Report No. 50, 168 pp.
Markin, G.P. and Nagata, R.F. (1989) Host preference and
potential climatic range of Cyanotricha necyria (Lepidoptera: Dioptidae), a potential biological control agent of the
weed Passiflora mollissima in Hawaiian forests. University of Hawaii at Manoa, Department of Botany, National
Park Service Technical Report No. 67, 35 pp.
Markin, G.P. and Nagata, R.F. (2000) Host suitability studies of
the moth, Pyrausta perelegans Hampson (Lepidoptera: Pyralidae), as a control agent of the forest weed banana poka,
Passiflora mollissima (HBK) Bailey, in Hawaii. Proceedings of the Hawaiian Entomological Society 34, 169179.
Markin, G.P., Nagata, R.F. and Taniguchi, G. (1989) Biology
and behavior of the South American moth, Cyanotricha
necyria (Felder and Rogenhofer) (Lepidoptera: Notodontidae), a potential biological control agent in Hawaii of the
forest weed, Passiflora mollissima (HBK) Bailey. Proceedings of the Hawaiian Entomological Society 29, 115123.
Martin, D.A. (1994) Statistics of Hawaiian Agriculture.
Hawaii State Department of Agriculture, Honolulu, HI,
USA, 100 pp.
Norrbom, A.L. and McAlpine, J.F. (1997) A revision of the
neotropical species of Dasiops rondani (Diptera: Lonchaeidae) attacking Passiflora (Passifloraceae). Memoir
Entomological Society of Washington 18, 189211.
Pemberton, R.W. (1989) Insects attacking Passiflora mollissima and other Passiflora species: Field survey in the Andes. Proceedings of the Hawaiian Entomological Society
29, 7184.
Posada, I.O., De Polonia, I.Z., De Arevalo, I.S., Saldarriage,
A.V., Garcia, F.R. and Cadenas, R.E. (1976) Lista de insectos danios y otras plagas en Colombia. Boletn tchnica No. 43 Oct. Instituto Colombiano Agropecuario,
Bogot, Colombia, pp. 337342.
Rojas de Hernndez, M. and Chacn de Ulloa, P. (1982)
Contribucin a la Biologa de Pyrausta perelegans
Hampson (Lepidoptera: Pyralidae). Brenisia 1920,
325331.
Rojas de Hernndez, M. and Chacn de Ulloa, P. (1983) Entomofauna Asociada al Cultivo de la Curuba en El Departamento del Valle. Coagro 45, 2127.
Trujillo, E.E. (2001) Effective biomass reduction of the invasive weed species banana poka by Septoria leaf spot.
Plant Diseases 85, 357361.
Trujillo, E.E. (2004) History and success of plant pathogens
for biological control of introduced weeds in Hawaii. Biological Control 33, 113122.
Warshauer, F.R., Jacobi, J.D., La Rosa, A.M., Scott, J.M.
and Smith, C.W. (1983) The distribution, impact, and
potential management of the introduced vine, Passiflora mollissima (Passifloraceae) in Hawaii. Coop.
National Park Studies Unit, University of Hawaii at
Manoa, Department of Botany, Technical Report No.
48, 39 pp.
675
A cooperative research model biological
control of Parkinsonia aculeata and
Landcare groups in northern Australia
V.J. Galea1
Summary
Parkinsonia, Parkinsonia aculeata L., is a woody shrub, which is classed as a weed of national significance in Australia. It is considered a major threat to both managed and natural ecosystems. Research
into the cause of a dieback disorder in Parkinsonia occurring at locations across northern Australia has
identified a range of fungal organisms to be associated with affected plants. Currently, these are being evaluated in conventional field trials at locations in north Queensland and the Northern Territory.
A cooperative research model has been developed to allow regional Landcare groups to participate
in this research programme. This standardized model for medium-scale trials will enable Landcare
groups to establish, monitor and evaluate the performance of a range of potential biological control
agents under local conditions. The development of a research kit is a key element of this programme.
The kit will include equipment needed to establish the trial and the fungal agents to be evaluated. An
instruction manual will outline the procedures required to select an appropriate trial site and provide
instructions on inoculation, data collection and ongoing maintenance of the trial. This cooperative
approach will both engage and enable Landcare groups in the development of solutions for their
regions.
Keywords: cooperation, research, Parkinsonia.
Introduction
Parkinsonia, Parkinsonia aculeata L., is a woody shrub,
which is classed as a Weed of National Significance
(WoNS) in Australia. It is considered a major threat to
both managed and natural ecosystems (Deveze et al.,
2004). Parkinsonia currently infests approximately 1
million ha of land, mainly along watercourses throughout northern Australia. Parkinsonia severely degrades
the economic and environmental value of land that it
invades (Deveze et al., 2004). The spread of this species threatens biodiversity, the health of river systems
and wetland areas and the productivity of pastoral enterprises (personal observation). It forms dense impenetrable thickets, which reduce native fauna habitat and
impede mustering activity and stock access to water
(Diplock et al., 2006). It replaces native and pasture
species and reduces the carrying capacity and productivity of pastoral land.
School of Land, Crop and Food Sciences, University of Queensland,
Gatton Campus, Gatton, QLD 4343, Australia <[email protected]>.
CAB International 2008
1
Biological control of Parkinsonia is a key element
of strategic management (van Klinken, 2006). While
chemical treatment of minor infestations is often effective, the establishment of biological control agents may
promise to give lasting control.
Parkinsonia dieback is a disorder, which causes
Parkinsonia plants to dieback from the tips; the leaves
droop, turn brown, but remain attached to the plant,
which eventually dies. Dieback progresses through
Parkinsonia populations as a front and kills both adult
and juvenile plants (Diplock et al., 2006). The disease
appears to be naturally occurring in the Northern Territory, Western Australia and Central Queensland.
There appear to be four key fungal organisms associated with Parkinsonia dieback, which are either native
fungal species, or species that are naturalized and now
widespread (Diplock et al., 2006). Field observations
and historical reports suggest that this dieback disorder
has the potential to be harnessed as a self-replicating
and potentially, self-dispersing biological control tool.
A better understanding of the way in which the fungal
species infect the plant causing disease and death, and
the related factors that contribute to disease develop-
676
A cooperative research model biological control of Parkinsonia aculeata
ment (i.e. environmental stresses, plant density and
age, impact of management practices, fire, etc.) could
lead to the development of an integrated management
programme for Parkinsonia involving the use of these
potential biological control agents as a key element.
Postgraduate research (Diplock et al., 2006) carried out by the University of Queensland (UQ) into
the cause of the dieback disorder in Parkinsonia has
identified a range of fungal organisms to be associated
with affected plants. A culture bank developed for this
project contains over 200 isolates taken from field affected Parkinsonia plants collected from across northern Australia. Within this collection, four key genera
(Fusarium, Lasiodiploidia, Phoma and Fusicoccum)
have been identified as being widely distributed and are
being treated as organisms of interest. Currently, these
four genera are being evaluated in conventional field
trials at locations in north Queensland and the Northern Territory and in laboratory and glasshouse studies
at UQ.
Objectives
To build on previous and ongoing university research
and to serve as a model for a national approach for
the cooperative development of knowledge and the
creation of regional capacity for pathogen-based
woody weed biocontrol programmes;
To develop a standardized field research model
including training and resources to build the capacity of stakeholders to undertake Parkinsonia
dieback trials;
To collect data that will contribute to the long-term
research into Parkinsonia biological control;
To communicate the outcomes of the project to the
community and other researchers; and
To develop appropriate delivery technologies for
Parkinsonia dieback pathogens that will be appropriate for on-ground managers.
Methodology
This project involves trials being set up within two National Parkinsonia Management Zones. Thirty sites are
to be set up in total with resources available to set up
more sites if required. Fifteen sites will be established
in Zone A (Victoria River District) and 15 in Zone B
(Barkly region). This project will develop a standardized model for medium-scale trials (based on methods
currently used for a PhD programme on Parkinsonia
dieback) which will enable Landcare groups to establish, monitor and evaluate the performance of a range
of potential biological control agents under local conditions. A support and communication officer will be appointed to develop materials and support stakeholders.
The current inoculation method involves the introduction of a formulated inoculum pellet into the stem of
Parkinsonia trees. Protocols for site and tree selection,
assessment of critical parameters for tree health, inoculation and site mapping have been established, but may
be modified after consultation with client groups.
There are seven key elements to this programme:
1. D
evelopment of an instruction manual. An instruction manual will be prepared, outlining the procedures required to select an appropriate trial site
and outline the procedure for selecting appropriate
trees, and performing pre- and postinoculation assessment for size and vigour. The manual will also
provide instructions on how to inoculate trees, postinoculation data collection (both quantitative and
qualitative) and ongoing maintenance of the trial.
2. Preparation of project kits. Project kits will be
prepared for each of the cooperator teams. These
will include the Instruction Manual along with some
of the materials required to establish the trial, that
is, the fungal inoculum in a stable form, tree tags, a
water squirt bottle, PVC tape to seal the tree wound,
flexible tape measure for stem circumference measurement, and additional data sheets. The kits will
be prepared and mailed out as required. Additional
inoculum and tags can be supplied for groups wishing to establish more than one trial site.
3. Running instructional workshops. Two training
workshops/field days will be conducted to outline
the procedures and rationale for this work as well
as assessment and data management procedures.
These workshops will also provide an opportunity
to gain information from land and landscape managers about Parkinsonia and their expectations from
the cooperative research programme.
4. Ongoing support for trials. Support for trials will
be available through e-mail and telephone communication. A programme coordinator and project
support officer will be available to visit trial sites in
each region.
5. Use of data. Data collected by cooperators will be
used internally in property management planning.
Additionally, data will be shared among groups and
also shared with the overall project coordinator to
contribute to the knowledge base on this biocontrol
system, assisting UQ research. The type of data to
be collected includes plant vigour before and after
inoculation with a pathogen, stem circumference
and symptoms of plant disease such as loss of stem
integrity, presence of basal stem lesions, or other related symptoms. Collection of (digital) photographic
evidence will also be supported.
6. Reporting and communications. Outcomes from
the trials will be shared by a biannual newsletter
to be distributed by electronic means. Additionally,
outcomes will be presented at appropriate sym
posia and conferences and published in appropriate newsletters (Barkly Beef, Weed All About it,
Barkly Land Care Association and Victoria River
677
XII International Symposium on Biological Control of Weeds
District Conservation Association newsletter, Network Notes) to ensure maximum possible sharing
of information.
7. Extension of new knowledge from laboratory
research. As new biocontrol isolates or inoculation
techniques become available from the associated
UQ projects, these will be communicated to the cooperator groups, and where possible made available
for field testing.
Through the involvement of local stakeholders, it will
be easier to achieve research goals. Training workshops
and the production of kits will increase the capacity of
land managers to undertake Parkinsonia dieback trials.
This project will raise awareness of adoption of integrated weed management practices.
Significance of expected outcomes
Identified collaborators
Strong support for this project has been demonstrated
by existing collaborators and new potential collaborators. Victoria River District Conservation Association
(VRDCA) and Roper River Landcare Group (RRLG)
are currently supporting the project with preliminary
sites already set up. Members of the VRDCA and the
RRLG are currently being trained in inoculation techniques and plant vigour assessment.
The Weed Management Section of the Department
of Natural Resources, Environment and the Arts (NRETA) will also support the project by providing technical advice and assistance to all stakeholders, attending
workshops and participating in setting up and monitoring trial sites.
Aboriginal Ranger Groups (Muru-warinyi Ankkul
Rangers) are keen to participate in the project to increase skills, build capacity among the groups and improve networks between stakeholders.
The Department of Primary Industry, Fisheries and
Mines (DPIFM) and the Northern Territory Cattlemens
Association (NTCA) also support this project as Parkinsonia impacts on pastoral operations and the pastoral industry within the Northern Territory. Both support
research into effective biological control agents for use
as a tool.
Justification
Current research on Parkinsonia dieback being conducted by UQ has over the past 2 years attracted significant attention among landscape managers from Parkinsonia-affected regions and researchers involved in
woody weed management. Through the involvement
of VRDCA, RRLCG, NRETA Weed Management
and Aboriginal ranger groups, UQ will deliver a project that achieves a strong community-based landcare
movement through improving communication and collaboration between land managers across a range of
tenures. This project involves local stakeholders, that
is, Landcare groups, aboriginal rangers, NRETA weeds
officers and UQ creating a cooperative collaborative
approach across regions and state boundaries. The
partnerships created will improve regional planning.
Parkinsonia biological control is still in development.
A partnership between Government agencies,
regional community leaders, indigenous ranger
groups and pastoralists will promote implementation of VRD components of the Katherine Regional
Weed Management Plan (M. Kassman, personal
communication, 2006). This ensures that all stakeholders have a voice in the execution of strategic weed management, supporting the VRDCA
model of holistic management in the spheres of
production, community and conservation, and the
Integrated Natural Resource Management Plan
(Northern Territory) vision of fostering strong local contribution to natural resource management.
Participation in, and visual confirmation of, the efficacy of sustained weed control programmes will
provide an impetus for further long-term planning
and investment from stakeholders in the regions.
This project directly contributes to the goal of sustainable natural resources and community capacity building by integrating planning, training and
action.
Concluding comments
Experience from previously established pilot trials
with the Barkly Landcare and Conservation Association (BCLA) and the Roper River Landcare Group
(RRLG) have indicated both a high level of willingness
to engage in cooperative research and accelerated outcomes through the sharing of knowledge and experience held by such groups. This approach has proven
to be an excellent way of guiding research to ensure
that outcomes are more appropriately aligned with the
needs of landholders and landscape managers. Adoption of outcomes will be greater for cooperatively developed management strategies; furthermore, this approach may ensure that localized environmental and
operational conditions are considered.
Acknowledgements
The author acknowledges the support of the following
organisations: Muru-warinyi Ankkul Rangers Central
Land Council, Tennant Creek, NT; Victoria River District Conservation Association, Katherine, NT; Australian Agricultural Company (AAC), Anthony Lagoon
Station, NT; Department of Natural Resources, Envi-
678
A cooperative research model biological control of Parkinsonia aculeata
ronment and the Arts, NT; Barkly Landcare and Conservation Association, NT; and Roper River Landcare
Group, Katherine, NT.
References
Deveze, M., March, N. and van Klinken, R. (2004) Parkinsonia ecology and threat. National Case Studies Manual
Parkinsonia, Approaches to the Management of Parkinsonia (Parkinsonia aculeata) in Australia. Department of
Natural Resources, Mines and Energy, Brisbane, QLD,
Australia, pp. 210.
Diplock, N., Galea, V., van Klinken, R. and Wearing, A.
(2006) A preliminary investigation of dieback on Parkinsonia aculeata. In: Preston, C., Watts, J.H. and Crossman,
N.D. (eds) 15th Australian Weeds Conference Proceedings: Managing Weeds in a Changing Climate. Weed
Management Society of SA Australia, pp. 585587.
van Klinken, R.D. (2006) Parkinsonia biocontrol: What are
we trying to achieve? Australian Journal of Entomology
45, 268271.
679
A global view of the future for biological
control of gorse, Ulex europaeus L.
R.L. Hill,1 J. Ireson,2 A.W. Sheppard,3 A.H. Gourlay,4 H. Norambuena,5
G.P. Markin,6 R. Kwong7 and E.M. Coombs8
Summary
Gorse (Ulex europaeus L.) has become naturalized in at least 50 countries outside its native range,
from the high elevation tropics to the subantarctic islands and Scandinavia. Its habit, adaptability and
ability to colonize disturbed ground makes it one of the worlds most invasive temperate weeds. It is
80 years since New Zealand first initiated research into biological control for gorse. This paper briefly
reviews the progress made worldwide since then, and examines future opportunities for biological
control of this weed. The range of available agents is now known, and this list is critically assessed.
Ten organisms have been released variously in six countries and islands and their performance is
reviewed. In most cases, agent populations have been regulated either from top-down or bottomup, and there is no evidence anywhere of consistent outbreaks that could cause significant reduction in existing gorse populations in the medium term. Habitat disturbance and seedling competition
are important drivers of gorse population dynamics. Existing agents may yet have long-term impact
through sublethal effects on maximum plant age, another key factor in gorse population dynamics.
Along with habitat manipulation, seed-feeding insects may yet play a long-term role in reducing seed
banks below critical levels for replacement in some populations. In the short term, progress will rely
on rational and integrated weed management practices, exploiting biological control where possible.
Keywords: integrated weed management, population dynamics, modelling.
Introduction
Gorse, Ulex europaeus L. (Fabaceae), is a thorny shrub
native to the temperate Atlantic coast of Europe and the
British Isles including Ireland. It has become naturalized elsewhere in Europe, North Africa and the Middle
Richard Hill and Associates, Private Bag 4704, Christchurch, New
Zealand.
2
Tasmanian Institute of Agricultural Research, 13 St Johns Street, New
Town, TAS 7008, Australia.
3
CSIRO Entomology, GPO Box 1700, ACT 2601, Australia.
4
Landcare Research, PO Box 40, Lincoln 7640, New Zealand.
5
Instituto de Investigaciones Agropecuarias Carrillanca, PO Box 58-D,
Temuco, Chile.
6
USDA Forest Service, Bozeman Forestry Sciences Lab, 1648 S.
7th Avenue, Bozeman, MT 59717, USA.
7
Department of Primary Industries, Frankston Centre, Ballarto Road,
VIC 3199, Australia.
8
Oregon Department of Agriculture, 635 Capitol Street NE, Salem, OR
97301-2532, USA.
Corresponding author: R.L. Hill <[email protected], john.ireson@
dpiw.tas.gov.au, [email protected], gourlayh@landcareresearch.
co.nz, [email protected], [email protected], [email protected].
gov.au, [email protected]>.
CAB International 2008
1
East. In other parts of the world, gorse has proven to be
an aggressive invader, forming impenetrable, largely
monotypic stands that reduce access of grazing animals
to fodder, modify native ecosystems and ecosystem
processes, and outcompete trees in developing forests.
It has now been recorded in more than 50 countries and
islands, and is considered to be a major weed in New
Zealand (Hill et al., 2000), the USA (Markin et al.,
1995, 1996), Chile (Norambuena et al., 2007) and Australia (Ireson et al., 2006).
Gorse is tolerant of a wide range of conditions, but
in temperate latitudes, it is limited altitudinally by cold
temperatures. This has not stopped it becoming established in the montane regions of tropical Hawaii, Sri
Lanka and Costa Rica, and it grows in Scandinavia, on
St. Helena, on some subantarctic islands, and in coastal
NE USA, where maritime influences moderate the
climate. New Zealand is one of the few places where
gorse has largely achieved its maximum distribution.
Land managers on intensively managed land with
adequate economic returns have many excellent alternatives for managing gorse. However, biological control may provide the only option for limiting the effects
680
A global view of the future for biological control of gorse, Ulex europaeus L.
of gorse on land that provides low economic returns,
land that is managed for biodiversity values, or where
the infestation is simply too difficult to manage.
It is now more than 80 years since research into the
biological control of gorse in New Zealand was first
commissioned (Zwlfer, 1963), and the first biological
control agent, Exapion ulicis Forster (Brentidae), was
first released on Maui, HI, in 1926 (Markin et al., 1996).
Ten biological control agents have now been released
as classical biological control agents in six countries or
regions: two seed-feeding insects and eight organisms
that attack green stems (Table 1). It is now time to examine the state of gorse biological control worldwide,
and its future role in gorse management. The purpose
of this paper is to:
evaluate the effects to date of biocontrol agents released worldwide;
examine the options for the development of new
biocontrol agents; and
explore how biological control might interact with
other management techniques.
Current role of biocontrol agents
in gorse management worldwide
strong focus of research and development since 1982
has been the search for new control agents that could
augment the activity of the univoltine weevil in spring,
and also reduce the autumn seed crop that currently escapes attack. C. succedana was chosen for release in
New Zealand to fill this role (Table 1) as it has two
generations per year in its home range; one in spring
on pods of U. europaeus and another on pods of late
summer- and autumn-flowering gorse species. The
moth is now abundant throughout gorse-infested areas
of New Zealand. Gourlay et al. (2004) used insecticide exclusion to show that C. succedana augmented
control gorse seed predation by E. ulicis and recorded
an overall 81% loss in spring seed production at one
site. However, the predicted reduction in autumn seed
production has not occurred. There appears to be lack
of synchrony between the emergence of moths and
the peak occurrence of U. europaeus pods in autumn,
and infestation rates rarely exceed 10%. As seed set in
autumn forms the bulk of seed production in warmer
parts of New Zealand, adequate control of seed production has not yet been achieved.
Tetranychus lintearius Dufour
Exapion ulicis (Forster)
The gorse seed weevil, E. ulicis, is now widely established in New Zealand, Australia, the USA (the West
Coast and Hawaii) and Chile (Table 1). In New Zealand, the weevil only attacks pods in spring, whereas
gorse sets seed in both spring and autumn. Where the
bulk of annual seed production is in autumn, almost all
seeds escape attack. Where the bulk of seed production
is in the spring, infestation rates are low because of the
abundance of the food source available to the weevil
(Hill et al., 1991a). Either way, this results in approximately 65% of the annual seed crop escaping attack
(Cowley, 1983). As in New Zealand, Davies (2006)
showed that larvae feed on seeds produced in spring
and summer in Australia (Tasmania), and were not
present during a second period of seed production during autumn and winter. He found that damage to gorse
seed ranged from 12.4% to 55.4% and varied annually
within and between sites. In Chile, E. ulicis is able to
reduce gorse seed production and dispersal (Norambuena and Piper, 2000) but has had only limited impact
upon gorse invasiveness to date. The same is true in the
western United States (Markin et al., 1995) and on the
islands of Maui (1953) and Hawaii (1984) (Markin and
Yoshioka, 1998).
Cydia succedana (Denis and
Schiffermller)
With the realization that E. ulicis alone was unlikely
to reduce seed production to low levels at most sites, a
Populations of the mite T. lintearius (Tetranychidae) initially increased rapidly in the countries where it
was released (Table 1). Colonies formed massive webs
over gorse and caused severe bronzing of the foliage.
However, in New Zealand and Australia, populations
have declined at all sites after the initial increase, and
although localized outbreaks still occur, widespread
outbreaks are now rare. Mite populations in New Zealand appear to be regulated by the predators Stethorus
bifidus Kapur (Coccinellidae) and Phytoseiulus persimilis (Athias-Henriot) (Phytoseiidae).
In Australia (Tasmania), Davies et al. (2007) showed
that the presence of mite colonies on gorse bushes over
a period of 2.5 years from the time of release reduced
foliage dry weight by around 36%. However, predation
of T. lintearius colonies by Stethorus sp. and P. persimilis is widespread in Australia (Ireson et al., 2003)
and probably a key factor in restricting the impact of
the mite. P. persimilis has been associated with the destruction of entire colonies in both Tasmania and Victoria (Ireson et al., 2003) as well as in Oregon, USA
(Pratt et al., 2003).
Regulation by predators does not yet appear to be as
severe in Hawaii and Chile. The mite has been abundant in Hawaii since 2000. Chemical exclusion from
paired bushes (n = 10) indicated that mite feeding reduced gorse shoot elongation by 37% and flowering
by 82% (G. Markin, unpublished data). Similar effects
have been measured in unpublished New Zealand studies, but only where infestation is persistent. Predation
usually precludes such long-term damage. In Chile,
mite populations have grown strongly at 90% of release
sites, especially in relatively dry areas, despite predation
681
XII International Symposium on Biological Control of Weeds
Table 1.
tatus of biological control agents released worldwide against gorse (N = not established, R = recovered, not yet
S
established, E = established, HI = Hawaii, WC = west coast).
Control agent
Exapion ulicis
(Forster)
Apion sp.
Eutrichapion
scutellare Kirby)
Tetranychus
lintearius Dufour
Taxonomic
group
Status
(N, R, E)
Comments
New Zealand 1931
Hill et al., 1991a
USA, HI
1926
1958,
1984
N
E
Larvae attack seeds
in pods
First on Maui and
then Hawaii
USA, WC
1953
Australia
1939
Chile
1976
Brentidae
USA, HI
USA, HI
1958
1961
N
N
Established in TAS,
VIC, NSW and SA
Seed production and
dispersal decreased
Seed in pods
Forms galls
Markin et al., 1996;
Markin and Yoshioka, 1998
Davis, 1959; Markin
et al., 1995
Ireson et al., 2006
Acari
New Zealand 1989
Extracts mesophyll
USA, HI
USA, WC
Australia
1995
1994
1998
E
E
E
Chile
1997
2006
1995
Scythrididae
New Zealand 1990
Oecophoridae
New Zealand 1990
USA, HI
USA WC
Australia
1988
Chile
1997
2006
New Zealand 1990
USA, HI
1990
Australia
2001
Brentidae
Target
country
St. Helena
Scythris grandipennis
(Haworth)
Agonopterix
ulicetella (Stainton)
Sericothrips
staphylinus Haliday
Thripidae
Date of
release
682
Established in TAS,
VIC, NSW, SA and
WA
Released and
recovered at 50
new points between
4153 and 43S
Larvae defoliate
from solitary web
Larvae defoliate
from solitary web
Release approved;
population ex NZ
currently in quarantine F31
Recovered the
following year but
establishment not
confirmed; good
damage potential in
confinement
Released at 40 new
points between
4153 and 43S
Extracts mesophyll
Established in TAS
and VIC, establishment in NSW and
SA not confirmed
References
Norambuena et al.,
1986
Markin et al., 1995
Hill et al., 1991b,
1993
Pratt et al., 2003
Ireson et al., 2003
Norambuena et al.,
2007
H. Norambuena,
personal
communication, 2006
S.V. Fowler, personal
communication, 2006
Hill et al., 2000
Hill et al., 1995
Markin et al., 1996
Markin et al., 1995
Norambuena et al.,
2004
H. Norambuena,
personal
communication, 2006
Hill et al., 2001
Markin et al., 1996;
Hill et al., 2001
Ireson et al., 2006
A global view of the future for biological control of gorse, Ulex europaeus L.
Table 1.
(Continued) Status of biological control agents released worldwide against gorse (N = not established,
R = recovered, not yet established, E = established, HI = Hawaii, WC = west coast).
Control agent
Cydia succedana
(Denis and
Schiffermller)
Pempelia genistella
Duponchel
Taxonomic
group
Target
country
Date of
release
Status
(N, R, E)
Comments
References
Tortricidae
New Zealand 1992
Larvae feed on
seeds in pods
Hill and Gourlay,
2002
Pyralidae
New Zealand 1996
Hill et al., 2000
USA, HI
1996
Larvae defoliate
from communal web
USA, HI
2000
Uredinales
Uromyces pisi (DC.)
Otth. f. sp. europaei
Wilson and Henderson
Single pustule after
2 years, but not seen
since
Markin et al., 1996,
2002
Culliney et al., 2003
by Oligota centralis Sharp (Staphylinidae). The stress
of mite attack has slowed plant growth, flowering has
become almost totally disrupted, and occasional seedlings have been killed. The mite is now being distributed widely in Chile (Norambuena et al., 2007), but in
Hawaii, populations have declined recently, and it is
feared that regulation by a phytoseiid new to this montane region is underway.
shoots, and the effect on growth rate or biomass accumulation per plant is not known. This is also true in
New Zealand, where outbreak populations have begun
to appear in some sites since 2005, 15 years after the
moth was first released. While the moth produced substantial damage to gorse shoots enclosed within a fine
mesh sleeve in Chile (Norambuena et al., 2004), field
establishment has not been confirmed.
Sericothrips staphylinus Haliday
Pempelia genistella Duponchel
Sericothrips staphylinus (Thripidae) is now widespread in Hawaii and parts of New Zealand, and
has now become established in Australia (Table 1).
No field studies on the efficacy of this species under field conditions have been conducted. However,
a glasshouse study in Australia (Tasmania) showed
that a combination of feeding by gorse thrips, ryegrass competition and simulated grazing resulted in
a gorse seedling mortality of 93% compared with no
mortality in the untreated control, and reduced shoot
dry weight of seedlings (Davies et al., 2005). This experiment indicates the potential of S. staphylinus in
an integrated control programme if field populations
are eventually able to increase to sufficient densities.
As yet no visible damage attributable to S. staphylinus has been observed at Tasmanian field sites up to
6 years after release. The maximum estimated field
population density of juveniles and adults has been ca.
1.5 thrips cm1 of new growth. In comparison, population densities of ca. 7 thrips cm1 of new growth have
been measured in glasshouse cultures, from plants on
which severe feeding damage was recorded (J. Ireson,
unpublished data).
A small population of this pyralid moth was observed for several years following its release in Hawaii.
No larvae have been detected for some years. Recent
applications of herbicides and fire destroyed the original release sites, and the persistence of this species is
in doubt. In New Zealand, P. genistella has established
well at only a limited number of sites near Christchurch,
despite widespread releases nationwide (Table 1). The
reasons for this are not known, and its future role in
gorse management remains uncertain.
Agonopterix ulicetella (Stainton)
Despite heavy parasitism, larvae of the oecophorid
moth A. ulicetella destroy a high proportion of gorse
shoot tips in the Hawaiian infestation each spring.
However, the control agent is univoltine, and the period
of damage is short. Gorse plants appear to compensate
for the loss within the growing season by initiating new
Uromyces pisi (DC.) Otth. f. sp. europaei
Wilson and Henderson
In 2002 a single pustule of this rust was detected in
Hawaii near where it was released 2 years previously.
It was a new infection locus, and urediniospores were
being produced. No additional rust pustules have been
found since, so its continued establishment must be in
doubt.
Bioherbicides for gorse management
In addition to the classical biocontrol agents, two
pathogens are being formulated in New Zealand as
bioherbicides of gorse. It has proven difficult to formulate Fusarium tumidum Sherb. in a way that produces consistent damage to gorse foliage, but Bourdt
et al. (2006) recently showed that both F. tumidum and
Chondrostereum purpureum (Pers.:Fr.) Pouzar may
have potential as mycoherbicides for gorse regenerating after mowing or trimming.
683
XII International Symposium on Biological Control of Weeds
Options for the development
of new agents
Surveys of potential biocontrol agents for gorse have
been conducted in the native range of gorse over a
long period. Zwlfer (1963) reported on European
literature records, and CABI staff undertook surveys
of the fauna in France. Zwlfers report did not specifically state the time of year of the surveys, but they
were generally considered to have been carried out in
spring (Sheppard, 2004). R.L. Hill (1982, unpublished
results) completed a detailed study of the seasonality of
gorse insects in southern England. ODonnell (1986)
surveyed NW Spain and Portugal in spring, listing brief
site descriptions and the agent species identified. Other
less formal surveys by authors of this paper explored
the fauna of gorse as far south as Sintra, NW of Lisbon
in spring (R. Hill and G. Markin, unpublished data).
Sheppard (2004) combined the information from all of
these sources and this summary is now considered to
be the definitive list of invertebrates that have potential
as control agents for gorse.
The rate at which gorse spreads into new habitats
is likely to be strongly related to the amount of seed
produced, and reducing the annual seed crop using
seed-feeding agents may slow spread and give land
managers opportunities to protect vulnerable habitats.
E. ulicis and C. succedana singly and in combination
reduce the annual seed crop of gorse in New Zealand
(Hill et al., 2004) but not sufficiently to cause population decline (although it is possible that the long-lived
seed bank masks such an outcome, and these effects
may become apparent in the future). For New Zealand, the solution lies in finding control agents that
attack pods formed in autumn. Several Apion species
are known to take this role (Zwlfer, 1963). Cydia internana (Guerin) also fills this role, but this is a rare
species in the UK (Hill, 1982). The introduction of C.
succedana to regions outside New Zealand remains an
option, but its unpredicted appearance on hosts related
to gorse following release in New Zealand demands
caution and further research. Such research is in progress (Fowler et al., 2004).
The introduction of additional seed-feeding agents
would also be useful for Australia, especially as there
are many sites where gorse sets seed only once per
year. The planned introduction of C. succedana is now
unlikely (Ireson et al., 2006) although it may still be
considered once additional data on its true host range
and the level of damage caused to alternative hosts
is considered. Ultimately, the introduction of an additional seed feeder to Australia may depend on the
discovery of host-specific biotypes of known species
from Europe, although no such autumn seed feeders
were found in recent surveys (Sheppard, 2004). Foliage-feeding agents may also reduce annual seed production by reducing plant vigour, but as yet we know
little about the strength of this effect.
The primary aim of the field surveys conducted in
Europe during 2003 (Sheppard, 2004) was to identify
agents capable of reducing seed production in autumn.
There were only low levels of pod production during
the period of the survey, and losses of Ulex spp. seed to
pod moths (Cydia spp.) and E. ulicis were considered
minor. There was also no evidence that there were any
other autumn-specific seed-feeding agents active during this period (Sheppard, 2004).
Surveys for root-feeding agents (Sheppard and
Thomann, 2005) revealed low levels of damage over a
large part of the native range of gorse and it was concluded that this guild of insects is unlikely to contain
useful biological control agents.
It is now considered unlikely that additional invertebrate species with potential as gorse biocontrol agents
will be found in Europe, although one further survey
of insects inhabiting gorse pods in NW Spain and Portugal will be conducted in 2007. Seed produced from
flowers set in autumn contribute consistently and heavily to the annual seed crop of gorse in New Zealand
and elsewhere, and it is surprising that this resource is
not exploited by natural enemies in Europe. In contrast,
autumn seed production in Europe appears to be inconsistent, and the stochastic nature of the resource may let
these seeds escape predation.
As the most recent surveys have shown that the options for additional invertebrate biocontrol agents are
limited, surveys for host-specific fungal pathogens of
gorse commenced in the European autumn of 2006
(Ireson et al., 2006). Diseased gorse specimens were
collected in SW France, NW Spain and northern Portugal to enable isolation, culturing and identification.
Surveys for pathogens will also continue during 2007.
Discussion
Hill et al. (2004) pointed out that effective gorse management relies on selecting the most appropriate suite of
management tactics for each situation. Where gorse is
newly naturalized or of limited distribution, the highest
priorities for investment in management should be to determine the extent of the infestation, develop appropriate public policy, contain the distribution, and if possible
eradicate the weed. There is little place for biological
control here, at least in the short term. Widespread gorse
can be managed successfully to protect production or
environmental values using conventional methods such
as herbicides, although this is expensive and technically
difficult. For this reason most gorse infestations worldwide are not managed for economic or environmental
gain and yet are already too widespread for containment
to be feasible. Biological control appears to be the only
means to achieve gorse control in such habitats.
Not all available agents have yet been distributed
worldwide (Table 1), but given the lack of immediate
success where they have been released, investment in
the development of any of these agents should be made
684
A global view of the future for biological control of gorse, Ulex europaeus L.
only after critical asessment of their potential contribution to gorse management. With the conclusion of
recent surveys, our knowledge of the natural enemies
of gorse is assumed to be almost complete. Most of
the agents identified to date seem to have lower potential for impact, and appear to be less host-specific than
those already released, and there appear to be no compelling new candidate agents. In short, there appears to
be no classical biological control solution for gorse in
areas where management by conventional means cannot be brought bear as well.
Opportunities remain for augmenting and enhancing
classical biological control worldwide through integration with conventional management tactics. Spatially
explicit simulation modelling showed that seedling
survival (in particular the poor ability of gorse seedlings to compete against grasses) and disturbance were
key determinants in the population dynamics of gorse
(Rees and Hill, 2001). The model predicted that under
a limited range of scenarios of high disturbance and
high seedling mortality 7585% reduction in the annual seed production (initially set at 20,000 seeds m2)
as a result of predation by biocontrol agents could lead
to a decline of equilibrium cover in the long term.
Davies et al. (2005) showed that under laboratory
conditions the competitive ability of gorse seedlings
growing with grass can be severely reduced by insect attack although this has not been confirmed in the
field. The simulation model suggests that by reducing
the competitive ability of gorse seedlings in this way,
foliage-feeding agents may increase the probability that
seed-feeding agents will be effective in achieving longterm control of gorse, bringing the levels of reduction
in the annual seed crop required for such control within
reach of the known agents.
Gorse population simulations were also sensitive to
lifetime fecundity of gorse plants, which is directly related to the maximum age of the plants (Rees and Hill,
2001). None of the control agents released have yet
shown a propensity to cause lethal damage to mature
plants, but we know little about the chronic effects of
these control agents on maximum age in the field. It is
possible that sublethal attack may already be reducing
the vigour and longevity of gorse, affecting its long-term
population dynamics. Additional agents might enhance
that effect, even though not greatly damaging in their
own right. Management techniques such as the appropriate use of fire, grazing and overseeding may augment
this effect (Rees and Hill, 2001; Hill et al., 2004).
High control agent populations that might prove
damaging to gorse appear to be constrained from the
top-down by predation (in the case of T. lintearius), or
possibly from the bottom-up by the effect of seasonality and plant quality on the voltinism and intrinsic rate
of increase of agents (Hill, 1982; J. Ireson, unpublished
data). Even in Hawaii, where predation constraints on
T. lintearius appear to be absent, severe attack leads
to reduction in biomass and flower production, but not
plant death. In Australia, the ability of these agents to
have any significant long-term impacts on gorse growth
and development is considered to be limited without
the establishment of additional agents.
The synergy that can exist between conventional
control tactics and biocontrol agents in the management
of legume shrub weeds such as gorse are clear (Rees
and Hill, 2001; Buckley et al., 2004). Integrating weed
control techniques may offer the best prospects for longterm control in areas where gorse is actively managed,
but the extent to which biological control will play a role
in this will only be determined by future research once
the full complement of available agents are established.
References
Bourdt, G.W., Barton, J., Hurrell, G.A., Gianotti, A.F. and
Saville, D. (2006) Chondrostereum purpureum and Fusarium tumidum independently reduce regrowth in gorse
(Ulex europaeus). Biocontrol Science and Technology 16,
307327.
Buckley, Y.M., Rees, M., Paynter, Q. and Lonsdale, M.
(2004) Modelling integrated weed management of an
invasive shrub in tropical Australia. Journal of Applied
Ecology 41, 547560.
Cowley, J.M. (1983) Life cycle of Apion ulicis (Coleoptora:
Apionidae) and gorse seed attack around Auckland, New
Zealand. New Zealand Journal of Zoology 10, 8386.
Culliney, T.W., Nagamine, W.T. and Teramoto, K.K. (2003)
Introductions for biological control in Hawaii, 1997
2001. Proceedings of the Hawaiian Entomological Society 36, 145153.
Davies, J.T. (2006) The efficacy of biological control agents
of gorse, Ulex europaeus L., in Tasmania. PhD dissertation. School of Agricultural Science and Tasmanian Institute of Agricultural Research, University of Tasmania,
Australia, 180 pp.
Davies, J.T., Ireson, J.E. and Allen, G.R. (2005) The impact
of gorse thrips, ryegrass competition, and simulated grazing on gorse seedling performance in a controlled environment. Biological Control 32, 280286.
Davies, J.T., Ireson, J.E. and Allen, G.R. (2007) The impact
of the gorse spider mite, Tetranychus lintearius, on the
growth and development of gorse, Ulex europaeus. Biological Control 41, 8693.
Davis, C.J. (1959) Recent introductions for biological control
in Hawaii IV. Proceedings of the Hawaiian Entomological Society 17, 6266.
Fowler, S.V., Gourlay, A.H., Hill, R.L. and Withers, T. (2004)
Safety in New Zealand weed biocontrol: a retrospective
analysis of host-specificity testing and the predictability
of impacts on non-target plants. In: Cullen, J.M., Briese,
D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott,
J.K. (eds) Proceedings of the XI International Symposium
on Biological Control of Weeds. CSIRO Entomology,
Canberra, Australia, pp. 265270.
Gourlay, A.H., Partridge, T.R. and Hill, R.L. (2004) Interactions between the gorse seed weevil (Exapion ulicis) and
the gorse pod moth (Cydia succedana) explored by insecticide exclusion in Canterbury, New Zealand. In: Cullen,
J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin,
685
XII International Symposium on Biological Control of Weeds
L. and Scott, J.K. (eds) Proceedings of the XI International Symposium on Biological Control of Weeds. CSIRO
Entomology, Canberra, Australia, pp. 520522.
Hill, R.L. and Gourlay, A.H. (2002) Host-range testing,
introduction and establishment of Cydia succedana (Lepidoptera: Tortricidae) for biological control of gorse, Ulex
europaeus L., in New Zealand. Biological Control 25,
173186.
Hill, R.L., Gourlay, A.H. and Martin, L. (1991a) Seasonal
and geographic variation in the predation of gorse seed,
Ulex europaeus L., by the seed weevil Apion ulicis Forst.
New Zealand Journal of Zoology 18, 3743.
Hill, R.L., Grindell, J.M., Winks, C.J., Sheat, J.J. and Hayes,
L.M. (1991b) Establishment of gorse spider mite as a control agent for gorse. Proceedings of the 44th New Zealand
Weed and Pest Control Conference 44, 3134.
Hill, R.L., Gourlay, A.H. and Winks, C.J. (1993) Choosing
gorse spider mite strains to improve establishment in different climates. In: Prestidge, R.A. (ed.) Proceedings of the 6th
Australasian Conference on Grassland Invertebrate Ecology. AgResearch, Hamilton, New Zealand, pp. 377383.
Hill, R.L., ODonnell, D.J., Gourlay, A.H. and Speed, C.B.
(1995) The suitability of Agonopterix ulicetella (Lepidoptera: Oecophoridae) as a control for Ulex europaeus (Fabaceae: Genisteae) in New Zealand. Biocontrol Science
and Technology 5, 310.
Hill, R.L., Gourlay, A.H. and Fowler, S.V. (2000) The biological control programme against gorse in New Zealand.
In: Spencer, N.R. (ed.) Proceedings of the X International
Symposium on Biological Control of Weeds. Montana
State University, Bozeman, MT, USA, pp. 909917.
Hill, R.L., Markin, G.P., Gourlay, A.H., Fowler, S.V. and
Yoshioka, E. (2001) Evaluation, release, and establishment of Sericothrips staphylinus Haliday (Thysanoptera:
Thripidae) as a biological control agent for gorse, Ulex
europaeus L. (Fabaceae) in New Zealand and Hawaii.
Biological Control 21, 6374.
Hill, R.L., Buckley, Y., Dudley, N., Kriticos, D., Conant, P.,
Wilson, E., Beaudet, B. and Fox, M. (2004) Integrating
biological control and land management practices for
control of Ulex europaeus in Hawaii. In: Cullen, J.M.,
Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L.
and Scott, J.K. (eds) Proceedings of the XI International
Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 407411.
Ireson, J.E., Gourlay, A.H., Kwong, R.M., Holloway, R.J.
and Chatterton, W.S. (2003) Host specificity, release
and establishment of the gorse spider mite, Tetranychus
lintearius Dufour (Acarina: Tetranychidae), for the biological control of gorse, Ulex europaeus L. (Fabaceae), in
Australia. Biological Control 26, 117127.
Ireson, J.E., Davies, J.T., Kwong, R.M., Holloway, R.J. and
Chatterton, W.S. (2006) Biological control of gorse, Ulex
europaeus L. in Australia: where to next? In: Hanson, C.
and Stewart, K. (eds) Proceedings of the First Tasmanian
Weeds Conference. Tasmanian Weed Society Incorporated,
Devonport, Australia, pp. 1519.
Markin, G.P. and Yoshioka, E.R. (1998) Introduction and establishment of the biological control agent Apion ulicis
(Forster) (Coleoptera: Apionidae) for control of the weed
gorse (Ulex europaeus L.) in Hawaii. Proceedings of the
Hawaiian Entomological Society 33, 3542.
Markin, G.P.,Yoshioka, E.R. and Brown, R.E. (1995) Gorse, Ulex
europaeus L. In: Necholls, J.R., Andres, L.A., Beardsley,
R.D., Goeden, R.D. and Jackson, C.G. (eds) Biological
Control in the Western United States: Accomplishments
and Benefits of Regional Project W-84, 19641989. Division of Agriculture and Natural Resources, Publication
3361, University of California, Oakland, CA, USA, pp.
299302.
Markin, G.P., Yoshioka, E.R. and Conant, P. (1996) Biological control of gorse in Hawaii. In: Moran, V.C. and Hoffman, J.H. (eds) Proceedings of the IX International Symposium on Biological Control of Weed. University of Cape
Town, Rondebosch, South Africa, pp. 371375.
Markin, G.P., Conant, P., Killgore, E. and Yoshioka, E.
(2002) Biocontrol of gorse in Hawaii: a program review.
In: Smith, C., Smith W., Denslow J. and Hight, S. (eds)
Proceedings of a Workshop on Biological Control of Invasive Plants in Native Hawaiian Ecosystems. Cooperative National Park, Resource Studies Unit, University of
Hawaii, Manoa. Technical Report 129, 5361.
Norambuena, H. and Piper, G.L. (2000) Impact of Apion ulicis on Ulex europaeus seed dispersal. Biological Control
17, 267271.
Norambuena, H., Carrillo, R. and Neira, M. (1986) Introduccion, establicimento y potencial de Apion ulicis como
antagonista de Ulex europaeus en el sur de Chile. Entomophaga 31, 310.
Norambuena, H., Escobar, S. and Diaz, J. (2004) Release strategies for the moth Agonopterix ulicetella in the biological
control of Ulex europaeus in Chile. In: Cullen, J.M., Briese,
D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L. and Scott,
J.K. (eds) Proceedings of the XI International Symposium
on Biological Control of Weeds (eds). CSIRO Entomology, Canberra, Australia, pp. 440446.
Norambuena, H., Martinez, G., Carillo, R. and Neira, M.
(2007) Host specificity and establishment of Tetranychus
lintearius (Acari: Tetranychidae) for biological control of
gorse (Ulex europaeus). Biological Control 26, 4047.
ODonnell, D. (1986) A survey of the natural enemies of
gorse (Ulex spp.) in Northern Spain and Portugal. Report,
CIBC, Imperial College, Silwood Park, Ascot, UK.
Pratt, P.D., Coombs, E.M. and Croft, B.A. (2003) Predation
by phytoseiid mites on Tetranychus lintearius (Acari:
Tetranychidae), an established weed biological control
agent of gorse (Ulex europaeus). Biological Control 26,
4047.
Rees, M. and Hill, R.L. (2001) Large-scale disturbances,
biological control and the dynamics of gorse populations.
Journal of Applied Ecology 38, 364377.
Sheppard, A. (2004) A search in Spain and Portugal for potential biocontrol agents for gorse (Ulex europaeus europaeus L.) in Hawaii. CSIRO Entomology contracted
research report for Parker Ranch Inc., Hawaii, USA, 25 pp.
Sheppard, A. and Thomann, T. (2005) European survey for
potential root-feeding biological control agents of gorse
(Ulex europaeus europaeus L.). CSIRO Entomology contracted research report no. 87 for the Tasmanian Institute
of Agricultural Research, 12 pp.
Zwlfer, H. (1963) Ulex europaeus project: European investigations for New Zealand, Report No. 2. Commonwealth
Instititute of Biological Control European Station, Delmont, Switzerland, 30 pp.
686
Assigning success in biological weed
control: what do we really mean?
J.H. Hoffmann and V.C. Moran
Summary
The biological control literature is filled with the terms success and failure used in various guises
but in very few cases is the meaning clearly defined. The implicit assumption is that agents generally
have failed unless they have caused a substantial decline in the overall abundance of the target weed.
This perception is probably a legacy of the early outstanding biological control programmes against
cactus weeds, St. Johns wort and, more recently, against floating aquatic weeds. These examples
have raised expectations that almost-complete extermination of the target should be the objective
of every biological control programme. Consequently, the other types of benefits conferred by biological control are undervalued. One way of ensuring that agents are properly credited is to ask the
crucial question: What would the situation have been without any biological control? This question
provides the focus for assessing the biological control programme against Opuntia stricta (Haworth)
Haworth in South Africa, using the cactus moth, Cactoblastis cactorum Berg., and a cochineal insect,
Dactylopius opuntiae (Cockerell), which demonstrates how prolonged monitoring can reveal subtle
but very real benefits that accrue from otherwise seemingly ineffective agents.
Keywords: monitoring, Opuntia stricta, Cactoblastis, Dactylopius.
Introduction
Almost without exception, early publications on biological control provide definitions that use reduction in
either density or abundance of the target pest as the
yardstick of success. Examples include: (i) Utilisation
of [natural enemies] for the regulation of host population densities (DeBach, 1964); (ii) Effective weed
control implies reduction of the population densities
of the weed below its level of economic importance.
(Andres and Goeden, 1971); (iii) the regulation
by natural enemies of another organisms population
density at a lower level than would otherwise occur.
(DeBach, 1974); (iv) the science that deals with
the role that natural enemies play in the regulation of
the numbers ofanimal or plant pests. (Wilson and
Huffaker, 1976); (v) Biological control is the use of
[natural enemies] to suppress a pest population, making it less abundantthan it would otherwise be. (Van
Driesche and Bellows, 1996).
Zoology Department, University of Cape Town, Rondebosch 7700,
South Africa.
Corresponding author: J.H. Hoffmann <[email protected],
[email protected]>.
CAB International 2008
The basis of these definitions is a few early projects that resulted in spectacular declines in the density
and abundance of the target plants [e.g. the use of
Cactoblastis cactorum (Berg) against Opuntia stricta
(Haworth) in Australia; Chrysolina quadrigemina (Suffrian) against St Johns wort, Hypericum perforatum
L., in California] and various projects against aquatic
weed species. These overwhelming successes established a reputation that biological control is a process
that can replace all other control methods. In most upto-date publications, lectures and discussions, success
continues to be equated with substantial reductions in
densities of the target weed. A corollary is that projects
are perceived as failures when weed densities do not
decline, or do so only marginally. Unfortunately, many
biological control programmes against weeds fall into
this latter category.
More recently, pleas have been made to rationalize what is meant by success (e.g. Hoffmann, 1995;
McFadyen, 1998; Briese, 2000; Fowler et al., 2000)
and to develop more precise performance criteria for
the role of biological control in weed management
(van Klinken and Raghu, 2006). While this needs to
be done if biological control is to receive the recogni
tionit deserves, it seems that old habits die hard and
that, in the main, biological control practitioners still
687
XII International Symposium on Biological Control of Weeds
hanker after agents that will all but eliminate their target weeds.
Ironically, one of the most successful biological
control agents of all time, C. cactorum, provides a useful example of how seemingly ineffective agents, in
terms of conventional expectations, can provide very
real benefits. In this case, the realization was a posteriori, but it nevertheless clearly demonstrates how success in biological control need not be an all or nothing
process.
O. stricta is a widespread weed in South Africa, with
one of the most badly affected areas being the Kruger
National Park (KNP), the flagship region of South African conservation, where the weed became naturalized
during the 1950s. By the 1980s, the problem was out
of control in spite of a protracted and expensive herbicide programme that had been in place for several
years. In desperation and as a last resort, moves were
made to initiate a biological control programme against
O. stricta in KNP (Hoffmann et al., 1998a). This biological control project was potentially straightforward
because O. stricta had been controlled so superbly by
C. cactorum in Australia (Dodd, 1940). The moth was
already well established on O. ficus-indica (L.) Mill.
elsewhere in South Africa and was immediately available for introduction into KNP.
The release of C. cactorum in KNP during 1987 was
greeted with much enthusiasm, fuelled by visible and
impressive evidence of larval damage that became apparent over an ever-increasing area of the cactus invasion. The initial euphoria gradually waned as the extent
of damage equilibrated at lower levels than were optimistically expected on the basis of the Australian precedents and the cactus remained abundant over a wide
area. By 1993, most observers were disillusioned and
some were openly critical of the biological control programme, considering it to have aggravated rather than
contained, let alone alleviated, the problem. This scepticism and hostility to biological control was the catalyst for the initiation of a research project to quantify
the impact of the moth and thereby determine whether
or not there have been any benefits from its presence
within KNP. The findings of the evaluation programme
are presented here.
Methods
Details of the materials and methods used to accumulate the data presented in the results are given in Hoffmann et al. (1998a,b). In essence, the numbers of C.
cactorum colonies and of plants and fruits of O. stricta
were monitored over a 9-year period (19932001).
Measurements of the relative numbers of C. cactorum
colonies, and of plant size and density were made in
two different types of infestations: (1) an area were the
cactus had been sprayed with herbicides a year before
the initiation of the study (designated the treated area);
and (2) an adjacent area where no herbicides had been
used (designated the untreated area). The residual
population of O. stricta in the treated area was sparse
and consisted of small plants that had been overlooked
during the herbicide treatment. In the untreated area,
the plants were large (up to 2.5 m in height) and dense,
forming clusters with 100% ground cover over several
square meters.
Counts were made along permanent transects which
were 100 m in length and 1 m wide. There were five
transects in the untreated infestation and ten transects
in the treated infestation. The size of the plants was recorded by counting the number of cladodes that made
up each plant. To estimate the mean growth rate of
undamaged plants, the sizes of individual plants (n =
67) that had not been colonized by C. cactorum was
measured at the beginning and end of an annual growth
cycle.
Results and discussion
In temperate regions, C. cactorum has two generations
a year (a short summer one and a long winter one) but
in the warm tropical climate of the KNP, the moth passes through two generations in summer and one in winter. Population levels of C. cactorum fluctuated both
annually and seasonally (Figure 1A). In the treated
infestations, the moths abundance increased between
1993 and 1995 as the populations recovered from being reduced in numbers by the herbicides (Figure
1B), but there was no trend toward an overall increase
or decrease in abundance of C. cactorum during the
study period, 19932001.
Despite the presence of C. cactorum, there was a
steady increase in the density of O. stricta plants over
the study period in both the untreated and treated infestations, increasing by fourfold and sixfold, respectively (Figure 2). During the same period, there was a
substantial increase in the area of the research plots in
which O. stricta occurred. In the untreated area, quadrat occupancy rose from 17.5% in 1993 to 45.4% in
2001, whereas in the treated area, it rose from 7.9% to
13.7% between 1993 and 2001, showing that the weed
was dispersing in spite of damage caused by C. cactorum. The biomass of the cactus (measured as cladodes
m2) increased in both areas but at a much higher rate
in the treated area (Figure 3). Finally, fruit production
in the infestations did not diminish as a result of the
damage caused by C. cactorum (Figure 4). Fruits have
been a major concern because one of the immediate
objectives in the management of the weed has been to
curtail its long-range dispersal by preventing seeding
and consequent spread via fruit-eating animals, mainly
elephants and baboons.
When considered in the light of conventional definitions of success in biological control (i.e. reduction in
density or abundance of the target weed), C. cactorum
on O. stricta in KNP must rank as an outright failure.
So how can the situation be perceived differently? One
688
Assigning success in biological weed control: what do we really mean?
0.25
colonies m-2
0.2
0.15
0.1
0.05
0
1993
0.05
1994
1995
1996
1997
1998
1999
2000
2001
1994
1995
1996
1997
1998
1999
2000
2001
colonies m-2
0.04
0.03
0.02
0.01
0
1993
Figure 1.
Abundance of Cactoblastis cactorum (mean 1 SE colonies
m2) in an untreated infestation (A) and a herbicide-treated
infestation (B) of Opuntia stricta in KNP between 1993 and
2001 (solid bars = winter; open bars = summer).
A
Plants m -2
0
1993
1994
1995
1996
1997
1998
1999
2000
2001
1997
1998
1999
2000
2001
plants m-2
0.8
0.6
0.4
0.2
0
1993
Figure 2.
1994
1995
1996
The density of Opuntia stricta plants (mean 1 SE plants m2)
in an untreated infestation (A) and a herbicide-treated infestation (B) in KNP between 1993 and 2001.
689
XII International Symposium on Biological Control of Weeds
50
Cladodes m-2
40
30
20
10
0
1993
cladodes m-2
1994
1995
1996
1997
1998
1999
2000
2001
1994
1995
1996
1997
1998
1999
2000
2001
0
1993
Figure 3.
The biomass (mean 1 SE cladodes m2) of Opuntia stricta in an untreated infestation (A) and a treated infestation (B) in KNP between
1993 and 2001.
25
Fruits m-2
20
15
10
5
0
1993
Fruits m -2
0.3
1994
1995
1996
1997
1998
1999
2000
2001
1994
1995
1996
1997
1998
1999
2000
2001
0.2
0.1
0
1993
Figure 4.
Fruit production (mean 1 SE numbers m2) on Opuntia
stricta in an untreated infestation (A) and a herbicide-treated
infestation (B) in KNP between 1993 and 2001.
690
Assigning success in biological weed control: what do we really mean?
scribed by Hoffmann et al. (1998b) and Lotter and
Hoffmann (1998). Essentially the process entailed less
frequent applications of herbicides and treatment of
only the largest fruit-bearing plants while leaving small
plants, which are difficult to locate, for C. cactorum.
This procedure resulted in a substantial financial saving
and enabled coverage of a greater area of the infested
parts of the park with the same allocation of resources.
This study has shown that qualitative assessments
that regarded C. cactorum as something of a failure
were not justified. The key element in changing this
misconception was to ask What would the situation
have been without biological control? rather than
What has been achieved? This approach forces the observer to consider scenarios of where the weed would
be without biological control and therefore to look at
the system from a different and more-telling perspective. In the case of C. cactorum, it has been possible
to show, through careful evaluation over almost a decade that the moth was having a dramatic effect, even
though the weed remained abundant in the presence of
this agent acting alone.
More recently, the situation has changed dramatically
with the introduction of a cochineal insect, Dactylopius
way is to ask What would the situation have been without any biological control? and to use a simple spreadsheet model to estimate how much cactus there would
have been if C. cactorum was not in the system.
The counts of cladodes on healthy, undamaged
O. stricta plants in KNP showed that there is an annual
increment of 1.8. This value can be used to calculate
the expected increase in biomass of the cactus over a
given time assuming absence of C. cactorum. Comparisons of expected and observed biomass over an 8-year
period show that there is an escalating difference with
time (Figure 5), which is attributable to damage caused
by C. cactorum. A substantial divergence accrues even
when annual growth of the cactus is reduced to a factor of 1.6 each year (i.e. 90% of capacity). The comparisons show that the infestations of the weed were
much less prolific than they would have been without
biological control and, even though the problem was
getting worse, the rate at which this was happening had
been substantially curtailed by C. cactorum.
The reduction in growth rates due to C. cactorum
took on special significance when biological control
was integrated with herbicide treatment of the weed.
The details of such an integrated programme are de700
600
Cladodes m-2
500
400
300
200
100
0
1
Year
35
30
Cladodes m-2
25
20
15
10
5
0
1
Year
Figure 5.
Observed (closed circles) and expected (closed squares =
growth at 100% of capacity; open squares = growth at 75%
of capacity) densities of cladodes (number m2) in an untreated infestation (A) and a herbicide-treated infestation (B) of
Opuntia stricta in KNP over an 8-year period.
691
XII International Symposium on Biological Control of Weeds
opuntiae (Cockerell), into KNP. This step has resulted
in a massive decline in the abundance of O. stricta in
KNP (unpublished results) and the weed is now considered by everyone to be under excellent biological control. The situation continues to be monitored to confirm
that this level of control is sustainable. The integrated
control programme has been replaced by a fully fledged
biological control initiative with cochineal being manually dispersed at every opportunity and herbicides no
longer used. C. cactorum persists in the system and the
interactions of the two agents are being monitored.
In spite of this recent drastic change in the status of
O. stricta in KNP, the message remains that we need to
get away from the perception that a reduction in density
is the primary, if not the sole, requirement for success
in biological control. In doing so, the status of many of
our supposedly unsuccessful agents is going to change
and biological control is going to be perceived much
more favourably, both by ourselves, by the public at
large and, perhaps most importantly, by the people
holding the purse strings.
Acknowledgements
We are grateful to many people who have been involved
in our KNP programme over the years; especially Dave
Zeller, Wayne Lotter and Llewellyn Foxcroft of the
National Parks Board, and Helmuth Zimmermann and
Fiona Impson. The National Research Foundation,
University of Cape Town, Agricultural Research Institute and National Parks Board have provided funding
and logistical support for this long-term project. Carien
Kleinjan provided valuable advice on presentation of
the manuscript and talk.
References
Andres, L.A. and Goeden, R.D. (1971) The biological control
of weeds by introduced natural enemies. In: Huffaker, C.B.
(ed.) Biological Control. Plenum, New York, pp. 143164.
Briese, D.T. (2000) Classical biological control. In: Sindel,
B.M. (ed.) Australian Weed Management Systems. R.G. &
F.J. Richardson, Melbourne, Australia, pp. 161192.
DeBach, P. (1964) The scope of biological control. In: DeBach, P. (ed.) Biological Control of Insect Pests and
Weeds. Reinhold, New York, pp. 320.
DeBach, P. (1974) Biological Control by Natural Enemies.
Cambridge University Press, Cambridge, UK, 323 pp.
Dodd, A.P. (1940) The Biological Campaign Against Prickly
Pear. Commonwealth Prickly Pear Board Bulletin, Brisbane, Australia, 177 pp.
Fowler, S.V., Syrett, P. and Hill, R.L. (2000) Success and
safety in the biological control of environmental weeds in
New Zealand. Austral Ecology 25, 553562.
Hoffmann, J.H. (1995) Biological control of weeds: the
way forward, a South African perspective. Weeds in
a Changing World. British Crop Protection Council,
Symposium Proceedings No. 64. BCPC, Brighton, UK,
pp. 7789.
Hoffmann, J.H., Moran, V.C. and Zeller, D.A. (1998a) Evaluation of Cactoblastis cactorum (Lepidoptera: Phycitidae)
as a biological control agent of Opuntia stricta (Cactaceae)
in the Kruger National Park, South Africa. Biological
Control 12, 2024.
Hoffmann, J.H., Moran, V.C. and Zeller, D.A. (1998b) Longterm population studies and the development of an integrated management programme for control of Opuntia
stricta in Kruger National Park, South Africa. Journal of
Applied Ecology 35, 156160.
Lotter, W.D. and Hoffmann, J.H. (1998) An integrated management plan for the control of Opuntia stricta (Cactaceae) in the Kruger National Park, South Africa. Koedoe
41, 6368.
McFadyen, R.E.C. (1998) Biological control of weeds. Annual Review of Entomology 43, 369393.
Van Driesche, R.G. and Bellows, T.S. (1996) Biological Control. Chapman & Hall, New York, 539 pp.
van Klinken, R.D. and Raghu, S. (2006) A scientific approach
to agent selection. Australian Journal of Entomology 45,
253258.
Wilson, F. and Huffaker, C.B. (1976) The philosophy, scope,
and importance of biological control. In: Huffaker, C.B.
and Messenger, P.S. (eds) Theory and Practice of Biological Control. Academic Press, New York, pp. 315.
692
Combination of a mycoherbicide with
selected chemical herbicides for control
of Euphorbia heterophylla
K.L. Nechet,1 B.S. Vieira,1 R.W. Barreto,1 E.S.G. Mizubuti1 and A.A. Silva2
Summary
The leaf-spot fungus Lewia chlamidosporiformans is being developed as a mycoherbicide for wild
poinsettia (Euphorbia heterophylla), a serious weed of many crops, and particularly of soybeans, in
Brazil. A comparative study of the levels of control of wild poinsettia obtained for the fungus alone,
the fungus plus selected herbicides and these herbicides alone was undertaken. The levels of control
obtained with application of the fungus alone was equivalent to the chemical herbicides fomesafen,
carfentrazone and atrazine for plants with up to five leaves, and equivalent to the herbicide glyphosate
for plants with five to ten leaves for one of the weed populations being tested. The fungus alone had
a better performance than both imazethaphyr and fomesafen. The combination of fomesafen with
the fungus produced complete control for the three weed populations, whereas its combination with
imazethaphyr significantly improved the control levels as compared with those obtained with imazethaphyr alone. Another experiment performed in microplots in the field, explored the combination
of L. chlamidosporiformans with fomesafen and clorimuron-ethyl for the control of an imazethaphyrresistant biotype of the weed. The best control levels were achieved with the application of an equivalent of 300 l/ha of a suspension of conidia (2.5 105 conidia/ml) in a solution of fomesafen (25% of
recommended dose). Death of all plants resulted from such application after 10 days.
Keywords: Lewia, fungus, weed management, wild poinsettia, bioherbicide.
Introduction
Euphorbia heterophylla L. (wild poinsettia, local name
in Brazil amendoim-bravo or leiteiro) is a plant native
to the Neotropics that became an aggressive invader of
important crops such as corn, sugarcane, common bean
and soybean in Brazil. It was recognized as a potential
biocontrol target with the pioneering development of
a mycoherbicide aimed at controlling wild poinsettia
in Brazil (Yorinori, 1985, 1987; Yorinori and Gazziero,
1989). This was based on Bipolaris euphorbiae (Hansford) Muchovej. Although interest in this fungus
still remains (Marchiori et al., 2001; Nechet et al.,
Universidade Federal de Viosa, Departamento de Fitopatologia, CEP
36571-000, Viosa, MG, Brazil.
2
Universidade Federal de Viosa, Departamento de Fitotecnia, CEP
36571-000, Viosa, MG, Brazil.
Corresponding author: R.W. Barreto <[email protected]>.
CAB International 2008
1
2006), this research never resulted in a commercially
viable product. Chemical control through application
of acetolactate synthase (ALS)-inhibiting herbicide
remains the method of choice for control of E. heterophylla in soybeans. Herbicides with this action mechanism are efficient in low doses, have a low toxicity for
mammals and are selective for several important crops
(Saari et al., 1994). Unfortunately, the continuous use
of this group of herbicides led to the widespread emergence of wild poinsettia populations that are resistant
to these products, rendering them ineffective in many
situations (Gazziero et al., 1997; Melhorana and
Pereira, 1999).
Problems with herbicide resistance represent windows of opportunity for the development and use of
mycoherbicides, either separately or in mixture with
chemical herbicides (Charudattan, 2001). Surveys of
the mycobiota of E. heterophylla in Brazil yielded additional fungi that might be of interest for use as mycoherbicides (Barreto and Evans, 1998). An isolate
693
XII International Symposium on Biological Control of Weeds
of Lewia chlamidosporiformans Vieira and Barreto,
a fungus recently described (Vieira and Barreto, 2005),
was selected as a potential mycoherbicide. It was
capable of causing high mortality in nine populations
of E. heterophylla, including one resistant to ALSinhibiting herbicides.
Combining a biocontrol agent with a chemical
herbicide may result in the reduction of the dose of
the chemical herbicide needed to control a specific
weed. It may also serve to increase the spectrum
of weeds controlled by the chemical product or increase its efficiency of control of a problem weed
species (Hoagland, 1996). It was hypothesized that
in the case of E. heterophylla, L. chlamidosporiformans might have such an effect when combined with
ALS-inhibiting herbicides, as well as other groups
of herbicides.
Materials and methods
Isolate of L. chlamidosporiformans and
biotypes of E. heterophylla
An isolate of L. chlamidosporiformans (KLN06)
was previously selected, among six other isolates,
because it was capable of causing high disease severity when tested against nine different populations of
E. heterophylla (including one known to be resistant to
ALS-inhibiting herbicides and one resistant to B. euphorbiae). KLN06 was the isolate used in the experiments described here.
Three different populations of E. heterophylla were
selected to be used in the first experiment: EKLN19,
previously selected as susceptible to L. chlamidosporiformans, B. euphorbiae (Hansford) Muchovej and
Sphaceloma poinsettiae Jenkins and Ruehle; ETRB
resistant to B. euphorbiae, collected in experimental fields of Paran; and ERH resistant to ALS-inhibiting herbicides. In the second experiment, only
EKLN19 was used. Seeds to be used in the experiments were harvested from greenhouse-grown potted plants and kept at 5C until use. Plants used in
the experiments were produced from pregerminated
seeds that were then transferred to 500 ml pots containing sterile soil and the plants were maintained in
a greenhouse (26 2C) and watered daily in the first
experiment.
Inoculum production
Spores of L. chlamidosporiformans to be used as
inoculum in the experiments were produced with a
biphasic technique modified from Walker (1980) as
follows: ten culture disks obtained from the margin of
7-day-old cultures grown in VBA (Pereira et al., 2003)
were transferred to each of a series of Erlenmeyer flasks
containing 100 ml of VB (the same medium described
by Pereira et al., 2003, but without agar). The Erlen-
meyers were left on a shaker at 140 rpm for 7 days
at room temperature. After this period, the mycelial
mass was blended together with the remaining liquid medium in each flask and poured and spread onto
20 28 cm aluminium trays, each already containing
100 ml of solidified VBA. Trays were kept in a controlled temperature room at 26 2C under a 12-h
photoperiod (light from two 40-W daylight fluorescent
lamps and two 40-W fluorescent, near ultraviolet light
lamps). After 2 days, spores were collected by pouring
50 ml of sterile water on the culture surface and scraping it with a rubber spatula. The resulting suspension
was then filtered through two layers of cheesecloth and
the final concentration of the suspension was evaluated
and adjusted for the experiments.
Wild poinsettia control with
mycoherbicide, selected herbicides
and their combination greenhouse
Only herbicides that did not inhibit spore germination in a previous experiment were used in this experiment. These were atrazine, carfentrazone, fomesafen,
glyphosate and imazethaphyr. Plants belonging to the
three populations above were sprayed with (1) the
herbicides alone; (2) the fungus alone at a concentration of 2 105 conidia/ml supplemented with Assist
3% (mineral oil) and Break thru 320 l/100 ml
(an organosilicone surfactant); and (3) with mixtures
of the fungus inoculum on the concentration mentioned above with each of the herbicides diluted in
water supplemented with Assist 3% and Break thru
320 l/100 ml. Plants were sprayed at the end of the
afternoon, left outside to be exposed to natural dew formation and moved next morning to a greenhouse (26
2C). The experiment was carried out in a completely
randomized design with a factorial of 12 treatments,
three plant populations and five replications per treatment. Each replication consisted of one pot containing
one plant. Plants were treated at an age of 4 weeks (four
leaf stage).
The evaluations were made 5 and 10 days after the
application of the treatments using a scale from 0 to
10 (Table 1). The nonparametric KruskalWallis (KW)
test was done on the rating values for each isolate and
time after inoculation combination. When KW tests
were significant, multiple comparisons of the means
were conducted to assess differences among treatments
at = 0.05. All nonparametric analyses were conducted
using the R package ver 2.3.1 (R Development Core
Team, 2006).
Wild poinsettia control with
mycoherbicide, selected herbicides
and their combination field
The experimental unit consisted of an area of 1.5 m2
(1.5 1.0 m), where three parallel 0.5-cm-deep rows set
694
Combination of a mycoherbicide with selected chemical herbicides for control of Euphorbia heterophylla
Table 1.
Scale of notes for evaluation of the control of populations of Euphorbia heterophylla.
Rank
0
1
2
3
4
5
6
7
8
9
10
Description of ranking
0.5 m apart were prepared for planting. An amount of
fertilizer equivalent to 20 kg/ha of N, 100 kg/ha of P2O5
and 100 kg/ha of K2O was added and a row equivalent
to 2530 soybean seeds/m was sown. Additional parallel rows, 25 cm to each side of the soybean rows, were
prepared where ETRB E. heterophylla seeds (equivalent to 30 seeds/m) were planted. Treatments consisted
of spraying with the following:
1. suspension of 2.5 105 conidia/ml of L. chlamidosporiformans;
2. herbicide clorimuron-ethyl (Classic 80 g/ha);
3. herbicide clorimuron-ethyl (Classic 80 g/ha) +
2.5 105 conidia/ml of L. chlamidosporiformans
suspended in the herbicide solution;
4. herbicide fomesafen (Flex, 1.0 l/ha recommended
dose);
5. herbicide fomesafen (Flex, of recommended
dose);
6. herbicide fomesafen (Flex, of recommended
dose) + 2.5 105 conidia/ml of L. chlamidosporiformans suspended in the herbicide solution;
Table 2.
Control percentage
No symptom
Yellowing
Terminal buds not damaged, less than 20% of the leaves showing injuries
Terminal buds not damaged, 2050% of the leaves showing injuries
Terminal buds not damaged, more than 50% of the leaves showing injuries
Terminal buds not damaged, all leaves with injuries
Terminal buds necrosed, leaves healthy
Terminal buds necrosed, less than two leaves with injury
Terminal buds necrosed, more than two leaves injured
Terminal buds necrosed, stems green, complete defoliation
Dead plants
0
10
20
30
40
50
60
70
80
90
100
7. herbicide fomesafen (Flex, of recommended dose)
+ 2.5 106 conidia/ml of L. chlamidosporiformans
suspended in the herbicide solution;
8. control 1: soybean challenged with wild poinsettia
sprayed with water;
9. control 2: soybean growing free of wild poinsettia
competition sprayed with water.
The volume of solution/suspension sprayed on each
treatment was equivalent to an application of 300 l/ha.
Water used to suspend or dilute the active ingredient or
inoculum was always supplemented with Assist 3%
(mineral oil) and Break thru 320 l/100 ml. Plants
were inoculated at the age of 2 weeks (three to four
leaf stage) at the end of the afternoon. Evaluations were
made 5 and 10 days after the application of the treatments using the criteria adopted by SBCPD (legend,
Table 2). The experiment was carried out in a completely randomized design with five repetitions; each
repetition was as described above. Data obtained were
graphed with the aid of Microsoft Excel 2000 (Microsoft Corporation, Redmond, WA).
Percentage of control of Euphorbia heterophylla in the field after spraying with Lewia chlamidosporiformans,
selected herbicides and fungusherbicide combinations.
Treatment
Control percentage
Five days
Clorimuron-ethyl
Clorimuron-ethyl + fungus (105 conidia/ml)
Fomesafen standard dose
Fomesafen dose
Fomesafen dose + fungus (105 conidia/ml)
Fomesafen dose + fungus (106 conidia/ml)
Fungus (105 conidia/ml)
Control
Ten days
Percentage
control
Control
description
Percentage
control
Control
description
80
98
100
94
98
92
97
0
B
A
A
A
A
A
A
E
70
99
100
86
100
98
99
0
C
A
A
B
A
A
A
E
Capital letters represent the control description according to SBCPD (1995). A = Excellent control or total control of the target species
(86100%); B = Good control of the target-species, acceptable for an infested area (6685%); C = Moderate control of the target-species,
insufficient for the infested area (4165%); D = Control of the target-species deficient or irrelevant (040%); E = No control (0%).
695
XII International Symposium on Biological Control of Weeds
Results
Wild poinsettia control with
mycoherbicide, selected herbicides and
their combination greenhouse
There were differences among treatments at both 5
and 10 days. The P values were less than 0.0001 for all
tests and the chi-square values for populations TRB,
KLN19 and RH at 5 days were 58.9, 59.0 and 58.8, respectively. At 10 days, the chi-square values were 58.8,
55.8 and 56.3, respectively.
Five days: When assessing plants of the TRB population, the control differed from the treatments: Atrazine +
L. chlamidosporiformans, Carf + L. chlamidosporiformans, Flex + L. chlamidosporiformans and Pivot +
L. chlamidosporiformans. All other pairwise comparisons involving the control were not significant. The
only mixture of herbicide plus L. chlamidosporiformans
that did not differ from the control was Glyphosate +
L. chlamidosporiformans. The pairwise comparisons
involving L. chlamidosporiformans vs Atrazine + L.
chlamidosporiformans and Atrazine vs L. chlamidosporiformans + Imazethaphyr were significant.
When assessing plants of the KLN19 population,
the control differed from the treatments: L. chlamidosporiformans; Atrazine; Atrazine + L. chlamidosporiformans; Carfentrazone; Carfentrazone + L. chlamidosporiformans; Fomesafen + L. chlamidosporiformans;
and Glyphosate + L. chlamidosporiformans. All other
pairwise comparisons involving the control were not
significant. Fomesafen, Glyphosate, Imazethaphyr and
Imazethaphyr + L. chlamidosporiformans treatments
did not differ from the control. The pairwise comparisons involving Atrazine vs Atrazine + L. chlamidosporiformans and Atrazine + L. chlamidosporiformans vs
Imazethaphyr were significantly different.
When assessing plants of the RH population, the
control differed from the treatments: Atrazine + L.
chlamidosporiformans; Carfentrazone + L. chlamidosporiformans; Fomesafen + L. chlamidosporiformans;
and Imazethaphyr + L. chlamidosporiformans. All other
pairwise comparisons involving the control were not
significant. Similarly to the reported for KNL19, pairwise comparisons involving Atrazine vs Atrazine + L.
chlamidosporiformans and Atrazine + L. chlamidosporiformans vs Imazethaphyr were significantly different.
Ten days: For both the TRB and RH population, the control differed from the following treatments: Atrazine +
L. chlamidosporiformans; Carfentrazone + L. chlamidosporiformans; Fomesafen + L. chlamidosporiformans; and Imazethaphyr + L. chlamidosporiformans.
All other pairwise comparisons involving the control
were not significant. Pairwise comparisons between
Atrazine vs Atrazine + L. chlamidosporiformans; Atrazine + L. chlamidosporiformans vs Imazethaphyr; and
Atrazine vs Imazethaphyr were significantly different.
For the KLN19 population, there were no pairwise differences between treatments.
All plants of population EKLN19 died 5 days after
being sprayed with L. chlamidosporiformans as well as
with the herbicides atrazine, carfentrazone and with the
mixtures of these herbicides with the fungus. Equivalent levels of control were observed for the populations
ERH and ETRB submitted to the same herbicides and
with their mixture with the fungus. The level of control obtained with the pure application of L. chlamidosporiformans in those populations was lower then
that obtained for EKLN19. Control was inadequate
for the application of fomesafen and imazethaphyr,
confirming their known inefficiency for the control of
E. heterophylla plants at the stage of development of
application in this experiment (three to four leaf stage).
Nevertheless, complete control of wild poinsettia was
obtained when fomesafen was mixed with L. chlamidosporiformans (Figure 1). The combination of imazethaphyr with the fungus led to significant increase in
control levels for the three weed populations, whereas
they were totally inadequate for imazethaphyr alone. Ten
days after the inoculation, the level of control, obtained
for the combination of L. chlamidosporiformans with
imazethaphyr, was considered acceptable for EKLN19
and for ERH although still insufficient for ETRB.
Glyphosate sprays led to total control of the three
wild poinsettia populations, after 10 days when applied
alone, but control was inadequate for EKLN19 and
ERH initially, that is 5 days after the application. However, when mixed with the fungus, glyphosate gave
complete control of all weed populations in just 5 days,
speeding weed control.
Wild poinsettia control with
mycoherbicide, selected herbicides
and their combination field
All E. heterophylla plants treated with the fungus +
fomesafen ( of the dose) were killed 10 days after
being sprayed (Table 2). In the first evaluation, 5 days
after the application, levels of control obtained for the
mixtures fungus + fomesafen ( of the dose) and fomesafen alone ( of the dose) did not differ statistically.
Nevertheless, on the second evaluation, 10 days after
the applications, total control was reached only with the
mixtures fungusherbicide. Level of control obtained
with fomesafen alone ( of the dose) and clorimuronethyl alone decreased from the first evaluation to the
second evaluation indicating a recovery of the weed
population from the damage caused by the herbicides.
Meanwhile, a tendency of increased control was noted
for treatments involving the use of L. chlamidosporiformans probably mirroring the progressive advance
of the disease. The level of control obtained with the
isolated application of L. chlamidosporiformans was
also ranked as excellent and statistically equivalent
696
Combination of a mycoherbicide with selected chemical herbicides for control of Euphorbia heterophylla
Figure 1.
Control of population EKLN19 of Euphorbia heterophylla, 5 days after spraying
with water, fomesafen, a conidial suspension of Lewia chlamidosporiformans and
a mixture of fomesafen + L. chlamidosporiformans.
to that obtained with the application of the herbicide
fomesafen at the recommended dose and superior to
that obtained with the application of of the dose of
this herbicide (Table 2). The combination of the herbicide clorimuron-ethyl with the fungus resulted in a
percentage of control higher than that obtained for the
herbicide alone.
Discussion
The level of control obtained with application of
L. chlamidosporiformans varied for each weed population, but when in combination with various herbicides
included in the test, the level of control was either kept
as complete control or increased as compared with the
herbicide alone. The efficiency of control of E. heterophylla by fomesafen is seriously restricted if applications occur at a later phenological state of the plant, as
observed in greenhouse experiment (where older plants
were used) and Table 2 (where younger plants were
used), but total control was achieved with a mixture of
fomesafen and L. chlamidosporiformans of older plants
(Figure 1). The fungus in mixture with glyphosate also
allowed the anticipation of the wild poinsettias control
by this product. These results are highly encouraging
for an integration of biological and chemical control
of wild poinsettia. This work seemingly is the first to
focus on that aspect in the development of a mycoherbicide in Brazil.
The herbicide imazethaphyr did not control the
population ERH (known to be resistant to this product) and the population ETRB. The population ETRB
was selected because it was known to be resistant to
B. euphorbiae; the present results show that it is also
resistant to imazethaphyr. The level of control of wild
poinsettia was significantly increased when L. chlamidosporiformans was mixed with that herbicide, but
control percentage was for most situations lower than
that obtained by the fungus alone. The incompatibility
between mycoherbicide and chemical herbicides is still
a little investigated subject, but it is known that herbicides can interfere with disease development, either because of a direct toxicity to the pathogen or indirectly
by triggering defense responses in the plants (Sanogo
et al., 2000). The negative effect of imazethaphyr observed on L. chlamidosporiformans was indirect because this herbicide did not inhibit conidial germination in vitro (unpublished results) in the recommended
dose. Possibly, some plant defense mechanism of the
plant is activated after the application of imazethaphyr
and this slowed disease development. Conversely, it
was recently observed that imazethaphyr applications
can increase significantly the severity of attack to soybean by Rhizoctonia solani Khun both in resistant and
susceptible cultivars (Bradley et al., 2002).
Lewia chlamidosporiformans can become an important tool in the management of populations of E.
heterophylla resistant to (ALS)-inhibiting herbicides in
separate applications as a mycoherbicide, but further
investigation of its interaction with imazethaphyr as a
possible combination is needed. Other options deserving to be investigated are the use of smaller doses of
imazethaphyr with the fungus and the effect of separate
sequential application of fungus and herbicide or herbicide followed by the fungus.
697
XII International Symposium on Biological Control of Weeds
The second experiment also yielded encouraging
results both for the potential of the mycoherbicide
alone and for fungusherbicide combinations. Often
the best alternative for weed management depends on
the association of different methods of control (Silva
and Alto, 1993). In this experiment, one of the most
noteworthy results was achieved with the combination
of L. chlamidosporiformans and fomesafen at of the
recommended dose. Complete control of wild poinsettia was achieved even with a significant reduction of
the use of the chemical herbicide and a relatively small
inoculum concentration for the fungus. Except for the
isolated application of clorimuron-ethyl that did not
result in an appropriate control of the weed, the other
treatments resulted in a good control of the weed. Adding a relatively low concentration of fungus inoculum
to clorimuron-ethyl also led to complete control of wild
poinsettia.
Although a series of studies involving L. chlamidosporiformans have been yielding excellent results
that are presently being published (see Vieira et al.,
this volume) and firmly indicate that this fungus may
be used soon as a mycoherbicide, particularly against
herbicide-resistant populations of E. heterophylla, further work is still necessary. Improvement in mass production, formulation and application technology are still
required as well as further studies on fungusherbicide
combinations.
Acknowledgements
Seeds used in the experiments were provided by the
Laboratrio de Herbicida na Planta Departamento de
Fitotecnia, Universidade Federal de Viosa or by J.T.
Yorinori (Embrapa Soja). The authors acknowledge
CNPq and CAPES for financial support.
References
Barreto, R.W. and Evans, H.C. (1998) Fungal pathogens of
Euphorbia heterophylla and E. hirta in Brazil and their
potential as weed biocontrol agents. Mycopathologia 141,
2136.
Bradley, C.A., Hartman, G.L., Wax, L.M. and Pedersen, W.L.
(2002) Influence of herbicides on Rhizoctonia root and
hypocotyl rot of soybean. Crop Protection 21, 679687.
Charudattan, R. (2001) Biological control of weeds by means
of plant pathogens: Significance for integrated weed
management in modern agro-ecology. BioControl 46,
229260.
Gazziero, D.L.P., Brighenti, A.M., Maciel, C.D.G., Christofolleti, P.J., Adegas, F.S. and Voll, E. (1997) Resistncia
de amendoim-bravo aos herbicidas inibidores da enzima
ALS. Planta Daninha 16, 117125.
Hoagland, M.R.E. (1996) Chemical interactions with bioherbicides to improve efficacy. Weed Technology 10, 651674.
Marchiori, R., Nachtigal, G.F., Coelho, L., Yorinori, J.T. and
Pitelli, R.A. (2001) Comparison of culture media for the
mass production of Bipolaris euphorbiae and its impact
on Euphorbia heterophylla dry matter accumulation.
Summa Phytopathologica 27, 428432.
Melhorana, A.L. and Pereira, F.A.R. (1999) Eficincia do
herbicida lactofen no controle de Euphorbia heterophylla,
resistente aos herbicidas inibidores da enzima acetolactato sintase (ALS). Documentos-Embrapa Agropecuria
Oeste 3, 1114.
Nechet, K.L., Barreto, R.W. and Mizobuti, E.S. (2006) Bipolaris euphorbiae as a biological control agent for wild
poinsettia (Euphorbia heterophylla): host-specificity and
variability in pathogen and host populations. BioControl
51, 259275.
Pereira, J.M., Barreto, R.W., Ellison, A.C. and Maffia, L.A.
(2003). Corynespora casiicola f. sp. lantanae: a potential
biocontrol agent from Brazil for Lantana camara. Biological Control 26, 2131.
R Development Core Team (2006). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0.
Available at: http://www.R-project.org.
Saari, L.L., Cotterman, J.C. and Thill, D.C. (1994) Resistance
to acetolactate syntase inhibiting herbicides. In: Pauls,
E.D. and Holtum, J.A.M. (eds) Herbicide Resistance in
Plants: Biology and Biochemistry. CRC Press, Boca Raton, USA, pp. 83142.
SBCPD (1995) Procedimentos para Instalao, Avaliao e
Anlise de Experimentos com Herbicidas. Sociedade Brasileira da Cincia das Plantas Daninhas, Londrina, Brazil,
42 pp.
Sanogo, S., Yang, X.B. and Scherm, H. (2000) Effects of
herbicides on Fusarium solani f.sp. glycines and development of sudden death syndrome in glyphosate-tolerant
soybean. Phytopathology 90, 5765.
Silva, A.A. and Alto, I.F. (1993) Efeitos do nicosulfuron sobre a cultura do milho e no controle de plantas daninhas.
In: Resumos XIX Congresso Brasileiro de Herbicidas e
Plantas Daninhas. IAPAR, Londrina, Brazil, p. 153.
Vieira, B.S. and Barreto, R.W. (2005) Lewia chlamidosporiformans sp. nov. from Euphorbia heterophylla. Mycotaxon 94, 245248.
Walker, L. (1980) Production of spores for field studies.
Advances in Agricultural Technology 12, 15.
Yorinori, J.T. (1985) Biological control of wild poinsettia
(Euphorbia heterophylla) with pathogenic fungi. In:
Delfosse, E.S. (ed.) Proceedings of the VI International
Symposium on Biological Control of Weeds. Agriculture
Canada, Vancouver, Canada, pp. 677681.
Yorinori, J.T. (1987) Controle biolgico de ervas daninhas
com microrganismos. In: Anais da II Reunio sobre Controle Biolgico de Doenas de Plantas. Escola Superior
de Agricultura Luis de Queiroz, Piracicaba, So Paulo,
Brazil, pp. 2030.
Yorinori, J.T. and Gazziero, D.L.P. (1989) Control of wild
poinsettia (Euphorbia heterophylla) with Helminthosporium sp. In: Delfosse, E.S. (ed.) Proceedings of the VII
International Symposium on Biological Control of Weeds.
Rome Istituto Sperimentale per la Patologia Vegetale,
Rome, Italy, pp. 571576.
698
Sustainable management based on
biological control and ecological restoration
of an alien invasive weed, Ageratina
adenophora (Asteraceae) in China
F. Zhang,1 W.-X. Liu,2 F.-H. Wan2 and C.A. Ellison3
Summary
Crofton weed, Ageratina adenophora (Sprengel) R. King and H. Robinson, originally from Central
America, was introduced into China in the 1940s. The weed spreads rapidly and is seriously damaging grasslands and hindering livestock production in southwestern China. To tackle the weed problem
and allow the sustainable use of pastures, an integrated strategy, based on biological control and
habitat restoration, is being explored. In 1983, a gall fly Procecidochares utilis Stone, originating
from Mexico, was introduced from Tibet into Yunnan Province for the control of crofton weed. The
current efficacy of this agent was investigated, but no significant control effect was found due to
more than 60% parasitism by native parasitoids. Surveys were undertaken to identify any indigenous
fungal pathogens infecting the weed. At least six strains of Alternaria alternata (Fr.) Keissler and a
Pestalotiopsis sp. were isolated from leaves. One strain of Alt. alternata was selected for further study
as a prospective mycoherbicide. Field trials on ecological restoration using competitive native plants
and forages showed that A. adenophora was less interspecifically competitive than Setaria sphacelata
(Schum.) Stapf. ex Massey cv. Narok.
Keywords: crofton weed, natural enemies, competitive weed replacement, IPM.
Introduction
The crofton weed Ageratina adenophora (Sprengel)
R. King and H. Robinson (Asteraceae) (Synonym
Eupatorium adenophorum) from Central America was
introduced into China in the 1940s. The weed has distributed rapidly and is seriously damaging grasslands
and livestock production in the southwestern China
provinces of Yunnan, Guizhou, Sichuan, Guangxi and
Tibet (Xiang, 1991; Qiang, 1998). Its successful invasion can be attributed to its strong adaptability, competitive ability in new invaded areas, abundant seed
production and paucity of natural enemies, compared
with its native range (Qiang, 1998; Wang, 2005). A.
CABI South-East and Eastern Asia China, 12 Zhongguancun Nandajie, Beijing 100081, China.
2
Center for Management of Invasive Alien Species, Ministry of Agriculture, Institute of Plant Protection, Chinese Academy Agricultural Sciences, Beijing 100094, China.
3
CABI Europe Bakeham Lane, Egham, Surrey TW20 9TY, UK.
Corresponding author: C. Ellison <[email protected]>.
CAB International 2008
1
adenophora is still continuing to spread with an average expansion rate of 20 km/year throughout the south
and middle subtropical zones, and 6.8 km/year in north
subtropical areas (Wang and Wang, 2006). It has invaded meadow, forest and wetland, forming single
dominant communities over a short period of time and
thus has caused the decline and disappearance of the
original plant community. It releases allelopathic substances into the soil, via the roots, which can inhibit
seed germination of neighbouring plant species (Tripathi et al., 1981; Baruah et al., 1994; Zheng and Feng,
2005). However, allelopathic effects on susceptible
plants, such as Chromolaena odorata (L.) R.M. King
and H. Robinson, Bidens pilosa L., Ageratum conyzoides L. and Gynura sp., were not significant at the seedling stage of A. adenophora in shaded plots (Wang and
Feng, 2006). This weed also exhausts arable soil fertility
due to its strong absorption of soil nutrients (Liu et al.,
1989). Moreover, A. adenophora threatens the health
of livestock as the branches and leaves are poisonous
to domestic animals, particularly horses (OSullivan,
1979; Kaushal et al., 2001; Wang, 2005).
699
XII International Symposium on Biological Control of Weeds
A. adenophora has become one of the worst invasive
alien species (IAS) in China. In 2003, A. adenophora
was one of 15 IAS listed in the White Paper of Primary IAS, released by State Environmental Protection
Administration of China. Enormous efforts have been
made by Central Government (through the Ministry of
Agriculture) and local government, to control and/or
eliminate A. adenophora in newly invaded areas, and
control methods have been relying on chemical herbicides and mechanical removal. However, A. adenophora
continues to spread, expanding its range principally toward the vast area of southern and southcentral China
(Wang and Wang, 2006). This paper summarizes a
sustainable management approach, based on biological control and ecological restoration, currently being
developed for crofton weed in China.
Sustainable management measures
for Ageratina adenophora
Classical biological control with a gall fly,
Procecidochares utilis Stone
Procecidochares utilis, originating from Mexico, is
a gall fly that forms galls in the stem of A. adenophora.
It lays eggs on the stem apex and on hatching the larvae tunnel into the stem. In response to larvae presence, a gall forms in the stem, which may contain from
1 to 23 larvae (Bennett and Van Staden, 1986). Galls
have been shown to cause severe stunting, a reduction
in flowering and seed set and may result in ultimate
death of the plant when they occurred in large numbers (Bennett and Van Staden, 1986). In 1984, P. utilis
was introduced from Nienamu county Tibet into Kunming, Yunnan Province, for the control of crofton weed
(Wei et al., 1989). Since its introduction, field populations of P. utilis have been successfully established in
the introduced areas in Yunnan, Guizhou and Sichuan
provinces (Dai et al., 1991; Chen and Guan, 1994; Li
et al., 2006; Wang et al., 2006). The gall fly is multivoltine with the number of generations varying across
the weeds invaded range in China (four in Kunming,
Yunnan Province; five in southwest Guizhou Province; and six generations in the warmer lowlands of
Sichuan).
A field study in 19901994 showed that infestation
rate of P. utilis in the released area was approximately
1037% on average, but could reach up to 8596%. The
average annual dispersal rate of P. utilis was 2025 km
(max. 40 km) from original release sites, facilitated by
the southwest monsoon (Chen and Guan, 1994). Infected
plants had a significantly decreased flower number,
as well as seed weight, size and percentage germination (Wang et al., 2006). It was concluded that the gall
fly plays an important role in suppressing population
growth of crofton weed as well as reducing its rate of
spread (Chen and Guan, 1994; Wang et al., 2006).
However, the dispersal of the gall fly has been lagging behind the spread of A. adenophora (Li et al.,
2006). In addition, the overall impact on the weed in
the field has been disappointing, and a number of contributory factors have been investigated.
P. utilis shows highly selective oviposition behaviour, leading to a restricted population size in the
field. The gall fly will only oviposit on the stem apex,
showing a preference for younger plant branches.
This results in a reduction in eggs laid during the
flowering season of the plant, when only less optimal oviposition sites are available (Wei et al., 1989;
Li et al., 2006).
Li et al. (2006) argued that P. utilis had no significant effects on the number of blossom branches, capitula and seeds produced by A. adenophora, which
indicated limited impact on reproduction and spread
of the weed.
Significant differences were found by Li et al.
(2006) in the infestation rates based on individual
plants (71.7%) and plant branches (17.3%). Since
A. adenophora produces a mean of 21 branches per
plant, this suggests that P. utilis is unlikely to have a
significant impact on the whole weed population.
Recent field surveys have also revealed that P. utilis
is parasitized by indigenous parasitoids; Torymoides kiesenzuetteri (Mayr), Eupelmus urozonus Dalman, Bracon sp., Eurytoma spp., Sphegigaster spp.
and Pteromalus spp., which might also explain its
lower-than-anticipated control effect. B.P. Li (2006,
personal communication) observed heavy parasitism of more than 60% in the field.
Biological control with fungal pathogens
Fungal pathogens of A. adenophora have been studied since the first report of Phaeoramularia eupatoriiodorati (Yen) Liu and Guo (= Cercospora eupatorii
Peck) in Australia in 1954 (Auld, 1969). This fungus
was accidentally introduced, probably carried by P.
utilis adult flies when they were introduced from Hawaii. This leaf-spot fungus is widespread and is responsible for significant premature leaf abscission, and is
particularly prevalent in cleared areas. It is believed
that the impact of the weed in Australia is considerably lessened by the pathogen (R.E. McFadyen, 2007,
personal communication). It was released in South Africa in 1987, but although it has established, its impact
on weed populations is minimal (Julien and Griffiths,
1998).
In Yunnan province, P. eupatorii-odorati is a common disease of A. adenophora and reaches epidemic
levels during June and July. It causes reductions in the
photosynthetic rate, transpiration rate and chlorophyll
content, thus reducing plant height and number of
leaves and flowers (Yang and Guo, 1991; Guo et al.,
1992). In laboratory studies, the optimal temperature for
700
Sustainable management based on biological control and ecological restoration of an alien invasive weed
conidia germination of P. eupatorii-odorati was 25oC
and the optimal pH was 5.0, and, like most pathogens,
it requires free water on the plant surface to develop
infection (Guo et al., 1992). This study recommended
that the ideal time, considering temperature and water
requirements, to release this pathogen into the area
where it does not occur in Yunnan province, was July
and August.
In China, indigenous fungal pathogens infecting
A. adenophora, which may have potential as mycoherbicides to control the weed (Qiang, 1998), have also been
investigated. At least six strains of Alternaria alternata
(Fr.) Keissler and a Pestalotiopsis sp. were isolated
from leaves of A. adenophora (Wan et al., 2001; W.X.
Liu, unpublished data). Necrotic leaf spots on A. adenophora leaves were observed 24 hours after treatment
with a toxin extract produced by Alt. alternata, at a dosage of 50300 mg/ml (Dai et al., 2004). Recent work by
Qiang et al. (2006) has found that mycelial preparations
of Alt. alternata are more effective infection propagules
than conidia. The assessment of the potential of these
local fungal pathogens is at the early stages of laboratory research and small-scale field trials.
Ecological restoration by competitive
replacement with native plants and
forage grasses
Ecological restoration involves the re-establishment
of the structure and function of an ecosystem, primarily to achieve a pre-disturbance state. In China, studies
have focused on restoration of both natural habitats with
the native flora, and pastureland with forage grasses.
The selection of alternative plants has been based on a
number of factors:
A. adenophora has a strong allelopathic effect on
other plants, and degrades the soil; therefore, replacement plants need to be tolerant of the prevailing soil
conditions, as well as being strongly competitive.
Seeds of crofton weed are very sensitive to sunlight
during germination and demonstrate less interspecific competitive ability at the seedling stage. Therefore, the chosen alternative plants should be easy to
grow, with a high rate of growth, and a canopy density that can reach 70% within a short period.
In the pasturelands of southwestern China, it is also
critical to consider field application and technology
dissemination within the rural communities. Therefore, to enhance development in rural areas, it would
be favourable to choose alternative plants that have
a high economic value.
Following these criteria, the candidate species selected were Brachiaria subquadripara (Trin.) Hitchc,
Cajanus cajan (L.) Mjllspaugh, Lolium perenne L.,
Pennisetum clandestinum Hochst. ex. Chiov., Pennisetum sinese Roxb., Setaria sphacelata (Schum.) Stapf.
ex. Massey cv. Narok, Setaria viridis (L.) Beauv. and
Trifolium repens L.
In a greenhouse study using plants grown in pots,
crofton weed was shown to be less interspecifically
competitive than S. viridis and S. sphacelata, when
comparing their shoot height and biomass at seedling
stage (Jiang, 2007). The addition of nitrogen enhanced
the relative competitive ability of the forage species,
which indicated that improvement of fertilization could
be of additional benefit to the restoration of the pasture.
Field plots studies by Jiang (2007) confirmed that the
growth (i.e. biomass) of A. adenophora was heavily
suppressed by S. sphacelata, when both species were
planted in a mixture at different densities, but no impacts were observed on the growth of S. sphacelata.
Compared with the biomass produced in its monoculture plots, the biomass of A. adenophora decreased by
over 70% in the mixed plots. The relative competitive
ability of S. sphacelata was still higher than that of
A. adenophora after 16 months in the field. This approach has been widely used and recommended in the
invaded areas for pastures and animal farming.
Discussion
Field management of A. adenophora in China has been
practised extensively using chemical and mechanical
control, and hand removal. However, these methods
are not sustainable, result in environmental problems,
are labour-intensive and expensive, and still fail to
contain the weed. Classical biological control was
not effective probably due to the impact of the indigenous parasitoids on P. utilis. Biological control using fungal pathogens as mycoherbicides is still at the
research stage and it is too early to say how effective
they will be in the field. Recently many efforts have
been made on ecological restoration of invaded pastures by native plants and forage grasses. However,
ecological restoration itself is also not a sole solution
to resolve the weed problem since it would not be feasible to restore all of the vast areas already invaded
by the weed.
Studies on the mechanisms involved in invasion
biology have also shed light on sustainable management of A. adenophora. It appears that A. adenophora
colonizes human-altered environments (i.e. roads and
streams) to which it is better adapted than native species, rather than invading undisturbed habitats and displacing locally adapted native species (Sax and Brown,
2000; Lu and Ma, 2006). Its invasion success is significantly negatively correlated with native plant diversity,
and reduced native species cover appears to facilitate
the invasion. Management to control A. adenophora in
southwest China should also focus on habitats along
roads and streams, which provide a significant source
of seed and facilitate the spread of the weed into other
habitats. Increasing canopy cover of native species
701
XII International Symposium on Biological Control of Weeds
could help control A. adenophora invasion in these
area (Lu and Ma, 2006).
The biological control programme against mist flower,
Ageratina riparia (Regel) R. King and H. Robinson
(Asteraceae) in Hawaii is considered one of the most
successful undertaken anywhere in the world (Morin
et al., 1997). The most important control agent for
A. riparia in Hawaii appears to be the fungus Entyloma
ageratinae Barreto and Evans, followed by the gall fly
Procecidochares alani Steyskal and then the plume moth
Oidaematophorus beneficus Yano and Heppner. It is interesting to note that a range of parasitoids destroyed up
to 50% of P. alani in galls in Hawaii and parasitism was
the main factor limiting its effectiveness in Queensland
(Morin et al., 1997). Morin et al. (1997) suggested that
the fungus and the gall fly were complementary in their
activity, and that both should be introduced into New
Zealand for a biological control of A. riparia.
With all these experiences, we believe that a sustainable management programme could be established
for A. adenophora in China using an integrated approach, underpinned by classical biological control.
New agents need to be sourced from the centre of origin of the weed (Mexico), which are complementary
to the two agents already impacting (albeit mildly) on
A. adenophora. For example, a lepidopterous and a
curculionid stem borer have been identified from Mexico (Osborn, 1924) as well as Baeodromus eupatorii
(Arthur) Arthur, a rust fungus from Central America
(Buritic and Hennen, 1980). In high value land, the
use of indigenous fungal pathogens as mycoherbicides
may be economically viable. Ecological restoration
with native plants and forage grasses, integrated with
other measures such as mechanical removal, fire and
chemical control, may provide the optimum approach.
Over the vast areas of natural habitats, classical biological control will need to form the foundation of the
integrated approach.
Acknowledgements
This study was funded by the International Science
and Technology Cooperation Programme of Ministry
of Science and Technology, China (2005DFA31090),
the National Basic Research and Development Programme, China (2002CB111400), and the CABI Partnership Facility (VMO56105).
References
Auld, B.A. (1969) Incidence of damage caused by organisms
which attack crofton weed in the Richmond-Tweed region
of New South Wales. Australian Journal of Science 32,
163.
Baruah, N.C., Sarma, J.C., Sarma, S. and Sharma, R.P. (1994)
Seed germination and growth inhibitory cadinenes from
Eupatorium adenophorum Spreng. Journal of Chemical
Ecology 20, 18851892.
Bennett, P.H. and Van Staden, J. (1986) Gall formation in
crofton weed, Eupatorium adenophorum Spreng. (syn.
Ageratina adenophora), by the Eupatorium gall fly Procecidochares utilis Stone (Diptera: Trypetidae). Australian
Journal of Botany 34, 473480.
Buritic, P. and Hennen, J.F. (1980) Pucciniosireae (Uredinales, Pucciniaceae). Flora Neotropica Monograph 24,
2425.
Chen, S.B. and Guan, D.S. (1994) Biological character observation and biological control of Procecidochares utilis
Stone. Southwest China Journal of Agricultural Sciences
7, 98102.
Dai, C., Wei, Y. and He, D.Y. (1991) A study on effect of
Procecidochares utilis on control of Eupatorium adenophorum. Journal of Weed Science 5, 2429.
Dai, X.B., Chen, S.G., Qiang, S., An, C.F. and Zhang, R.X.
(2004) Effect of toxin extract from Alternaria alternata
(Fr.) Keissler on leaf photosynthesis of Eupatorium adenophorum Spreng. Acta Phytopathologica Sinica 34,
5560.
Guo, G.Y., Yang, Y.R., Liu, Y., Ma, J., Xu, L.H. and Jiang,
C.L. (1992) Studies on the biological characteristics of
Mycovellosiella eupatorii odorai (Yen) Yen, a potential pathogen for the biological control of crofton weed,
Eupatorium adenophorum. Chinese Journal of Biological
Control 8, 120124.
Jiang, Z.L. (2007) Ecophysiological mechanisms of competition between the invasive weed Ageratina adenophora
(Asteraceae) and non-invasive herbaceous plants. PhD
dissertation. Chinese Academy of Agricultural Sciences,
Beijing, China, 107 pp.
Julien, M.H. and Griffiths, M.W. (1998) Biological Control
of Weeds: A World Catalogue of Agents and their Target Weeds, 4th edn. CABI Publishing, Wallingford, UK,
223 pp.
Kaushal, V., Dawra, R.K., Sharma, O.P. and Kurade, N.P.
(2001) Biochemical alterations in the blood plasma of rats
associated with hepatotoxicity induced by Eupatorium
adenophorum. Veterinary Research Communications 25,
601608.
Li, A.F., Gao, X.M., Dang, W.G., Huang, R.X., Deng, Z.P. and
Tang, H.C. (2006) Parasitism of Procecidochares utilis
and its effect on growth and reproduction of Eupatorium
adenophorum. Journal of Plant Ecology 30, 496503.
Liu, L.H., Liu, W.Y., Zheng, Z. and Jing, G.F. (1989) The
characteristic research of autecology of pamakani (Eupatorium adenophorum). Acta Ecologica Sinica 9, 6670.
Lu, Z.J. and Ma, K.P. (2006) Spread of the exotic crofton
weed (Eupatorium adenophorum) across southwest China
along roads and streams. Weed Science 54, 10681072.
Morin, L., Hill, R.L. and Matayoshi, S. (1997) Hawaiis
successful biological control strategy for mist flower
(Ageratina riparia) can it be transferred to New Zealand? Biocontrol News and Information 18, 7788.
Osborn, H.T. (1924) A preliminary study of the pamakani
plant (Eupatorium glandulosum H.B.K.) in Mexico with
reference to its control in Hawaii. Hawaiian Planters
Records 24, 546559.
OSullivan, B.M. (1979) Crofton weed (Eupatorium adenophorum) toxicity in horse. Australian Veterinary Journal
62, 3032.
Qiang, S. (1998) The history and status of the study on crofton
weed (Eupatorium adenophorum Spreng.), a worst world-
702
Sustainable management based on biological control and ecological restoration of an alien invasive weed
wide weed. Journal of Wuhan Botanical Research 16,
366372.
Qiang, S., Zhu, Y., Summerell, B.A. and Li, Y. (2006) Mycelium of Alternaria alternata as a potential biological
control agent for Eupatorium adenophorum. Biocontrol
Science and Technology 16, 653668.
Sax, D.F. and Brown, J.H. (2000) The paradox of invasion.
Global Ecology and Biogeography 9, 363371.
Tripathi, R.S., Singh, R.S. and Rai, J.P.N. (1981) Allelopathic
potential of Eupatorium adenophorum, a dominant ruderal weed of Meghalaya. Proceedings of Indian Academy
of Sciences 47, 458465.
Wan, Z.X., Qiang, S., Xu, S.C., Shen, Z.G. and Dong, Y.F.
(2001) Culture conditions for production of phytotoxin by
Alternaria alternata and plant range of toxicity. Chinese
Journal of Biological Control 17, 1015.
Wang, J.J. (2005) Crofton weed Eupatorium adenophorum
Spreng. In: Wan, F.H., Zheng, X.B. and Guo, J.Y. (eds) Bio
logy and Management of Invasive Alien Species in Agriculture and Forestry. Science Press, Beijing, pp. 650661.
Wang, J.F. and Feng, Y.L. (2006) Allelopathy and light acclimation characteristic for Eupatorium adenophorum seedlings grown in man-made communities. Acta Ecologica
Sinica 26, 18091817.
Wang, R. and Wang, Y.Z. (2006) Invasion dynamics and potential spread of the invasive alien plant species Ageratina
adenophora (Asteraceae) in China. Diversity and Distributions 12, 397408.
Wang, W.Q., Wang, J.J. and Zhao, Z.M. (2006) Effects of
parasitizing on the sexual reproduction of Eupatorium
adenophorum Spreng. by Procecidochares utilis Stone at
different microhabitats. Acta Phytophylacica Sinica 33,
391395.
Wei, Y., Zhang, Z.Y. and He, D.Y. (1989) Mass rearing of
Procecidochares utilis (Diptera: Trypetidae), a biological control agent of Eupatorium adenophorum. Chinese
Journal of Biological Control 5, 4142.
Xiang, Y.X. (1991) The distribution, perniciousness and control of Ageratina adenophora. Chinese Journal of Weed
Science 4, 1011.
Yang, Y.R. and Guo, G.Y. (1991) Study on the effect of Mycovellosiella eupatorii odorai upon growth and physiological parameters of Eupatorium adenophorum. Journal
of Weed Science 5, 611.
Zheng, L. and Feng, Y.L. (2005) Allelopathic effects of
Eupatorium adenophorum Spreng. on seed germination
and seedling growth in ten herbaceous species. Acta Ecologica Sinica 25, 27822787.
703
XII International Symposium on Biological Control of Weeds
Trans-Atlantic opportunities for collaboration on classical
biological control of weeds with plant pathogens
D.K. Berner and W.L. Bruckart
Foreign Disease-Weed Science Research Unit, USDA, ARS, Ft. Detrick, MD 21702-5023, USA
In North America, introduced invasive weeds are having catastrophic effects on agricultural and natural, wild ecosystems. Many of these weeds have been introduced from Eurasia, and the only economically feasible means for controlling them is through classical biological control. This situation is the
same in Europe with invasive weeds introduced from North America. Development of pathogens for
classical biological control involves collection of pathogens in the native habitat of the target weed,
testing the pathogens in quarantine in the country of prospective release or in fields in the country(ies)
of origin, and release in the non-native environment. However, discovery of new pathogens has mostly
been the purview of the explorer from the country with the weed problem. A more efficient approach
would be for explorers from the country of origin to discover pathogens on native species and send
them to the country(ies) where the weed is a problem. Given established collaboration, this would
work reciprocally with pathogens being discovered in North America and Eurasia, tested in the native
range, and sent to the collaborator(s) for rapid advancement of pathogens toward release. However,
there are some basic requirements for this to work. Mutual advantages, requirements, and an example
are presented.
Factors affecting success and failure of Diorhabda
elongata releases for control of Tamarix spp.
in western North America
T.L. Dudley,1 P. Dalin,1 D.W. Bean,2 D.L. Thompson,3 D. Kazmer,4
D. Eberts5 and C.J. DeLoach6
Marine Science Institute, University of California, Santa Barbara, CA, USA
2
Colorado Department of Agriculture, Palisade, CO, USA
3
Entomology, Plant Pathology and Weed Science, New Mexico State University, Las Cruces, NM, USA
4
USDA Agricultural Research Service, Sidney, MT, USA
5
USDI Bureau of Reclamation, Denver, CO, USA
6
USDA Agricultural Research Service, Temple, TX, USA
1
The 1999 cage releases and 2001 research releases of a central Asian form of the saltcedar leaf beetle,
Diorhabda elongata (Chrysomelidae) has led to successful establishment of the agent in some locations, substantial target mortality at four release sites, and numerous desired responses by ecosystem
and community elements. However, the majority of releases have resulted in limited establishment or
failure. Three ecological factors appear to account for these failures: (1) this beetle from 44N latitude
enters reproductive diapause in response to daylength too early in the season for populations to establish at latitudes lower than ca. 38; (2) invertebrate predation can inhibit population establishment,
particularly where developmental responses are not ideal; and (3) the target species in some regions is
a poor quality host for the approved agents. Other ecotypes from the D. elongata species complex from
different latitudes and bioregions may provide phenotypic traits that allow successful establishment at
different latitudes and against different target plant genotypes. Five ecotypes are currently being cagetested across three latitudinal gradients in North America to determine which develops, reproduces,
and over-winters most successfully in regions where saltcedar biocontrol is desired.
CAB International 2008
704
Abstracts: Theme 9 Management Specifics, Integration, Restoration and Implementation
Advances in Striga mycoherbicide research
and development: implications and future
perspective for Africa
A. Elzein,1 J. Kroschel,2 P. Marley3 and G. Cadisch1
Institute for Plant production and Agroecology in the Tropics and Subtropics (380),
University of Hohenheim, D-70593 Stuttgart, Germany
2
Integrated Crop Management Division, International Potato Center (CIP), Av. La Molina 1895,
Apartado 1558, Lima 12, Peru
3
Department of Crop Protection, Faculty of Agriculture/Institute for Agricultural Research,
Ahmadu Bello University, Samaru, Zaria, Nigeria
1
Striga spp. are important pests of cereals in semiarid, tropical Africa. An integrated approach, in which
biocontrol represents an important component, appears to be an ideal management strategy for Striga.
Our recent research focuses on the development of appropriate mycoherbicidal formulations and delivery systems to facilitate practical field application of the potential Striga mycoherbicides Fusarium
oxysporum (Foxy 2 & PSM197). Hence, Pesta formulation, made by encapsulating fungal inoculum
in a matrix composed of durum wheat-flour, kaolin, and sucrose, was developed. Further, a seed treatment technology for coating sorghum and maize seeds for further minimizing the inoculum amount
and facilitating delivery of Striga mycoherbicides was provided. Both formulations showed promising
efficacy in controlling Striga. The integration of Striga mycoherbicides with Striga-resistant sorghum
and maize clearly enhanced the efficacy of both mycoherbicides in controlling Striga under field conditions. Further, both mycoherbicides maintained excellent viability on Pesta products and treated seeds
after one year of storage, sufficient for their use under practical conditions of storage, handling, and
delivery. The compatibility and suitability of Pesta and seed treatment technology for formulating
and delivering Striga mycoherbicides can contribute to solving the primary difficulties of large-scale
underemployment of Striga mycoherbicides in Africa. Strategies for integrated Striga control using
mycoherbicides are proposed for African subsistence farmers.
Multispectral satellite remote sensing of water hyacinth
at small extents a monitoring tool?
J.T. Fisher, B.F.N. Erasmus and M.J. Byrne
School of Animal, Plant and Environmental Sciences, University of the Witwatersrand,
Johannesburg, South Africa
Water hyacinth (Eichhornia crassipes) is widespread in South Africa, occurring in inaccessible and
often small water bodies. Our objective was to determine if medium resolution (1030 m) satellite
remote sensing using existing imagery can be used to monitor the growth and health of water hyacinth
populations at small extents in South Africa. Measuring the area covered by water hyacinth, as a result
of classification accuracy, at decreasing resolutions was evaluated. The usefulness of satellite imagery
for predicting health of plants using the Normalized Difference Vegetation Index, and the feasibility
of using satellite remote sensing, were assessed. Classification accuracy is greatest on water bodies
larger than 5 ha with greater than 30% cover of water hyacinth. SPOT IV imagery is the most costeffective per image. IKONOS imagery (4 m resolution) should be used to monitor sites smaller than
5 ha; however, ground sampling smaller sites is more cost-effective. SPOT IV imagery is recommended
to determine the extent of a water hyacinth mat. Seasonal or biannual monitoring is adequate to detect
change based on field biomass measurements. The nature of the change in extent can be determined at
sites larger than 50 ha using SPOT IV imagery, and at smaller sites using IKONOS imagery.
705
XII International Symposium on Biological Control of Weeds
Innovative tools for the transfer of invasive
plant management technology
M.J. Grodowitz,1 S.G. Whitaker,1 J.A. Stokes2 and L. Jeffers2
US Army Engineer Research and Development Center, 3909 Halls Ferry Road, Vicksburg,
MS 39180-6199, USA
2
BITS, US Army Engineer Research and Development Center, 3909 Halls Ferry Road, Vicksburg,
MS 39180-6199, USA
1
Managing nuisance/invasive plants is an ever growing problem throughout the world today. Different methods of control (i.e. biological, chemical, mechanical, etc.) are available but information is,
at times, difficult to obtain especially in one easy to access location. Innovative tools are needed that
allow for rapid and efficient access to information on the identification and management of these problem plant species. Toward this goal, the US Army Engineer Research and Development Center has
produced two plant management information/expert systems that allow for ready information access.
These systems, PMIS and APIS, allow for information access on a variety of plant management topics
including plant identification, plant biology as well as on various management techniques including
an in-depth section on biocontrol. The systems come in two formats including CD and web-based versions. A new version of APIS is currently in beta testing and runs entirely on handheld personal data
assistants (PDA) and smartphones using Windows Mobile technology. This mobile format allows for
information access in even the most remote locations. With the availability of these systems, plant
managers now have better and more efficient access to information on available technologies to manage invasive plant species.
Physiological age-grading techniques to assess
reproductive status of insect biocontrol agents of
aquatic plants
M.J. Grodowitz1 and L. Lenz2
US Army Engineer Research and Development Center, Vicksburg, MS 39180, USA
2
Former address: University of North Texas, Denton, TX 76203, USA
Physiological age grading is used to assess reproductive health and status of many insect species using
ovarian morphology to ascertain reproductive status, reproductive history, and eggs oviposited. Three
physiological age-grading systems have been developed for agents of aquatic plants including two
weevil species; Neochetina eichhorniae and Euhrychiopsis lecontei, and the leaf-mining fly, Hydrellia
pakistanae. All three systems utilize ovarian morphology to determine changes in reproductive condition. The two weevil species have similar ovarian developments/morphologies and thereby utilize
almost identical systems, where changes in fat body, cuticular hardness, and follicular relics give rise
to three nulliparous and three parous stages. Ovarian morphology/development differs in the fly but
also uses characteristics of follicular relics to distinguish three nulliparous and four parous classes.
Strategies of ovarian development differ between these species based on longevity and habitat. The
weevils are long-lived and reside in relatively protected habitats and mature/deposit eggs individually
throughout most of their life. The fly is short-lived, resides in open areas, and only oviposits when an
entire batch or compliment of eggs are mature. Interestingly, the fly emerges as an adult with mature
eggs ready for oviposition; another strategy to maximize egg production and minimize predation over
a short life span.
706
Abstracts: Theme 9 Management Specifics, Integration, Restoration and Implementation
Use of multi-attribute utility analysis for the identification
of aquatic plant restoration sites
M.J. Grodowitz,1 R.M. Smart,2 J. Snow,3 G.O. Dick3 and J.A. Stokes4
US Army Engineer Research and Development Center, Vicksburg, MS 39180, USA
US Army Engineer Research and Development Center, Lewisville Aquatic Ecosystem Research Facility,
Lewisville, TX 75057, USA
3
University of North Texas, Institute of Applied Science, Lewisville Aquatic Ecosystem Research Facility,
Lewisville, TX 75057, USA
4
BITS, US Army Engineer Research and Development Center, Vicksburg, MS 39180, USA
1
Presence of a diverse native plant community has been shown to enhance weed management especially
in the presence of a capable herbivore. Therefore, an important consideration when designing weed
biocontrol projects is the implementation of well-designed revegetation programmes. Such is the case
for the management of submersed aquatic plants; for example, the greatest hydrilla declines occur in
the presence of both native plants and sustained fly herbivory. While progress has been made in developing techniques for native aquatic plant culturing/planting, only limited information is available that
allows non-technical personnel the ability to select suitable sites for re-vegetation efforts. To solve this
problem, a decision support model was developed using multi-attribute utility analysis (MAU) where
re-vegetation experts identified 11 important site characteristics ranging from shoreline gradient to
sediment type. For each characteristic, utility functions were developed, which incorporate probabilities for site selection across a wide range of site characteristic values. Once the information is collected
and entered, the system provides an instantaneous, prioritized listing of sites suitable for re-vegetation
based on expert opinion and facts. A Web-based version has been developed allowing non-technical
personnel easy and efficient access to this important aquatic plant revegetation site selection tool.
Induced resistance in plants friend or foe to
biological control?
P.E. Hatcher
School of Biological Sciences, University of Reading, Reading, UK
Over the last 10 years, research into induced resistance (IR) in plants to pathogens and insects has developed from a theory with a little empirical evidence to a major field within plant products identified.
Does IR affect biological control? To date, reducing the effects of IR on biocontrol agents has only
been considered in a few cases, in which inhibitors of plant defence mechanisms have been combined
with mycoherbicides. Yet IR could have a wide effect, especially with multiple sequential releases (To
what extent is the interspecific competition and interference often recorded here due to IR?), or when
biocontrol agents are introduced to indigenous plants that already have some damage, and thus may already be induced. Can adjuvants already used in mycoherbicide formulations stimulate or inhibit IR?
These and other questions will be addressed in this paper, which seeks to raise awareness of this
topic by giving an overview of IR studies in weeds, its potential affects on biocontrol organisms, how
this can be managed, and whether IR can be harnessed to the benefit of biological control.
707
XII International Symposium on Biological Control of Weeds
Turning the tide using the sterile insect technique
to mitigate an unwanted weed biocontrol agent
S.D. Hight,1 J.E. Carpenter,2 S. Bloem3 and K.A. Bloem3
USDA-ARS-CMAVE, 6383 Mahan Drive, Tallahassee, FL 32308, USA
USDA-ARS-CPMRU, 2747 Davis Road, Bldg. 1, Tifton, GA 31794, USA
3
USDA-APHIS-CPHST, 1730 Varsity Drive, Raleigh, NC 27606, USA
1
The most successful programme of classical biological control of weeds has been the control of invasive prickly pear cactus (Opuntia spp.) by the Argentine cactus moth Cactoblastis cactorum. However,
the moth has now become an invasive pest in the southeastern USA and its ability to dramatically
control its host plant raises concerns for the safety and survival of the many ecologically, agriculturally,
and culturally important Opuntia spp. in southwestern USA and Mexico. The sterile insect technique
(SIT) has been developed for this insect as an areawide control measure. A validation/implementation
study of the SIT coupled with sanitation efforts (removal of eggsticks, infested pads/larvae, and pupae)
has limited the western spread of the moth. Sterile insects released in the field were highly competitive
against wild moths. Competitiveness was evaluated for males by their recapture rate in pheromonebased monitoring traps and the proportion of sterile eggsticks produced as a result of sterile males mating with wild females. Continued refinement of the SIT against C. cactorum represents an opportunity
to manage this biological control agent become pest. If implemented rapidly on new introductions, SIT
can also serve as an effective risk management tool to eradicate other new invasive pests.
Integrated weed control using a retardant dose of
glyphosate: a new management tool for water hyacinth
A.M. Jadhav,1 A. Kirton,2 M.P. Hill,3 M. Robertson4 and M.J. Byrne1
Private Bag 3, School of Animal, Plant and Environmental Sciences, University of Witwatersrand,
Johannesburg 2050, South Africa
2
Schools of Statistics and Actuarial Sciences, University of the Witwatersrand,
Johannesburg 2050, South Africa
3
Department of Zoology and Entomology, Rhodes University, PO Box 94, Grahamstown, South Africa
4
Department of Zoology, University of Pretoria, Pretoria, South Africa
1
Water hyacinth, Eichhornia crassipes (Mart.) Solms-Laubach has a major impact on aquatic ecosystems in South Africa despite biocontrol, which remains hampered by high nutrient levels and low
temperatures. Often, the biocontrol agents are unable to overcome rapid weed growth, necessitating
the need for intervention by herbicidal control. However, lethal doses of herbicides have harmful environmental consequences and kill the biocontrol agents by removing their habitat. The weed resurges
from seed or overlooked individuals. However, glyphosate sprayed at a 0.8% concentration retards the
growth and vegetative reproduction of the weed without detrimental effects on the biocontrol agents.
High nutrient levels and season did not override the retardant effects of the herbicide. If spray volumes
are adjusted to plant biomass, this method offers a low impact integrated weed management tool for
weed-affected water systems.
708
Abstracts: Theme 9 Management Specifics, Integration, Restoration and Implementation
Avoiding biotic interference with weed
biocontrol insects in Hawaii
M.T. Johnson
Institute of Pacific Islands Forestry, USDA Forest Service, PSW Research Station, Volcano, HI, USA
Equable climates in tropical habitats favor high year-round populations of generalist predators, parasitoids, and pathogens, which may hamper establishment and limit population growth of insects released
for weed biocontrol. In Hawaii, the many past introductions for insect biocontrol as well as accidental
introductions of herbivore enemies have led to awareness of problems with biotic interference in weed
biocontrol. A variety of strategies are available to help predict and avoid biotic interference with future introductions. Matching checklists of Hawaiian arthropod enemies against enemies of prospective
biocontrol agents known from their areas of origin provides a basis for early prioritization of candidates. For our most promising agents, we conduct pre-release quarantine tests to detect vulnerability to
selected enemies. Retrospective studies also are useful: evaluating populations of lepidopteran agents
released in the past helps us better understand impacts of natural enemies such as egg parasitoids.
Field studies with chrysomelids provide data on potential enemies of a candidate agent in this family, which has been seldom used for biocontrol in Hawaii. New approaches to studying multitrophic
interactions of herbivores will contribute other tools for predicting and avoiding barriers to successful
weed biocontrol.
Sustainable management, based on biological control
and ecological restoration, of the alien invasive weed,
Ageratina adenophora (Asteraceae), in China
W-X. Liu,1 F-H. Wan,1 F. Zhang2 and C.A. Ellison3
Center for Management of Invasive Alien Species, Ministry of Agriculture, Institute of Plant Protection,
Chinese Academy Agricultural Sciences, Beijing 100094, China
2
CABI China Office, 12 Zhongguancun Nandajie, Beijing 100081, China
3
CABI UK Centre (Ascot), Silwood Park, Ascot, Berkshire SL5 7TA, UK
Crofton weed, Ageratina adenophora (Asteraceae), originally from South America, was introduced
into China in the 1940s. The weed distributed rapidly and is seriously damaging grasslands and livestock production in southwestern China. To tackle the weed problem for sustainable use of pastures,
an integrated strategy, based on biological control and habitat restoration, is being explored. In 1983,
a gall fly Procecidochares utilis, originating from Mexico, was introduced from Tibet into Yunnan
Province for the control of crofton weed. The current efficacy of this agent was investigated, but no
significant control effect was found and heavy parasitism (70%) by native wasps was responsible. Surveys were undertaken to investigate indigenous fungal pathogens infecting the weed. Three strains of
Alternaria alternata and a Pestalotiopsis sp. were isolated from leaves; a strain of the former species
(YN01) was selected for further study as a prospective mycoherbicide. Surveys in Mexico to look for
potential new classical biocontrol agents are planned for 2007. Field trials on ecological restoration by
competition replacement of native plants and forages, under different fertility and plant density ratio
conditions, were conducted. The results are discussed together with planned work on the interaction of
natural enemies and this restoration approach.
709
XII International Symposium on Biological Control of Weeds
Biological control of emerging weeds in South Africa:
an effective strategy to halt alien plant invasions
at an early stage
A.J. McConnachie,1 T. Olckers,2 A. Fourie,3 K. Ntushelo,3 E. Retief,3
D.O. Simelane,4 L.W. Strathie,1 H. Williams4 and A.R. Wood3
1
ARC-PPRI, Weeds Division, Private Bag X6006, Hilton 3245, South Africa
School of Biological and Conservation Sciences, University of KwaZulu Natal, Private Bag X01,
Scottsville 3209, South Africa
3
ARC-PPRI, Weeds Division, Private Bag X5017, Stellenbosch 7599, South Africa
4
ARC-PPRI, Weeds Division, Private Bag X134, Queenswood 0121, South Africa
Biological control of incipient or emerging weeds (plants in the early stage of invasion) is internationally an uncommon practice due to restricted research funds being allocated to weeds that have already
reached detrimental levels. Previously in South Africa, because of limited funds and, as a result, few
opportunities to conduct overseas exploration work, researchers have capitalized on their survey trips
by collecting potentially useful biological control agents from as many target plants as possible. Earlier exploration for agents against high-priority weeds thus allowed simultaneous collection of natural
enemies of low-priority weeds in the same region. Such opportunistic programmes have been beneficial for South Africa in the management of several emerging weeds. Formal classification systems,
however, have since been developed in South Africa for the prioritization of invasive alien plants. In
light of this, the Working for Water Programme, the main funding agency for weed biological control
research in South Africa since 1996, officially recognized and awarded funds for five emerging weed
programmes in 2003. This paper reviews the cases where emerging weeds were targeted for biological control in South Africa and where successes were achieved; the use of classification systems to
prioritize the management of invasive plants, and the progress achieved with the emerging weed programmes currently underway in South Africa.
Routine use of molecular tools in Australian weed
biological control programmes involving pathogens
L. Morin1,2 and D. Hartley2
1
Cooperative Research Centre for Australian Weed Management, University of Adelaide,
Waite Main Building, Urrbrae, Adelaide, Australia
2
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Molecular techniques are increasingly being utilized in classical biological control programmes for
weeds to characterize agents and targets and streamline the agent selection process. They are now routinely employed during the development and implementation of weed biological control programmes
involving rust pathogens in Australia. For example, molecular characterization of pathogens has been
used to confirm identification of a potential biocontrol candidate (Puccinia lagenophorae on Senecio
madagascariensis), determine the genetic diversity of an illegally introduced pathogen before considering introduction of additional strains (Puccinia xanthii on Xanthium occidentale), and examine
intraspecific genetic and virulence variation of a rust fungus to assist selection of an appropriate strain
for release (Puccinia myrsiphylli on Asparagus asparagoides). Molecular diagnostic tools have been
investigated as a means to confirm establishment of new pathogen strains released in areas where populations of the pathogen already exist, as strains cannot be differentiated from each other using morphological characters (Phragmidium violaceum on Rubus fruticosus agg). Molecular detection based
on polymerase chain reaction amplification will also be used to confirm establishment after release of
a systemic rust pathogen that has a long incubation period before visible symptoms develop on plants
(Endophyllum osteospermi on Chrysanthemoides monilifera subsp. monilifera).
710
Abstracts: Theme 9 Management Specifics, Integration, Restoration and Implementation
An ecological approach to aquatic plant management
R.M. Smart1 and M.J. Grodowitz2
US Army Engineer Research and Development Center, Lewisville Aquatic Ecosystem Research Facility,
Lewisville, TX 75057, USA
US Army Engineer Research and Development Center, Vicksburg, MS 39180-6199, USA
A simple, yet often used concept of integrated pest or plant management (IPM) is one where all available management options are considered as part of a toolbox or arsenal. These tools/weapons are
then used singly or in combination in an effort to maximize control without impacting the use of one
or more strategies. While this approach can be effective, it tends to provide only short-term control by
neglecting the underlying reasons for the formation of the infestations. A more prudent and ecologically compatible approach would be the use of an ecosystem-based IPM programme that relies heavily
on ecosystem management and restoration strategies and addresses causative factors that allow such
formations. A key component of an ecosystem approach to managing aquatic plants is the use of hostspecific biological control agents. Most of the economically important invasive/nuisance aquatic plants
are introduced species that have escaped their host-specific herbivores and pathogens. In addition to
their high intrinsic rates of increase, this lack of sustained feeding and resultant damage allows the
formation of extensive monospecific infestations. By re-establishing a complex of host-specific herbivores and pathogens and implementing revegetation using native plants, these invasive species can be
held at non-problem levels.
A cooperative approach to biological control of
Parthenium hysterophorus (Asteraceae) in Africa
L.W. Strathie,1 A.J. McConnachie1 and M. Negeri2
Agricultural Research Council, Plant Protection Research Institute, Private Bag X6006,
Hilton 3245, South Africa
2
Ethiopian Institute of Agricultural Research, National Plant Protection Research Centre,
PO Box 37, Ambo, Ethiopia
1
Parthenium hysterophorus (Asteraceae) is a serious invader, impacting on agriculture, biodiversity,
and human and animal health in Africa, Asia, some Pacific Islands and Australia. Dense infestations
are present and spreading in large parts of eastern and southern Africa and islands off the mainland.
Despite decades of significant control efforts against parthenium in Australia, and to a lesser degree
in India, no biological control programmes were initiated on this species in Africa until 2003, when
natural enemies were introduced into South African quarantine in a programme targeting emerging (or
incipient) weeds. Three damaging insect species and a pathogen that target flowers, stems, and leaves
were imported. In 2005, a cooperative project was initiated for the management of parthenium in
eastern and southern Africa, under the auspices of USAID IPM CRSP, with South Africa and Ethiopia
the foci for biological control programmes within each region. The project encompasses host range
studies, assessments of agent impact, and transfer of technology regarding appropriate facilities and
training for weed biocontrol research. Additionally, studies of parthenium distribution, socioeconomic
and biodiversity impacts, and methods for improved pasture management are being undertaken. The
value of international cooperative research programmes is discussed.
711
XII International Symposium on Biological Control of Weeds
Biological control of Asparagus asparagoides
may favour other exotic species
P.J. Turner,1,2,3 H. Spafford Jacob1,2 and J.K. Scott2,3
School of Animal Biology (M085), University of Western Australia, Stirling Hwy, Crawley,
WA 6009, Australia
2
Cooperative Research Centre for Australian Weed Management, University of Adelaide,
Waite Main Building, Urrbrae, Adelaide, Australia
3
CSIRO Entomology, Private Bag 5, PO Wembley, WA 6913, Australia
Environmental weeds can have an impact on native biodiversity and nutrient cycling. The aim of
biological control of environmental weeds is to reduce these impacts and restore ecosystem health. A
biological control programme needs to not only evaluate agent establishment and subsequent decrease
in the target weed but also determine if the impacts of the weed have been reduced. This is a difficult
task because few studies on weed impacts are undertaken before biological control commences. This
is not the case for the weed Asparagus asparagoides (bridal creeper), which is targeted for biocontrol
within Australia. A. asparagoides invasion has resulted in a decrease in number and cover of native
plants. Areas of low species richness can be susceptible to weed invasions and therefore any reduction
in A. asparagoides may result in the expansion of other weeds. Invaded areas also contained elevated
soil nutrients, which in Australia favours exotic plants over natives. It has been suggested that restoration efforts should be dealt with post-control and is therefore a separate issue to biocontrol. However,
our study clearly demonstrates that biocontrol must be coupled with other restoration techniques.
The past, present, and future of biologically based
weed management on rangeland watersheds in the
western United States
L. Williams,1 R.I. Carruthers,2 K.A. Snyder1 and W.S. Longland
2
1
USDA ARS Exotic and Invasive Weeds Research Unit, 920 Valley Road, Reno, NV 89512, USA
USDA ARS Exotic and Invasive Weeds Research Unit, 800 Buchanan Street, Albany, CA 94710, USA
Saltcedars (Tamarix spp.) are exotic, invasive perennials introduced into the western USA from Eurasia, and are among the most damaging weeds in western riparian habitats. Here we provide an update
of Tamarix spp. control by Diorhabda elongata in the Intermountain West, and describe future research
plans for the USDA-ARS Exotic and Invasive Weeds Research Unit in Reno, NV, USA. Our goal is to
develop ecologically sustainable means of suppressing saltcedars and other exotic, invasive weeds of
the region. We have adopted a weed management pipeline approach that integrates classical biological control with ecological studies, aimed at maximizing the beneficial effects of biological control
agents while minimizing their potential detrimental effects on the soil and native flora and fauna.
This work includes use of hyperspectral imaging and other tools to characterize the spatio-temporal
dispersal and impact of biological control agents on a region-wide, long-term scale. Other studies will
address ecological interactions between biological control agents and their natural enemies, and the
effect of plantinsect interactions on plant ecophysiology and hydrology. Successful control of a target
weed usually requires decades of research; the proposed studies will be an important step toward ecologically rational management of some of the most important weeds in the western USA.
712
Abstracts: Theme 9 Management Specifics, Integration, Restoration and Implementation
An adaptive management model for the biological
control of water hyacinth
J.R.U. Wilson,1 I. Kotz,2 M.P. Hill,3 R. Brudvig,4 A. King4 and M. Byrne4
Centre for Invasion Biology, Department of Botany and Zoology, University of Stellenbosch,
Stellenbosch 7602, South Africa
2
CSIR, Stellenbosch, South Africa
3
Department of Entomology, Rhodes University, Grahamstown, South Africa
4
School of Animal, Plant and Environmental Sciences, University of the Witwatersrand,
Johannesburg, South Africa
As many sites affected by water hyacinth are inaccessible or remote, managers require a weed management guide that does not necessitate detailed and expensive monitoring. Here we focus on two sitespecific questions: What will the water hyacinth population look like if no control measures are used?
What type of control will classical biological control using Neochetina spp. provide? We developed
management recommendations for a given set of abiotic conditions based on an extensive review of
experimental and field observations. In tropical situations, given time, water hyacinth should be well
controlled by water hyacinth weevils, but in subtropical and temperate regions, there are many conditions under which water hyacinth still prospers, but weevils provide little control. The recommendations produced were found to agree with the preliminary results of a long-term multi-site monitoring
project in South Africa. In addition to providing a readily accessible guide for managers, the model
presented here also creates a focus for the development of new control methods and an assessment of
integrated control options.
Monitoring garlic mustard populations in anticipation
of future biocontrol release
L.C. Van Riper,1 L.C. Skinner2 and B. Blossey3
University of MinnesotaTwin Cities, 411 Borlaug Hall, 1991 Upper Buford Circle, St. Paul,
MN 55108, USA
2
Minnesota Department of Natural Resources, 500 Lafayette Road, St. Paul, MN 55155, USA
3
Cornell University, 122E Fernow Hall, Ithaca, NY 14853, USA
Garlic mustard (Alliaria petiolata) is native to Europe, but has become invasive in forested regions
throughout the United States. Garlic mustard is a concern because of its ability to invade high-quality
forests, form dense populations, and decrease abundance of native species. The evaluation of potential
biocontrol agents may result in the availability of Ceutorhynchus weevils for biocontrol. Accurate and
well-designed monitoring is essential to provide data as to the success of the biocontrol agents and the
status of the ecosystem. Monitoring data can be used to determine if the target species has been reduced
and if the native species are returning. Garlic mustard is a biennial and its populations can vary from
year to year. Early monitoring is necessary to accurately characterize the population before biocontrol
release. Two years of garlic mustard monitoring data from 12 sites has provided information about
garlic mustard population dynamics, a characterization of the plant communities associated with garlic
mustard, and a documentation of the low levels of herbivory currently found on garlic mustard in Minnesota (USA). Pre-release monitoring is an important component of biocontrol release.
713
This page intentionally left blank
Workshop Reports
This page intentionally left blank
Feasibility of biological control of common
ragweed (Ambrosia artemisiifolia) a noxious
and highly allergenic weed in Europe
Conveyors: D. Coutinot,1 U. Starfinger,2 R. McFadyen,3 M.G. Volkovitsh,4
L. Kiss,5 M. Cristofaro6 and P. Ehret7
Introduction
Established in Europe for more than a century, the common ragweed, Ambrosia artemisiifolia L., is the most
polluting invasive weed causing allergies to European
populations in various countries and is causing enormous health costs. At present, there is no effort from
European decision makers to implement biological control at the European level against common ragweed.
Presentation of results of the international
meeting of experts, Vienna (AGES),
27 September 2006 (Uwe Starfinger)
The experts considered biological control as an important tool in the strategy of managing A. artemisiifolia:
On the invitation of BBA, Braunschweig, and
AGES, Vienna, experts from the fields of agronomy,
botany, ecology, plant protection and road maintenance
from seven European countries gathered for a one-day
workshop to discuss the problems caused by Ambrosia
artemisiifolia and the availability and effectiveness of
control measures. Results and individual contributions
are published at the BBA website (www.bba.bund.de/
ambrosia). In particular, the experts
reported impacts of A. artemisiifolia in several European countries on human health, plant health and
nature conservation,
expressed their concern about an ongoing spread of
the species in Europe,
urged authorities in countries concerned to prevent
further import and spread or to control existing populations,
gave a set of recommendations for all private or
public bodies concerned.
European Biological Control Laboratory, USDAARS, Campus International de Baillarguet, CS 90013 Montferrier-sur-Lez, 34988 St. Gely
du Fesc, France <[email protected]>.
2
Biologische Bundesanstalt fr Land und Forstwirtschaft, Abteilung fr
nationale und internationale Angelegenheiten der Pflanzengesundheit,
Messeweg 11/12, 38104 Braunschweig, Germany <u.starfinger@bba.
de>.
3
CRC for Australian Weed Management, Block B, 80 Meiers Road, Indooroopilly, QLD 4068, Australia <[email protected]>
4
Laboratory of Insect Systematics, Zoological Institute of the Russian
Academy of Sciences, Universitetskaya Embankment 1, St.Petersburg
199034, Russia <[email protected]>.
5
Plant Protection Institute, Hungarian Academy of Sciences, H-1525
Budapest, PO Box 102, Hungary <[email protected]>.
6
ENEA-BIOTEC/BBCA, Lgo. Santo Stefano, 3 Anguillara Sabazia,
Rome 00061, Italy <[email protected]>.
7
DGAL/SDQPV, Ministre de lAgriculture et de la Pche, DRAF/
SRPV ZAC dAlco, BP 3056 34034 Montpellier, France <pierre.
[email protected]>.
A. artemisiifolia is a suitable target for biological
control in natural, ruderal and settlement areas, as
well as along traffic ways.
From a risk assessment point of view, the sub-tribe
Ambrosiineae does not include crop species in
Europe.
The risk to A. maritima, the only native species of
the sub-tribe in Europe, has to be assessed.
Success of biological control of a closely related
species (Parthenium hysterophorus) and A. artemisiifolia in Australia is a good argument to establish
biological control in Europe.
Additional suitable biological control agents have
been identified in the native area of A. artemisiifolia
and have to be investigated further.
A follow-up meeting is planned for 2008 at the next
NEOBIOTA conference in Prague.
Success of biological control against
common ragweed in Australia
(Rachel McFadyen)
Artemisia artemisiifolia is classified as a declared
class 2 weed in two eastern states of Australia: can potentially cause substantial economic and environmental
damage and land managers must take reasonable steps to
keep their land free of these weeds. Biological controls
717
XII International Symposium on Biological Control of Weeds
have been implemented as a management option for
ragweed in Australia. Several insect species have been
introduced for biological control of common ragweed,
two of which are established and giving good control.
The leaf-feeding beetle Zygogramma bicolorata Pallister (Coleoptera: Chrysomelidae) was introduced from
Mexico in 1980 as a biocontrol agent for the weed
Parthenium hysterophorus L. (Asteraceae). Defoliation
has immediate negative effects on plant performance
and as a consequence affects the growth, reproduction
and fitness of the plants. In some locations and years,
Z. bicolorata caused 85100% defoliation, resulting
in significant reductions in plant density, growth and
flower production. East coast areas receiving greater
and more frequent rainfall had higher weed density
and consequently higher beetle density. The stem-
galling moth Epiblema strenuana (Walker) (Lepidoptera: Tortricidae) was introduced from Mexico in 1982
for biological control of P. hysterophorus and became
widely established within 2 years of introduction.
E. strenuana now occurs throughout the range of its
host plants parthenium, Noogoora burr (Xanthium
occidentale Bertol.) and annual ragweed (A. artemisiifolia L.) in Australia. Galls can kill seedling plants and
severely reduce growth of larger plants, and as a result
annual ragweed is no longer a serious weed in eastern
Australia.
Biological control of common ragweed,
Ambrosia artemisiifolia L. in Russia
(Mark Volkovitsh)
The ragweed leaf beetle, Zygogramma suturalis F.
was introduced to Russia from the USA and Canada in
1978 to control the common ragweed, Ambrosia artemisiifolia. The initial phase of this introduction was a
population explosion with more than a 30-fold yearly
increase in number and population density (up to 5000
adults/m2 in aggregations).
In 20052006, we conducted selective quantitative
sampling over the whole area infested by A. artemisiifolia in southern Russia. The average population
density of the ragweed leaf beetle was very low: 0.001
adults/m2 in crop rotations and 0.1 adults/m2 in more
stable habitats. In an overwhelming majority of inspected populations, the impact on the targeted weed
was negligible. However, having regard to the spectacular success achieved in the permanent experimental
plot in 19831985, it is still possible that stable protected field nurseries could be a promising method of Z.
suturalis propagation for biological control of ragweed
in surrounding areas.
List of the known insect biological control
agents (Dominique Coutinot)
Many potential insect biological control agents exert a pressure on Ambrosia in its native environment
in North America. Certain of these have already been
introduced and established within the framework of a
classical biological control against Ambrosia in various countries but never in Europe. The list of all released biological control agents was presented during
the workshop; the list below is those who have been
reported as established:
Epiblema strenuana (Walker) (Lepidoptera: Tortricidae)
Introduced from Mexico in 1982 in Australia where
it is widely established.
Introduced from Mexico in 1991 in China: under
evaluation.
Zygogramma bicolorata Pallister (Coleoptera: Chrysomelidae)
Introduced from Mexico in 1980 in Australia where
it is well established.
Zygogramma disrupta Rogers (Coleoptera: Chrysomelidae)
Introduced from the US in 1990 in Russia; the species is under evaluation.
Zygogramma suturalis (Fabricius) (Coleoptera: Chrysomelidae)
Introduced from Canada & US in 1978 in the Republic of Georgia and released in Abhazia (West Georgia)
and Lagodekhi region(East Georgia): the establishment needs to be confirmed and evaluated.
Introduced from Canada & US in 1978 in Ukraine:
establishment is not confirmed. Introduced from Canada
& US in 1978 in Russia, where the species is established. Introduced from US in 1985 in the ex Yugoslavia,
where establishment needs to be evaluated.
Introduced from Canada in 1988 in China, where it
was recovered in some provinces. Introduced from US
in 1980 in Australia, but not established.
The use of insect biological control agents within
the framework of a classical biological control program
against Ambrosia does not seem to be under consideration to date by decision-makers in Europe.
List of potential fungal biological
control agents for Ambrosia artemisiifolia
& Species under consideration in Hungary
(Levente Kiss)
There are several fungal pathogens of common ragweed (A. artemisiifolia) that have already been considered as potential biological control agents of this
noxious weed. Protomyces gravidus, causing an endemic stem gall disease on giant ragweed (A. trifida) in
the USA, was evaluated as a potential mycoherbicide
against both giant and common ragweed. A Phoma sp.
718
Feasibility of biological control of common ragweed
isolated from A. artemisiifolia in Canada performed
well in inundative experiments alone and especially in
combination with a leaf-eating beetle, Ophraella communa. Puccinia xanthii, a microcyclic autoecious rust
reported to infect common ragweed in some parts of
the USA, was proposed as a classical biological control
agent outside North America. To our knowledge, none
of these, or any other fungal pathogens, are currently
investigated as potential biological control agents of
A. artemisiifolia except a strain of Sclerotinia sclerotiorum, which is under consideration as a mycoherbicide
in Hungary.
Biocontrol of Ambrosia artemisiifolia in
Hungary: ideas, options and problems
Currently, Hungary is the country where the air is
most polluted with ragweed pollen in Europe. Agricultural and other control measures have not helped to
reduce the amount of this airborne allergen so far. Classical biocontrol of ragweed could be a feasible way to
achieve this goal because most of the natural enemies
of A. artemisiifolia, known to occur in its native areas
in North America, are missing from Europe. Puccinia
xanthii, a microcyclic autoecious rust reported to infect
common ragweed in some parts of the USA, as well
as two leaf-eating beetles, Zygogramma suturalis and
Ophraella communa, have already been proposed as
classical biological control agents of ragweed in Hungary. In addition, a mycoherbicide product based on a
strain of Sclerotinia sclerotiorum is under development
in Hungary.
Species under consideration in Italy
(Massimo Christofaro)
Ragweed, Ambrosia artemisifolia, an alien weed that
originated from America, is now infesting the Po Valley regions in Italy. Due to the regional distribution of
the plant, Italy lacks national legislation for the prevention and control of A. artemisiifolia. In northern Italy,
regional governments as well as single municipalities
have issued local laws and guidelines that establish:
mandatory mowing in late June, late July and mid
August
general guidelines for early warning in urban areas,
improvement of the pollen monitoring station system.
In the Lombardy region alone, the direct costs of A.
artemisiifolia allergies on the public health system exceed 1 million euros (2003).
Mainly because of its alien origin and the territory
occupied, biological control can be considered a feasible approach. The following specific steps would be
required:
evaluation of of A. artemisiifolia distribution in Italy,
genetic characterization of weed and insect populations in Italy and North America,
host specificity studies of the potential selected candidate arthropods,
selection of one or more (specific) plant-pathogens
(fungi),
insectpathogen interaction studies,
evaluation of the impact of potential agents on the
weed in a confined field environment,
evaluation of potential synergisms with herbicides,
development of an information and educational
network.
In conclusion, real eradication (or at least suppression)
of the target weed can be achieved only if Italy becomes
involved together with other European Countries in an
integrated network, using biological control in combination with other different agronomic and mechanical
strategies.
Phytosanitary status of
Ambrosia artemisiifolia
a few examples (Pierre Ehret)
Experts participating on the workshop to discuss the
problems caused by Ambrosia artemisiifolia in Vienna
in 2006 recommended that countries should explore
legal possibilities to regulate Ambrosia.
It is clear that, because of its extremely large geographic distribution, this weed does not fit with quarantine regulations in most parts of the world. Different
types of official control and different levels of decision (national, regional or local) do exist nevertheless.
Some are in the phytosanitary field; more are related to
health or more global environment fields.
Local or regional approach is well adapted to an already widespread weed, and biological control seems
to warrant further exploration as part of the portfolio of
control methods.
Discussion
Biological control could be one of the tools to control
common ragweed in Europe. Taking advantage of the
examples from Australia and Russia with insect biological control agents, further investigation and research
on rust and other weed pathogens needs to be developed at a European level. Although the total eradication of this plant is not feasible, a European program
for the biological control of common ragweed would
be highly desirable.
This workshop had as objective to reply to the conclusions of the experts at the workshop organized in
Vienna by OLD (AT) and BBA (OF) at the NEOBIOTA
conference. The presentations and conclusions provide
a basis for forthcoming discussions in future scientific
or political conferences on the management of common ragweed including the feasibility of a biological
control program against Ambrosia in Europe.
719
Rearing Insects
Conveyors: R. De Clerck-Floate1 and H.L. Hinz2
Report
This informal workshop explored some of the challenges experienced in rearing insects for classical weed
biological control. In a recent global survey of weed
biological control, researchers revealed that problems
in rearing were providing significant obstacles during
biological control programs, particularly during the
early stages of development (e.g. host-specificity testing). Through a mutual sharing of some of the difficulties encountered, and also of some of the novel rearing
solutions that have been developed, it is hoped that ongoing and future programs can be improved, and perhaps difficult agents with biological control potential
may get a second chance.
During the workshop, it was suggested that it would
be useful to have a forum where researchers could
exchange information on insect-rearing techniques,
problems and solutions. Alec McClay volunteered to
set up a forum page at http://mcclay-ecoscience.com/
phpBB2/viewforum.php?f=2.
Anyone can read the forum, but to post or reply to
messages you will have to register. To do this, click on
the Register link at the top right of the page and provide the information requested. To reduce spam, Alec
has set the site up so that all registrations must be manually approved by the moderator (Alec). Please provide
information in the location, occupation and interests section so that he can identify you as someone
involved with weed biological control or with an associated field of work. He would prefer you to use your
real name as your username or signature and will try to
approve registrations soon after he receives them.
Please let Alec ([email protected])
know if you have any problems or any suggestions
to make the forum more useful, and feel free to forward this information to other weed biological control
researchers.
Agriculture and Agri-Food Canada, Lethbridge Research Centre, PO
Box 3000, Lethbridge, Alberta, Canada T1J 4B1 <[email protected]>.
2
CABI EuropeSwitzerland Centre, Rue des Grillons 1, 2800 Delmont,
Switzerland <[email protected]>.
720
Correction to a paper published in the
Proceedings of the XI International
Symposium on Biological Control of Weeds,
Canberra Australia, page 121: Population
structure, ploidy levels and allelopathy of
Centaurea maculosa (spotted knapweed)
and C. diffusa (diffuse knapweed) in
North America and Eurasia
Ruth A. Hufbauer,1 Robin A. Marrs,1 Aaron K. Jackson,2 Ren Sforza,3
Jorge M. Vivanco4 and Shanna E. Carney5
Correction
In Hufbauer et al. (2004), the root squash technique
used to gather the data for Figure 2 did not provide
adequate resolution. Other collections and approaches
suggest that C. diffusa (diffuse knapweed) within North
America are diploid (Ochsmann, 2000; Blair, personal
communication) and that C. maculosa (spotted knapweed) within North America are tetraploid (e.g. C.
stoebe micranthos) (Mller, 1989; Ochsmann, 2000;
Mller-Schrer, personal communication). Additionally, the approach used to assay catechin levels produced by individual plants for Figure 3, which followed
Bais et al. (2002), is not possible biochemically (Blair
et al., 2005). Research by Blair et al. (2005) shows that
spotted knapweed produces an average of 0.29 g/ml
(+/)-catechin (Blair et al. 2005) or 0.15g/ml ()-
catechin when grown following the protocol used here.
Department of Bioagricultural Science and Pest Management, Colorado State University, Fort Collins, CO 80523-1177, USA <hufbauer@
lamar.colostate.edu, [email protected]>.
2
Dale Bumers National Rice Research Center, USDAARS, 2890 Highway 130 East, Stuttgart, AR 72160, USA <[email protected].
gov>.
3
European Biological Control Laboratory, USDAARS, Campus Intl
de Baillarguet, CS90013 Montferrier-sur-Lez, 34988 St-Gely du Fesc,
France <[email protected]>.
4
Department of Horticulture and Landscape Architecture, Colorado
State University, Fort Collins, CO 80523-1173, USA <j.vivanco@
colostate.edu>.
5
USDAARS, 2150 Centre Avenue, Fort Collins, CO 80526, USA
<[email protected]>.
Figure 3 also contains two typos: the units for the yaxis should be microgram (g) rather than gram (g) and
the extraction was from 500 l rather than 500 ml. No
data are currently available on catechin production by
diffuse knapweed or hybrids between diffuse and spotted knapweed.
References
Bais, H.P., Walker, T.S., Stermitz, F.R., Hufbauer, R.A. and
Vivanco, J.M. (2002) Enantiomeric-dependent phytotoxic
and antimicrobial activity of (+/-)-catechin. A rhizosecreted
racemic mixture from spotted knapweed. Plant Physiology 128, 11731179.
Blair, A.C., Hanson, B.D., Brunk, G.R., Marrs, R.A., Westra,
P., Nissen, S.J. and Hufbauer, R.A. (2005) New techniques and findings in the study of a candidate allelochemical implicated in invasion success. Ecology Letters
8, 10391047
Hufbauer, R.A., Marrs, R.A., Jackson, A.K., Sforza, R., Bais,
H.P., Vivanco, J.M. and Carney, S.E. (2004) Population
structure, ploidy levels and allelopathy of spotted and diffuse knapweed in North America and Eurasia. In Cullen,
J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin,
L. and Scott J.K. (eds) Proceedings of the XI International
Symposium on Biological Control of Weeds. CSIRO Entomology, Canberra, Australia, pp. 121126.
Mller, H. (1989) Growth pattern of diploid and tetraploid
spotted knapweed, Centaurea maculosa Lam. (Compositae), and effect of the root-minimg moth Agapeta zoegana
(L.) (Lep.: Cochylidae). Weed Research 29, 103111.
Ochsmann, J. (2000) Morphologische und molekularsystematische Untersuchungen an der Centaurea stoebe L.Gruppe (Asteraceae-Cardueae) in Europa. Dissertations
Botanicae, 324, 242.
721
This page intentionally left blank
Author Index
Adair, R.J.
Alexandrova, .V.
Alfaro-Alpizar, M.A.
Altpeter, F.
Amer, W.M.
Anderson, C.L.
Anderson, F.E.
Anderson, L.W.J.
Andreas, J.E.
Andres, L.A.
Antonini, G.
Asadi, G.A.
Ash, G.J.
Audisio, P.
Averill, K.M.
Aveyard, R.
Badenes-Perez, F.R.
Baker, J.L.
Balciunas, J.
Bamba, J.
Baret, S.
Barreto, R.W.
Barton, J.
Bassett, I.E.
Batchelor, K.L.
Bean, D.W.
Becker, R.L.
Beed, F.
Beever, R.B.
Beggs, J.
Berner, D.K.
Biondi, M.
Blanfort, V.
Bloem, K.A.
Bloem, S.
Blossey, B.
Bollig, C.
Bon, M.C.
Bourchier, R.S.
Bourdt, G.W.
Bowman Gillianne, H.
Boyetchko, S.M.
Brndle, F.
Brandsaeter, L.O.
Bredow, E.
Briano, J.A.
122
251
129
354
246
443
245
353
75, 516
521
263
245
254, 306
263
251
637
129
503, 631
395
641
476
109, 182, 195, 206, 221,
248, 270, 358, 359, 577, 693
245, 364, 631
56
637
257, 704
358, 395
354
449
56
251, 396, 704
263
476
708
102, 708
56, 103, 356, 635, 713
450
263, 448
632
87,145, 246, 507
450
353
634
470
589
396
Briese, D.T.
Brown, B.
Bruck, D.J.
Bruckart, W.L.
Brudvig, R.
Bruzzese, E.
Buccellato, L.
Buckley, Y.M.
Burns, E.
Byrne, M.J.
Cabrera Walsh, G.
Cadisch, G.
Caesar, A.J.
Caesar-Ton That, T.
Cag, L.
Campobasso, G.
Campos, M.R.
Cardo, M.V.
Carney, V.A.
Carpenter, J.E.
Carruthers, R.I.
Carvalheiro, L.G.
Casagrande, R.
Castillo-Castillo, A.
Causton, C.E.
Center, T.D.
Chandler, M.
Charudattan, R.
Chkhubianishvili, C.
Chong, J.
Chumala, P.
Cliquet, S.
Cock, M.J.W.
Coetzee, J.A.
Cofrancesco, A.F., Jr
Colasanti, R.
Collison, L.M.
Colonnelli, E.
Cook, C.
Coombs, E.M.
Cortat, G.
Cother, E.J.
Coutinot, D.
Cozma, V.
Cripps, M.G.
Cristofaro, M.
723
91
398
37
396, 540, 704
632, 713
260
57
3, 52, 83, 101
589
57, 59, 512, 558, 632,
636, 705, 708, 713
252, 353
634, 705
7, 13
13
216
397
182, 359
211, 435
633
101, 102, 708
535, 712
83, 101
259
129
669
63, 641, 642, 655
252
102, 354
354
589
636
254, 306
451
512, 665
397
52, 64
640
173, 246
528
395, 516, 521, 680
133
254, 306
717
247
246
133, 150, 154, 173, 178, 189,
XII International Symposium on Biological Control of Weeds
Cruz, T.Z.
Cuda, J.P.
Cullen, J.M.
Curtet, L.
227, 246, 263, 321, 333,
361, 490, 717
641
60, 102, 270, 355, 589
91
490
Dalin, P.
Darbyshire, S.
Dvalos, A.
Davies, G.
Davis, A.S.
Day, R.
De Biase, A.
De Clerck-Floate, R.
de Lillo, E.
Delgado, O.
Delhey, R.
DeLoach, C.J.
deMeij, A.E.
Den Breeyen, A.
Dev, U.
Dhileepan, K.
Di Cristina, F.
Diaconu, A.
Diaz, R.
Dick, G.O.
Ding, H.
Ding, J.
Diplock, N.
DiTomaso, J.M.
DiTommaso, A.
Djeddour, D.H.
Doetzer, K.
Dolgovskaya, M.Yu
Duckett, C.
Dudley, T.L.
Dunlap, C.A.
704
448
56
470
635
360
263
528, 633, 720
178
256
211, 258
63, 249, 253, 398,
450, 535, 704
573
354
384
105
173, 189, 246, 333, 361
247, 278
355
707
75
103
247
649
251
463
254
227, 614
256
138, 355, 704
634
Eberts, D.
Edwards, G.R.
Edwards, P.J.
Ehlers, R.-U.
Ehret, P.
Eigenbrode, S.D.
Elliott, M.
Ellison, C.A.
Elzein, A.
Erasmus, B.F.N.
Eschen, R.
Evans, H.C.
Evans, K.J.
Everitt, J.
535, 704
246
360
369
717
75
102
165, 361, 384, 699, 709
634, 705
705
470
20, 455, 463, 491
637
249, 535
Fan, L.L.
Faria, A.B.V.
Fichera, G.W.
349
270
398
Fisher, A.J.
Fisher, J.T.
Fourie, A.
Fowler, S.V.
Foxcroft, L.C.
Freedman, J.E.
Freitas, H.
French, K.
Friesen, R.D.
Fumanal, B.
Galea, V.J.
Gard, B.
Gaskalla, R.
Gaskin, J.
Gassmann, A.
Geissler, J.
Gerber, E.
Getz, C.
Ghorbani, R.
Gianotti, A.F.
Goolsby, J.
Gourlay, A.H.
Gous, S.F.
Grecu, M.
Greizerstein, E.
Gresham, B.
Grevstad, F.S.
Grodowitz, M.J.
Groenteman, R.
Grosskopf, G.
Groves, R.H.
Gltekin, L.
Hfliger, P.
Haines, M.L.
Hanks, M.
Hansen, R.W.
Harizanova, V.
Harman, H.M.
Harmon, B.L.
Harms, N.
Hartley, D.
Hatcher, P.E.
Hayat, R.
Hayes, L.M.
Heard, T.
Hedderson, T.A.J.
Heil, J.
Hennecke, B.R.
Herr, J.C.
Herrick, N.J.
Hibbard, K.
Hiebert, E.
Hight, S.D.
Hill, M.P.
724
540
705
356, 710
87, 104, 145, 246, 248,
254, 495, 545, 631
452
44
359
103
669
448
247, 676
490
589
410, 448, 640
248, 252, 259
353
57, 133, 154, 635
390
245
364
249
104, 364, 545, 680
32
278
435
287
58
44, 61, 62, 639,
706, 707, 711
87, 145
252, 552, 644
91
133, 150, 154, 173
356
104
395
635
311, 328
248, 449, 495
429, 640
639
637, 710
357, 470, 707
189, 321, 361
376
528
452
395
103
535
292, 362
589
102
589, 708
59, 512, 558, 632, 708, 713
Author index
Hill, R.L.
Hinz, H.L.
Hoffmann, J.H.
Hona, S.R.
Horn, C.
Horrell, J.
Hough-Goldstein, J.A.
Hufbauer, R.A.
Hupka, D.
Hurard, C.
Hurrell, G.A.
Hynes, R.K.
Impson, F.A.C.
Ireson, J.
287, 680
133, 154, 278, 410,
429, 470, 528, 635
26, 359, 452, 687, 720
104
376
102
283
418, 449
353, 636
448
507
353, 636
26
680
Jackson, C.A.R.
Jackson, M.
Jadhav, A.M.
Jashenko, R.V.
Jeffers, L.
Johnson, M.T.
Joley, D.B.
Jones, W.
Jourdan, M.
Julien, M.H.
Jurovich, D.
58, 61, 561
634
708
249, 253, 398
706
129, 709
607
249, 448, 490
160, 637
91, 211, 255, 258, 435, 528
633
Kalibbala, F.N.
Karacetin, E.
Karimpour, Y.
Karova, A.
Kashefi, J.
Katovich, E.J.S.
Katz, M.L.
Kay, M.K.
Kazmer, D.
Keller, J.C.
Kelly, D.
Kharrat, M.
Kiehr, M.
Killgore, E.M.
King, A.M.
Kirk, A.
Kirton, A.
Kiss, L.
Kitchen, H.
Kleinjan, C.
Knutson, A.E.
Kohlschmid, E.
Kok, L.T.
Kolomiets, T.M.
Konstantinov, A.
Korotyaev, B.A.
Kotz, I.
Kovalenko, E.D.
Krebs, C.
636
37
250, 357
301
60, 568, 637
358, 395
58
59, 250, 287
704
257
87, 145, 248
238
211, 258
364, 594
59, 713
138, 249, 355
708
717
449
26
535
489
292, 362
251
263
133, 154, 173, 263
713
251
57
Kremer, R.J.
Kriticos, D.J.
Kroschel, J.
Kumar, P.S.
Kurose, D.
Kwong, R.
Laing, M.D.
Lambert, A.M.
Lamoureaux, S.
Landis, D.A.
Laukkenen, K.H.
Lavergne, C.
Lawrie, A.C.
Le Bourgeois, T.
Le Maguet, J.
Lecce, F.
Lecheva, I.
Lekomtseva, S.N.
Lenz, L.
Littlefield, J.L.
Liu, W.-X.
Livramento, M.
Longland, W.S.
Lonsdale, W.M.
Lowe, A.J.
Lscher, A.
Luster, D.G.
Lustosa, D.C.
Lyver, P.O.B.
MacDonald, G.E.
Macedo, D.M.
MacKinnon, D.K.
Madeira, P.T.
Makinson, J.
Malania, I.
Mancini, E.
Manhart, J.
Manrique, V.
Marchante, H.
Markin, G.P.
Marley, P.
Marrs, R.A.
Martin, J.F.
Massey, B.
Mattioli, F.
Maxwell, A.
Maywald, G.
McAvoy, T.J.
McClay, A.S.
McConnachie, A.J.
McEvoy, P.B.
McFadyen, R.
McKay, F.
McLaren, D.A.
725
7
32, 43
634, 705
384
463
680
365
138, 355
248
635
58
476
255
476
490
173, 189, 333, 361
301
251
706
60, 200, 552, 568, 573, 583, 637
699, 709
254
712
91
452
470
396
358
376
354
577
418
665
398
354
263
249
60, 102
359
60, 200, 227, 301, 516,
568, 573, 583, 620,
637, 643, 669, 680
634, 705
449
448
449
353
160
64
292, 362
252
253, 558, 710, 711
37
67, 717
252, 253
245, 255
XII International Symposium on Biological Control of Weeds
McNeil, J.N.
Medal, J.C.
Memmott, J.
Meyer, J.Y.
Michels, G.J, Jr
Michels, J.
Milan, J.D.
Milbrath, L.R.
Miller, J.C.
Mira, D.
Mityaev, I.D.
Mizubuti, E.S.G.
Moeri, O.E.
Monfreda, R.
Moore, G.
Moore, J.A.
Morais, E.G.F.
Moran, P.J.
Moran, V.C.
Moretti, M.
Morin, L.
Moulay, C.
Muegge, M.
Mueller-Schaerer, H.
Mukhina, Zh..
Mller-Schrer, H.
Mller-Stver, D.
Muniappan, R.
Murrell, C.
Myers, J.H.
Myers, J.V.
Nachtrieb, J.G.
Nagasawa, L.S.
Nastasa, V.
Navajas, M.
Neave, M.
Nechet, K.L.
Negeri, M.
Newcombe, G.
Newman, J.R.
Nibling, F.
Norambuena, H.
Norton, A.P.
Novak, S.J.
Nowierski, R.M.
Ntushelo, K.
Nuzzo, V.
OCasey, C.E.
Officer, D.
Ogutu, W.O.
Olckers, T.
Oleiro, M.I.
OMeara, S.
Osborne, L.
Overholt, W.A.
Owens, C.S.
390
102, 589
83, 101, 495
476, 594
633
535
640
251, 448
521
398
249, 253
693
102
178
249
26
182, 359
535
26, 687
57
637, 638, 639, 642, 710
254, 306
535
360
251
450
489
638, 641
57
58, 61, 561, 601
104
61, 639
364
278
403
639
221, 693
711
640
489
249
640, 680
62
422
620, 643
360, 710
635
58
255
360
710
252, 253
535
589
60, 102, 355, 589
61, 62
Palmer, W.A.
Palomares-Rius, F.J.
Pankratova, L.F.
Paolini, A.
Parepa, M.
Parker, P.
Parkes, S.
Parsons, L.K.
Paynter, Q.
Pedrosa-Macedo, J.H.
Pelser, P.
Peng, G.
Pepper, A.
Pereira, O.L.
Petersen, B.A.
Peterson, P.G.
Peterson, R.K.
Picano, M.C.
Pitcairn, M.J.
Pomella, A.W.V.
Potter, K.J.
Pratt, P.D.
Prody, H.D.
Puliafico, K.P.
Purcell, M.F.
Puzari, K.C.
Queiroz, R.B.
Quinn, H.
Rabindra, R.J.
Raghu, S.
Ragsdale, D.W.
Ramasamy, S.
Rambuda, T.
Randal, C.
Rashed, M.H.
Rattray, G.
Rayamajhi, M.B.
Rector, B.G.
Reddy, G.V.P.
Reid, A.
Rentera, J.L.
Retief, E.
Reznik, S.Ya
Richardson, B.
Richardson, D.M.
Robbins, T.O.
Robertson, M.
Roda, A.
Roderick, G.
Rollins, J.A.
Sadeghi, H.
Salom, S.M.
Samokhina, I.U.
Sanabria, J.
Sankaran, K.V.
726
365
357
251
173, 189, 333, 361
247
249
449
640
56, 104
102, 248, 254
363
636
249
195
257
104, 495
105
182, 359
607
577
32
63, 355, 641, 642, 655
60
200, 429, 573
256, 398, 655
384
182, 359
61
165
105
232, 358, 395, 643
255
645
535
245
545
63, 641, 655
263, 311, 317, 328
638, 641
638, 639, 642
361
362, 710
227, 333, 614
32
452
63, 450
43, 59, 632, 708
589
403
354
245
292, 362
251
63, 450, 535
384
Author index
Sauerborn, J.
489
Saville, D.J.
507
Sawchyn, K.
353
Scanlan, J.
105
Schaffner, U.
57, 200, 252, 360, 363, 450, 470
Schooler, S.S.
255
Schwarzlnder, M. 75, 154, 410, 429, 443, 640, 644
Scott, J.K.
91, 160, 399, 451, 712
Seier, M.
154, 451, 463
Sellers, B.
589
Sforza, R.
422, 448, 449, 490
Sharp, D.
645
Shaw, R.
463, 484, 489
Shearer, J.F.
44
Sheppard, A.W.
91, 364, 451, 452, 680
Shivas, R.G.
365
Silva, A.A.
693
Silva, G.
182, 359
Silvrio, M.
254
Simelane, D.O.
253, 363, 710
Sinclair, A.R.E.
58
Sing, S.E.
105, 620, 643
Skatenok, O..
251
Skinner, L.C.
232, 248, 358, 395, 643, 713
Skoracka, A.
317
Smart, R.M.
61, 62, 707, 711
Smith, L.
150, 173, 189, 263, 321,
333, 361, 495, 540
Smith, L.A.
104, 644
Snow, J.
707
Snyder, K.A.
712
Soares, D.J.
206
Sosa, A.J.
211, 258, 435
Soubeyran, Y.
476
Souissi, T.
238
Souza, P. G.
340
Spafford Jacob, H.
260, 712
Spasskaya, I.A.
614
Spencer, D.
249
Sreenivasam, D.
395
Sreerama Kumar, P.
165
Stansly, P.
589
Starfinger, U.
717
Steinger, T.
450
Stoeva, A.
311, 328
Stokes, J.A.
706, 707
Strathie, L.W.
256, 645, 710, 711
Suzuki, L.
254
Syrett, P.
644
Szcs, M.
443
Tabatadze, E.
Takahashi, N.
Talmaciu, M.
Tamagawa, Y.
Tanner, R.A.
Taputuarai, R.
Tarin, D.
354
463
247, 278
138
463, 491
594
249
Tate, C.D.
Taylor, G.S.
Telesnicki, M.C.
Terragitti, G.
Thines, M.
Thomann, T.
Thomas, H.Q.
Thompson, D.C.
Thompson, D.L.
Tipping, P.W.
Tosevski, I.
Tth, P.
Tthova, M.
Tracy, J.L.
Traversa, M.G.
Treier, U.
Tronci, C.
Tsivilashvili, L.
Turner, C.E.
Turner, P.J.
Turner, S.C.
van Klinken, R.D.
Van Riper, L.C.
Van, T.K.
Ventim, R.
Vieira, B.S.
Villegas, B.
Vitorino, M.
Vitou, J.
Volkova, S.A.
Volkovitsh, M.G.
Vrieling, K.
Vurro, M.
Waipara, N.W.
Walsh, G.C.
Wan, F.-H.
Wapshere, A.J.
Wardill, T.
Warner, K.D.
Watt, M.S.
Watts, D.
Wearing, A.
Weaver, D.K.
Webber, N.A.P.
Weed, A.S.
Wheeler, G.S.
Whitaker, S.G.
White, S.R.
Whitehead, D.
Wikler, C.
Wilkie, P.
Williams, A.M.
Williams, D.
Williams, H.
Williams, L.
727
101
256
435
397
634
160, 364
257
257, 535
704
641, 642, 655
248
216
216
63, 450
211, 258
450
133, 150, 154, 173,
189, 227, 321, 333, 361
354
521
712
259, 625
52, 64, 247
713
63
83
221, 693
607
589
160, 451
251
227, 614, 717
363
455
195, 246, 248, 364, 449
255
699, 709
91
528
390
32
249
247
620, 643
503, 631
259
252, 256
706
58, 61
32
340, 589
449
260
60
710
154, 712
XII International Symposium on Biological Control of Weeds
Willis, A.J.
Wilson-Davey, J.R.A.
Wilson, J.R.U.
Wilson, L.M.
Wineriter, S.A.
Winks, C.J.
Winston, R.L.
Withers, T.M.
Witkowski, E.T.F.
Witt, A.B.R.
Wood, A.R.
Woods, D.M.
Wright, A.D.
Wu, Z.
Xiang, M.M.
638
104
452, 713
552
655
104, 248
644
104
57, 636
452
345, 356, 360, 362,
365, 645, 710
540, 607
398
395
349
Yacoub, R.
Yan, S.
Yang, C.
Yobo, K.S.
Yoder, M.V.
Zachariades, C.
Zaitzev, V.F.
Zeehan, K.
Zeeshan, K.
Zeng, Y.S.
Zhang, F.
Zhang, X.
Zhou, Y.P.
Zilli, A.
Zonneveld, R.
Zouaoui Boutiti, M.
728
607
292, 362
249
365
232
43, 256, 645
614
254
306
349
699, 709
287
349
246
398
238
Keyword Index
A
Acacia
Acari
adaptation
agent impact
agent selection
air-drying
Allee effects
Alliaria petiolata
alligator weed
Alternanthera philoxeroides
Aphthona nigriscutis
aquatic weeds
arthropod herbivores
arthropods
Assam
athel
Aulacidea hieracii
Aulacidea pilosellae
Aulacidea subterminalis
Australia
26
317
403
512
43, 52, 109, 122, 429
306
495
410
349, 435
349, 435
503
206
232
160, 211
384
535
552
552
552
67
B
bacteria
7
beneficial non-target effects
87
biocontrol
507
bioherbicides
7, 109, 221, 455, 577, 693
biological control
3, 7, 26, 83, 91, 138, 160, 206,
232, 516, 594, 614, 649
biological control agent
476
biological control agent habitat requirements
625
biological control efficacy
601
biological control weeds
535
biological invasions
435
biology
227
bionomics
150
biotic interference
129
biphasic
221
Bradyrrhoa gilveolella
568
Brazilian peppertree
270
bronze skeleton weed root borer
227
broomrapes
238
Buddleja
287
Buddleja davidii
32
C
Cactoblastis
candidates
687
216
carbon addition
Carduus nutans
Carduus pycnocephalus
Carduus tenuiflorus
Centaurea solstitialis
centrifugal phylogenetic method
Ceratapion basicorne
Cheilosia psilophthalma
Cheilosia urbana
chemical ecology
chlamydospores
Chondrilla juncea
Chromolaena odorata
cinnabar moth (Tyria jacobaeae)
Cirsium vulgare
classical biological control
Cleopus
Cleopus japonicus
climate
climate-matching
CLIMEX
coevolution
coevolved pathogens
COI
communication
compensatory growth
competition
competitive weed replacement
conflict of interest
congeneric
conidia
Convolvulus
cooperation
cost/benefit
crofton weed
Cryptonevra
Cuscuta
Cyperus rotundus
Cytisus scoparius
D
Dactylopius
Dalmatian toadflax
data requirements
decision making
Diorhabda
Dipsacaceae
729
278
87, 145
87
87
150, 189
410
263
552
552
75
306
568
43
37, 583
145
109, 182, 195,
270, 384, 455, 463
287
32
561
512
43
20
206
263, 443
376
32
561
699
122
561
306
216
676
67, 91
699
138
216
577
516
687
625
369
376
535
328
XII International Symposium on Biological Control of Weeds
Dipsacus
distribution
dock
E
efficacy
emerging environmental weed
environmental weeds
Eriophyidae
Eriophyoidea
establishment
establishment rates
EU
Eucryptorrhynchus brandti
Euphorbia esula
evaluation
Everglades
evolution
Exapion fuscirostre
F
Fallopia japonica
feeding damage
fermentation
field biology
field experiments
field surveys
flea beetle
FloraMap
foreign exploration
Fumaria species
fungal mutualists
fungal pathogens
fungus
G
generalist insects
genetic analysis
genetic variation
genetics of biocontrol releases
geographical distribution
giant reed
grazing
growth modelling
H
habitat restoration
Hawaii
heather beetle
herbivore
herbivore niche
herbivory
historical review
hoary cress
host-choice behaviour
host plant specificity
host range
host range testing
216
227, 301
470
620
345
455
317
178
540
495
484
292
503
44
655
403
516
463
333
306
568
321
154, 160
333
43
91, 173, 669
160
20
160
693
429
665
403
495
178
138
278
32
655
182, 340
495
138
429
129
91
278
75
321
189, 227, 340
133, 301
host specificity
host specificity testing
houndstongue
hybrid strain
hybrids
Hydrellia pakistanae
I
impact
impact assessment
impact tests
Indian infrastructure for
biological control
insect
insect biotypes
insect exclusion
integrated management
integrated weed management
international survey
invasion
invasive alien species
invasive alien weeds
invasive grass
invasive plants
invasive weed
IPM
Ipomoea carnea subsp. fistulosa
Ipomoea fistulosa
Isatis tinctoria
island
37, 129, 512, 552
292
75
443
435
44
227
333
340
165
216
443
278
649
278, 680
528
607
384
165
422
311, 328, 476, 589,
601, 649, 655
195
649, 699
206
206
133
594
J
Japanese knotweed
463
K
Kazakhstan
Kerala
154
384
L
Larinus filiformis
legislation
Lepidium draba
Lepidium latifolium
Lewia
Linaria
Longitarsus jacobaeae
M
Macrolabis pilosellae
management
Melanaphis donacis
methodology
microbial ecology
microsatellite
microspora
Mikania micrantha
mites
730
150
484
410
154
693
418, 620
545, 573
552
665
138
83
13
418
37
165
328
Keyword index
modelling
modelling tritrophic interactions
monitoring
Montana
multilocus genotypes
multiple agents
multiple introductions
multiple species
Mycoleptodiscus terrestris
mycotoxin
N
natural enemies
non-target effects
non-target feeding
O
Onopordum acanthium
Opuntia stricta
organic farming
P
parasitic plants
Parkinsonia
Parthenium hysterophorus
Passiflora tarminiana
Passiflora tripartita
pathogenhost interaction
pathogenic fungi
pathogenic mechanism
pathogenicity
pathogens
Persicaria perfoliata
pest risk assessment
Phrydiuchus tau
phylogeny
plant community response
plant disease
plant distribution
plant fitness
plant pathogens
plantherbivore interactions
plantinsect interactions
plantation crops
Polygonum perfoliatum
polyploids
population dynamics
population regulation
post-release assessment
post-release evaluation
potential spread
pre-dispersal seed predation
project challenges
Psylliodes chalcomera
public engagement
public-interest science
public outreach
purple nutsedge
3, 680
37
589, 594, 687
200
422
561
422
26
44
349
173, 232, 292, 601, 699
75, 561
503, 552
145
687
470
238
676
165
669
669
37
211
349
221, 238
238
283
484
521
287, 321
620
195
211
20
13
435
418
384
283
435
3, 178, 182, 680
52
516
512, 558, 614
665
516
528
263
376
390
390
577
Q
quackgrass
R
ragwort flea beetle
rainfall
rainforest
range improvement
Ranunculus acris
realized host range
rearing
rearing difficulties
REBECA
registration
regulation
reinfection
release size
research
Rhamnaceae
Rhinoncomimus latipes
risk
risk assessment
risk management
risk perception
Rumex crispus
Rumex longifolius
Rumex obtusifolius
rush skeletonweed
Russian thistle
rust pathogens
S
Salsola
saltcedar
Schinia cognata
Sclerotinia sclerotiorum
seed dispersal mechanisms
selection
Senecio jacobaea
sequence variation
simulated herbivory
single agent
soilborne
South Africa
South America
spatial mapping
spread
spread rates
stage-based implementation model
Stenoplemus rufinasus
storage
strawberry
success rates
successful management
surveys
synergism
731
317
200, 545, 573
545
594
521
507
512
292
528
369
369
369
540
495
676
232
283
561
321
390
390
470
470
470
301, 568
173
165
173
535
301
507
52
403
200, 583
443
32
561
13
345
669
625
540
52
607
558
306
311
67
521
227, 270
7, 13
XII International Symposium on Biological Control of Weeds
T
Tamarix
tansy ragwort
target selection
taxonomic revision
taxonomy
teasel
Tetramesa romana
Tibouchina
Tingis grisea
topdown effects
Trichosirocalus horridus
Trichosirocalus mortadelo
trophic interactions
tropical island
tumbleweed
Turkey
Tyria jacobaeae
535
573, 583
109
287
227
311, 328
138
340
189
278
145
145
7, 13
476
173
154
37, 583
U
ultrastructure
349
V
virulence
238
W
wattle
weed biological control
weed control
weed management
weeds
weevil
wheat
wild poinsettia
Y
yellow bells
YST (Centaurea solstitialis)
732
122
67, 133, 577, 589, 669
507
693
178, 182, 614
521
317
693
345
150, 189
List of Delegates
Surname
Address
Robin
Adair
Primary Industries Research Victoria, PO Box 48, Frankston, VIC, Australia
Hala
Alloub
University of Gezira, Faculty of Agricultural Sciences, Wao Medani, Sudan
Freda
Anderson
CerzosConicet, Camino la Carrindanga, Km 7, 8000 Bahia Blanca, Argentina
Jennifer
Andreas
Washington State University, Suite 120, 99 SW Grady Way, Renton, WA, USA
Gloria
Antonini
Biotechnology and Biological Control Agency, Via del Bosco 10, Rome, Italy
Francisco Ruben
Badenes-Perez
Facultad de Medicina, Universidad Costa Rica, Studios U. Frente, San Pedro de Montes Oca, Costa Rica
Janita
Baguant
Department of Primary Industries, PO Box 48, Frankston, VIC, Australia
Karen
Bailey
Agriculture & Agrifood Canada, 107 Science place, Saskatoon, Canada
John Lars
Baker
Fremont Country Weed and Pest, 450 North Second Street, Lander, WY, USA
Myriam
Barat
Universitie de Rennes 1, Campus de Beaulieu, Ave du General Leclerc, Rennes, France
Robert
Barreto
Universidade Federal de Vicosa, Vicosa, MG, Brazil
Imogen
Bassett
University of Auckland, Tamaki Campus, Building 733, Auckland, New Zealand
Nicolas
Beck
Tours du Valat, Arles, France
Roger
Becker
University of Minnesota, 411 Borlaug Hall, 1991 Upper Buford Circle, St Paul, MN, USA
Dana
Berner
USDA-ARS, 1301 Ditto Avenue, Ft. Detrich, MD, USA
Bernd
Blossey
Cornell University, Ecology & Management of Invasive Plants Program, 202 Fernow Hall, Ithaca, NY, USA
Marie-Claude
Bon
USDA-ARS, EBCL, Campus de Baillarguet, Montferrier sur Lez, France
Robert
Bourchier
Agriculture and Agri-Food Canada, Lethbridge Research Centre, 5403 1st Avenue S., Lethbridge, Alberta,
Canada
Graeme
Bourdot
AgResearch, Gerald Street, Lincoln, Canterbury, New Zealand
Meriem
Boutiti Zouaoui
Institut National Agronomique de Tunisie, 43 Avenue Charles Nicolle, Tunis-Mahragene, Tunisia
Susan
Boyetchko
Agriculture & Agri-Food Canada, 107 Science Place, Saskatoon, SK, Canada
Juan
Briano
USDA-ARS-OIRP, SABCL, Bolivar 1559, Hurlingham, Buenos Aires, Argentina
William
Bruckart
USDA-ARS, FDWSRU, 1301 Ditto Avenue, Frederick, MD, USA
Ryan
Brudvig
University of the Witwatersrand, School of animal, plant & environmental sciences, Private Bag 3, Wits,
Lisa
Buccellato
Yvonne
Buckley
University of Queensland, School of integrative biology, St Lucia, Brisbane, QLD, Australia
Marcus
Byrne
University of the Witwatersrand, School of animal, plant and environmental sciences, Wits,
Johannesburg, South Africa
University of the Witwatersrand, School of animal, plant & environmental sciences, Private Bag 3, Wits,
Johannesburg, South Africa
Johannesburg, South Africa
Guillermo
Cabrera Walsh
USDAARS, SABCL, Bolivar 1559, Hurlingham, Buenos Aires, Argentina
XII International Symposium on Biological Control of Weeds
734
Given name
Surname
Address
Anthony
Caesar
USDAARS, 1500 North Central Avenue, Sidney, MT, USA
Gaetano
Campobasso
(Deceased)
Vanessa
Carney
Texas Agricultural Experiment Station, 2301 Experiment Station road, Bushland, TX, USA
James
Carpenter
USDAARS, PO Box 748, Tofton, GA, USA
Luisa
Carvalheiro
University of Bristol, School Biological Sciences, Woodland Road, Bristol, UK
Richard
Casagrande
University of Rhode Island, 9 East Alumni Avenue, Kingston, RI, USA
Ted
Center
USDAARS Invasive Plant, 3225 SW College Avenue, Fort Lauderdale, FL, USA
Raghavan
Charudattan
University of Florida, PO Box 110680, Gainesville, FL, USA
Hongyin
Chen
Chinese Academy of Agricultural Sciences, N2, West Yuan Ming Yuan Road, Beijing, China
Cisia
Chkhubianishbili Institute of Plant Protection, 82 Chavchavadze Avenue, Tbilisi, Georgian Republic
Julie
Coetzee
ARCARS, PPRI, Private Bag 3, Wits, Johannesburg, South Africa
Alfred
Cofrancesco
US Army Engineer Research & Development, CEERD-EM-W, 3909 Halls Ferry Road, Vicksburg, MS, USA
Eric
Coombs
Oregon Department Agriculture, 635 Capitol Street, Salem, OR, USA
Ghislaine
Cortat
CABI Europe Switzerland, 1 rue des Grillons, Delmont, Switzerland
Jenny
Cory
Algoma University College, 1520 Queen Street E., Sault Ste Marie, Canada
Dominique
Coutinot
USDAARS, EBCL, Campus International de Baillarguet, Montferrier sur Lez, France
Michael
Cripps
Lincoln University, PO Box 84, Lincoln, New Zealand
Massimo
Cristofaro
ENEABBCA, Via Anguillarese, 301, S.M. di Galeria, Rome, Italy
James
Cuda
University of Florida, Bluilding 970, Natural Area Drive, Gainesville, FL, USA
Jim
Cullen
CSIRO Entomology, GPO Box 1700, Canberra, ACT, Australia
Laurence
Curtet
Office Nat. de la Chasse & de la Faune Sauvage, Station de la Dombes, Birieux, France
Joanne
Daly
CSIRO Agribusiness, GPO Box 1700, Canberra, ACT, Australia
Andrea
Davalos
Cornell University, Fernow Hall, Ithaca, NY, USA
Rosemarie
de Clerck-Floate
Lethbridge Research Center, 5403-1 Avenue South, Lethbridge, Alberta, Canada
Enrico
de Lillo
University of Bari, Biologia e Chimica Agroforestale e Ambientale, Via Amendola, 165/A, Bari, Italy
Chantal
Dechamp
AFEDA, 25 rue Ambroise, Saint-Priest, France
Ernest
Delfosse
USDAARS, 5601 Sunnyside Avenue, 4-2238, Beltsville, MD, USA
Oona
Delgado
MIZAUCV, Avenida Universidad, El Limon, Macaray, Venezuela
Jack
DeLoach
USDAARS, 808, E. Blackland Road, Temple, TX, USA
Alana
Den Breeyen
University of Florida, 1453 Fifield Hall, Gainesville, FL, USA
Alecu
Diaconu
Institute of Biological Research, Biological Control Laboratory, Boulevard Carol I, 20-A, Iasi, Romania
Delegate List
735
Given name
Surname
Address
Rodrigo
Diaz
University of Florida, 2199 S. Rock Road, Fort Pierce, FL, USA
Franca
DiCristina
BBCA, Via del Bosco, 10, Sacrofano, Rome, Italy
Jianqing
Ding
Wuhan Botanical Garden, Chinese Academy of Sciences, Moshan, Hubei Province, China
Naomi
Diplock
University of Queensland, Gatton Campus, Gatton, QLD, Australia
Joe
DiTomaso
University of California, Mail stop 4, Davis, CA, USA
Djamila
Djeddour
CABI EuropeUK, Silwood Park, Buckhurst Road, Ascot, Berkshire, UK
Sarah
Dodd
Landcare Research, Private Bag 92-170, Auckland, New Zealand
Margarita
Dolgovskaya
Zoological Institute RAS, Universitetskaya, 1, St Petersburg, Russian Federation
Alan
Dowdy
USDAARSOIRP, 5601 Sunnyside Avenue, Beltsville, MD, USA
Christopher
Dunlap
NCAURUSDA, 1815 N. University street, Peoria, IL, USA
Ralf-Udo
Ehlers
Phytopathology, Hermann-Rodewald street 9, Kiel, Germany
Pierre
Ehret
MAP/DGAL/SDQPV, DRAF/SRPV, BP 3056, Montpellier, France
Carole
Ellison
CAB International, Silwood Park, Buckhurst Road, Ascot, UK
Abuelgasim
Elzein
University of Hohenheim, Garbenstrasse 13, Stuttgart, Germany
Harry
Evans
CABI UK Center, Silwood Park, Buckhurst Road, Berkshire, UK
Simon
Fowler
Landcare Research, PO Box 40, Lincoln, New Zealand
Rex
Friesen
Southern Kansas Cotton Growers Cooperative, PO Box 972, Oxford, KS, USA
Victor
Galea
University of Queensland, Gatton Campus, Gatton, QLD, Australia
Benjamin
Gard
Montpellier Sup Agro, 2, place Pierre Viala, Montpellier, France
Andr
Gassmann
CABI Europe Switzerland, 1, rue des Grillons, Delmont, Switzerland
Esther
Gerber
CABI Europe Switzerland, 1, rue des Grillons, Delmont, Switzerland
Reza
Ghorbani
Faculty of Agriculture, Ferdowsi Universtiy of Mashhad, PO Box 91775-1163, Mashhad, Iran
Christophe
Girod
CSIRO EL, Campus International de Baillarguet, Montferrier sur Lez, France
John
Goolsby
USDAARS, 2413 East Highway 83, Weslaco, TX, USA
Pierre
Gotteland
CNPAEI, 72, rue Leon Menabrea, Chambery, France
Hugh
Gourlay
Landcare Research, PO Box 40, Lincoln, Christchurch, New Zealand
Fritzi
Grevstad
University of Washington, Olympic Natural Resources Center, 2907 Pioneer Road, Long Beach, WA, USA
Michael
Grodowitz
US Army Engineer Research & Development, 3909 Halls Ferry Road, Vicksburg, MS, USA
Ronny
Groenteman
University of Canterbury, School of Biological Sciences, Christchurch, New Zealand
Gitta
Grosskopf
CABI EuropeSwitzerland, 1, rue des Grillons, Delmont, Switzerland
Patrick
Hfliger
CABI Europe Switzerland, 1, rue des Grillons, Delmont, Switzerland
Richard
Hansen
USDA, Suite 108, 2301 Research Boulevard, Fort Collins, CO, USA
XII International Symposium on Biological Control of Weeds
736
Given name
Surname
Address
Vili
Harizanova
Agricultural University, 12 Mendeleev Street, Plovdiv, Bulgaria
Helen
Harman
Landcare Research, Private Bag 92-170, Auckland, New Zealand
Paul
Hatcher
University of Reading, Whiteknights, PO Box 217, Reading, UK
Rustem
Hayat
Ataturk University, Faculty of Agriculture, Erzurum, Turkey
Lynley
Hayes
Landcare Research, PO Box 40, Lincoln, Christchurch, New Zealand
Tim
Heard
CSIRO, 120 Meiers Road, Indooroopilly, Brisbane, QLD, Australia
Bertie
Hennecke
University of Wollongong, School of Biological Sciences, Wollongong, NSW, Australia
Joseph
Hershenhorn
Newe Yaar Research Center, PO Box 2021, Ramat Yishay, Israel
Stephen
Hight
USDAARSCMAVE & FAMU Center, 6383 Mahan Drive, Tallahassee, FL, USA
Martin
Hill
Rhodes University, PO Box 94, Grahamstown, South Africa
Richard
Hill
Richard Hill & Associates, C/o Crop & Food research, Christchurch, New Zealand
Hariet
Hinz
CABI Europe Switzerland, 1, rue des Grillons, Delmont, Switzerland
John
Hoffmann
University of Cape Town, Rondebosch, South Africa
Judith
Hough-Goldstein University of Delaware, 531 South College Avenue, Newark, DL, USA
Ruth
Hufbauer
Colorado State University, Department of Bioagricultural Sciences and Pest Management, Fort Collins, CO, USA [email protected]
Geoffrey
Hurrell
Agrisearch, Gerald Street, Lincoln, Canterbury, New Zealand
Russell
Hynes
Agriculture & Agri-Food Canada, 107 Science Place, Saskatoon, SK, Canada
Fiona
Impson
University of Cape Town, Rondebosch, South Africa
John
Ireson
Tasmanian Insitute of Agricultural Research, 13, St. Johns Avenue New Town, Hobart, Australia
Caroline
Jackson
University of British Columbia, 6270 University Boulevard, Vancouver, BC, Canada
Ashwini
Jadhav
University of Witwatersrand, School of Animal, Plant and Environmental Sciences, Wits, Johannesburg, South [email protected]
Roman
Jashenko
Institute of ZoologyCABIOCL, 93 Al-Farabi Avenue, Almaty, Kazakhstan
Tracy
Johnson
USDA, PO Box 236, Volcano, HI, USA
Walker
Jones
USDAARS, EBCL, Campus de Bailarguet, Montferrier sur Lez, France
Mireille
Jourdan
CSIRO European Laboratory, Campus International de Baillarguet, Montferrier sur Lez, France
Mic
Julien
CSIRO European Laboratory, Campus International de Baillarguet, Montferrier sur Lez, France
Faith
Kalibbala
University of the Witwatersrand, School of Animal Plant and Environmental Sciences, Wits, Johannesburg,
Africa
South Africa
Yaowei
Kang
Novozymes Biologicals, 5400 Corporate circle, Salem, NC, USA
Younes
Karimpour
Faculty of Agriculture, URMIA University, Urmia, Iran
Delegate List
737
Given name
Given name
Surname
Address
Anna
Karova
Faculty of Plant Protection and Agroecology, Agricultural University of Plovdiv, 12 Mendelee Street,
Javid
Kashefi
USDAARS, EBCL, Tsimiski 43, 7th Floor, Thessaloniki, Greece
Elizabeth
Katovich
University of Minnesota, 411 Borlaug Hall, 1991 Buford Circle, St Paul, MN, USA
Nod
Kay
ENSIS, Private Bag 3020, Rotorua, New Zealand
Anthony
King
University of the Witwatersrand, School of Animal Plant and Environmental Sciences, Wits, Johannesburg,
Plovdiv, Bulgaria
Carien
Kleinjan
University of Cape Town, Rondebosch, South Africa
Eva
Kohlschmid
University of Hohenheim, Garbenstrasse 13, Stuttgart, Germany
Loke
Kok
738
Virginia Tech, 216 Price Hall, Blacksburg, VA, USA
Tamara Mikhailovna Kolomyets
All Russian Rice Research of Phytopathology, Russia, Moscow region, Bolshie Vjasjemi, Russian Federation
Darren
Kriticos
ENSIS, PO Box E4008, Kingston, Australia
Adam
Lambert
UC Santa Barbara, Marine Science Institute, Santa Barbara, CA, USA
Thomas
Le Bourgeois
CIRAD, TA A51 / PS2, Boulevard de la Lironde, 34398 Montpellier, France
Francesca
Lecce
BBCA, Via del Bosco, 10, Sacrofano, Rome, Italy
Ivanka
Lecheva
Agricultural University of Plovdiv, Mendeleev Street 12, Plovdiv, Bulgaria
Kiss
Levente
Hungarian Academy of Science, Herman Otto ut 15, Budapest, Hungary
Jeff
Littlefield
Montana State University, PO Box 173120, Bozeman, MT, USA
Tatamze
Malania
Institute of Plant Protection, 82, Chavchavadze Avenue, Tbilisi, Georgian Republic
Isabelle
Mandon
Conservatoire Botanique National , Le Castel Sainte Claire, Chemin Sainte Claire, Hyres, France
Veronica
Manrique
University of Florida, 2199 South Rock Road, Fort Pierce, FL, USA
George
Markin
US Forest Service, 1648 5. 7th Avenue, Bozeman, MT, USA
Peter
Mason
Agriculture Canada, K. W. Neatby Building, 960 Carling Avenue, Ottawa, Canada
Alec
McClay
McClay Ecoscience, 15 Greenbriar Crescent, Sherwood Park, Alberta, Canada
Andrew
McConnachie
ARCPPRI, Private Bag X6006, Hilton, South Africa
Peter
McEvoy
Oregon State University, 2082 Cordley Hall, Corvallis, OR, USA
Rachel
McFadyen
Weeds CRC, Block B. 80 Meiers Road, Indooroopilly, Brisbane, QLD, Australia
Fernando
McKay
USDAARS, SABCL, Bolivar 1559 , Hurlingham, Buenos Aires, Argentina
Julio
Medal
University of Florida, PO Box 110620, Gainesville, FI, USA
Jean-Yves
Meyer
Delegation a la Recherche, Government of French Polynesia, BP 20981 Papeete, Tahiti, French Polynesia
Joseph
Milan
Idaho State Department of Agriculture, 3948 Development Avenue, Boise, ID, USA
Louise
Morin
CSIRO Entomology, GPO Box 1700, Canberra, ACT, Australia
XII International Symposium on Biological Control of Weeds
South Africa
Given name
Surname
Address
Heinz
Mller-Schrer
University of Fribourg, Perolles, Fribourg, Switzerland
Rangaswamy
Muniappan
IPM CRSP, Virginia Tech, 1060 Litton-Reaves Hall, Blacksburg, VA, USA
Judith
Myers
University of British Columbia, 6270 University Boulevard, Vancouver, Canada
Julie
Nachtrieb
University of North Texas US Army Engineer, Research Development Center, 2312 James Street, Denton,
TX, USA
Navajas
INRA, 488, rue de la croix Lavit, Montpellier, France
Joseph
Neal
North Carolina State University, 262 Kilgore Hall, Raleigh, NC, USA
Patricia
Neenan
CABI, 47, Dundas drive, Rochester, NY, USA
Hernan
Norambuena
INIA, Casilla 58-D, Temuco, Chile
Andrew
Norton
Colorado State University, C129 Plant Sciences, Fort Collins, CO, USA
Stephen
Novak
Boise State University, 1910 University Dr., Boise, ID, USA
Victoria
Nuzzo
Natural Area Consultants, 1 West Hill School Road, Richford, NY, USA
Walter
Ogutu
University of Fribourg, Chemin du Muse 10, Fribourg, Switzerland
Allessandra
Paolini
BBCA, Via del Bosco 10, Sacrofano, Rome, Italy
Quentin
Paynter
Manaaki Whenua Landcare Research, 231 Morrin Road, St Johns, Auckland, New Zealand
Gary
Peng
Agricultural Agri-Food Canada, 107 Science place, Saskatoon, SK, Canada
Hlne
Petit
ATEN, Residence Le Coromandel, 6, rue Achille Bege, Montpellier, France
Michael
Pitcairn
CDFA, 3288 Meadowview Road, Sacramento, CA, USA
Paul D.
Pratt
USDAARS / IPRL, 3225 College Avenue, Fort Lauderdale, FL, USA
Kenneth
Puliafico
University of Idaho, 511 A Remuera Road, Remuera, Auckland, New Zealand
Matthew
Purcell
CSIRO Entomology, 120 Meiers Road, Indooroopilly, Brisbane, QLD, Australia
David
Ragsdale
University of Minnesota, 1980 Folwell Avenue, 219 Hodson Hall, St Paul, MN, USA
Sethu
Ramasamy
RMIT University, Level 223.1.61 A, Bundoora, Australia
Min
Rayamajhi
USDAARS, 3225 College Avenue, Fort Lauderdale, FL, USA
Brian
Rector
USDAARS, EBCL, Campus International de Baillarguet, Montferrier sur Lez, France
Gadi
Reddy
University of Guam, Agricultural Experiment Station, Mangilao, GU, USA
Adele
Reid
CSIRO Entomology, GPO Box 1700, Canberra, ACT, Australia
George
Roderick
University of California, Berkeley, CA, USA
Lo
Ruamps
USDA-ARS, EBCL, Campus International de Baillarguet, Montferrier sur Lez, France
Mohammed Hassan Safaralizadeh
Faculty of Agriculture, URMIA University, Urmia, Iran
Jean-Louis
Sagiolocco
Department of Primary Industries, PO Box 48, Frankston, Australia
Joaquin
Sanabria
Texas A&M University, 720 E. Blackland Road, Temple, TX, USA
Delegate List
739
Maria
Given name
Address
Sankaran
Kerala Forest Research Institute, Peechi, Trichur, Kerala State, India
Raghu
Sathyamurthy
University of Illinois, Illinois Natural History Survey, 1816 S. Oak Street, Champaign, IL, USA
Urs
Schaffner
CABI EuropeSwitzerland, 1, chemin des Grillons, Delmont, Switzerland
Steeve
Schawann
CSIRO EL, 93 rue Jean Francois Breton, Montpellier, France
Shon
Schooler
CSIRO Entomology, Long Pocket Laboratories, 120 Meiers Road, Indooroopilly, QLD, Australia
Mark
Schwarzlander
University of Idaho, E.J. Iddings Agricultural Sciences Building, Moscow, ID, USA
John
Scott
CSIRO Entomology, Private Bag 5, PO, Wembley, WA, Australia
Ricardo
Segura
CSIROMexican Field Station, A. Carlon no 5, Col. Ejido 10 de Mayo, Boca del Rio, CP94297, AP14
Veracruz, Mexico
740
Marion
Seier
CABI, Silwood Park, Buckhurst Road, Ascot, Berkshire, UK
Rene
Sforza
USDAARS, EBCL, Campus International de Baillarguet, Montferrier sur Lez, France
Richard
Shaw
CABI EuropeUK, Silwood Park, Buckhurst Road, Ascot, Berkshire, UK
Andrew
Sheppard
CSIRO Entomology, GPO Box 1700, Canberra, ACT, Australia
David
Simelane
ARCPPRI, P/Bag X134, Queenswood, Pretoria, South Africa
Sarah
Simons
CABI, United Nations Avenue, Gigiri, Nairobi, Kenya
Sharlene
Sing
Montana State University, PO Box 173120, Bozeman, MT, USA
Luke
Skinner
Minnesota Department of Natural Resources, 500 Lafayette Road, Saint-Paul, MN, USA
Tatyana
Skupko
All Russian Rice Research Institute, P/o Belozernoye, Krasnodar, Russian Federation
Michael
Smart
Corps of Engineers, 201 E. Jones Street, Lewisville, TX, USA
Lincolm
Smith
USDAARS, Western Regional Research Center, 800 Buchanan Street, Albany, CA, USA
Lindsay
Smith
Landcare Research, PO Box 40, Lincoln, New Zealand
Alejandro
Sosa
USDAARS, SABCL, Bolivar 1559, Hurlingham, Buenos Aires, Argentina
Helen
Spafford-Jacob
University of Western Australia, School of Animal Biology (MO85), 35 Stirling Hwy, Crawley, WA, Australia [email protected]
Uwe
Starfinger
Fed. Biol. Research Center, Messeweg 11/12, Braunschweig, Germany
Shelli
Stewart
University of Idaho, 2064 S. Gourley Street, Boise, ID, USA
Atanaska
Stoeva
Agricultural University, Plant Protection Research Institute, 12 Mendeleev St., Plovdiv, Bulgaria
Lorraine
Strathie
ARCPlant Protection Research Institute, Private Bag X6006, Hilton, South Africa
Marianna
Szucs
University of Idaho, E.J. Iddings Agricultural Sciences building, Moscow, ID, USA
Robert
Tanner
CABI Bioscience, Silwood Park, Ascot, Berkshire, UK
Gary
Taylor
University of Adelaide, Waite Campus, PMB I, Glen Osmond, Adelaide, Australia
Elizabeth
Tewksbury
University of Rhode Island, 9 East Alumni Avenue, Kingston, RI, USA
Thierry
Thomann
CSIRO European Laboratory, Campus International de Baillarguet, Montferrier sur Lez, France
XII International Symposium on Biological Control of Weeds
Surname
Surname
Address
Hillary
Thomas
University of California-Davis, 1 Shields Avenue, Davis, CA, USA
David
Thompson
New Mexico State University, MSC 3BE, Las Cruces, NM, USA
Philip
Tipping
USDAARS, 3225 College Avenue, Fort Lauderdale, FL, USA
Peter
Toth
Slovak Agricultural University, A. Hlinku 2, Nitra, Slovak Republic
Guadalupe
Traversa
University of Bahia Blanca, San Andres 800, Bahia Blanca, Argentina
Carlo
Tronci
Biotechnology and Biological Control Agency, Via del Bosco 10, Sacrofano, Rome, Italy
Susan
Turner
Ministry of Forests and Range, 515 Columbia Street, Kamloops, BC, Canada
Rieks
Van Klinken
CSIRO, 120 Meiers Road, Indooroopilly, Brisbane, QLD, Australia
Laura
Van Riper
University of Minnesota, 411 Borlaug Hall, 1991 Upper Buford Circle , St Paul, MN, USA
Julien
Vendeville
Biobest, ZAC Porte Sud, Orange, France
Marcelo
Vitorino
Blumenau University, R. Antonio da Veiga, 140, Blumeneau, Santa Catarina, Brazil
Janine
Vitou
CSIRO European Laboratory, Campus de Baillarguet, Montferrier sur Lez, France
Svetlana
Volkova
All Russian Rice Research Institute, ARRRI, Belozerny, Krasnodar, Russian Federation
Mark
Volkovitsh
Zoological Institute RAS, Universitetskaya, 1, St Petersburg, Russian Federation
Maurizio
Vurro
Istituto Tossine e MicotossineCNR, Via Amendola 122/O, Bari, Italy
Nick
Waipara
Landcare Research, Private Bag 92-170, Auckland, New Zealand
Keith
Warner
Santa Clara University, 580 El Camino Real, Santa Clara, CA, USA
David
Weaver
Montana State University, PO Box 173120, Bozeman, MT, USA
Aaron
Weed
University of Rhode Island, 9 East Alumni Avenue, Kingston, RI, USA
Charles
Wikler
Unicentro / FUPEF, BR 153, Irati, PR, Brazil
Livy
Williams
USDAARSEIWRU, 920 Valley Road, Reno, NV, USA
John
Wilson
Centre for Invasion Biology, Stellenbosch University, Matieland, South Africa
Linda
Wilson
University of Idaho, PO Box 442339, Moscow, ID, USA
Rachel
Winston
University of Idaho, 2064 S. Gourley Street, Boise, ID, USA
Arne
Witt
ARCPRI, Private Bag X134, Pretoria, South Africa CABI, United Nations Avenue, Gigiri, Nairobi, Kenya
Alan
Wood
ARCPPRI, P. Bag X5017, Stellenbosch, South Africa
Meimei
Xiang
Zhongkai University of Agriculture & Technology, 24 Dongsha Street, Guangzhou, China
Alice
Yeates
University of Queensland, School of Integrative Biology , St. Lucia, Australia
Kwasi
Yobo
University of KwaZulu-Natal, Private Bag X01, Scottsville, Pietermaritzburg, South Africa
Costas
Zachariades
ARC-PPRI, Private Bag X6006, Hilton, South Africa
Kashif
Zeehan
LUMAQ, 2, rue de lUniversit, Quimper, France
Delegate List
741
Given name
XII International Symposium on Biological Control of Weeds
742
Key to symposium photograph
743
Adele Reid
Keith Douglass Warner
Anna Karova
Guillermo Cabrera Walsh
Alecu Diaconu
Juan Briano
Ivanka Lecheva
Ralf-Udo Elhers
Raghu Sathyamurthy
Sethu Ramasamy
Freda Anderson
Esther Gerber
Laurence Curtet
Joseph Milan
Pierre Ehret
Alec McClay
Dominique Coutinot
Richard Hill
Uwe Starfinger
Peter Mason
Laura C. Van Riper
Tim Heard
Thierry Thomann
Gloria Antonini
Benjamin Gard
Harry C. Evans
Roman Jashenko
Carlo Tronci
Pietro Tronci
K.V. Sankaran
Aaron Weed
Atanaska Stoeva
Arne Witt
Vili Harizanova
Marcus Byrne
John Hoffmann
Paul Hatcher
John Goolsby
Enrico de Lillo
Jean Louis Sagliocco
40
41
42
43
44
45
46
47
48
48b
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
David Ragsdale
Alessandra Paolini
Francesca Lecce
Richard Hansen
Judith Myers
Rieks van Klinken
Marion K. Seier
Lynley Hayes
Walter Ogutu
Linda Wilson
Roger Becker
Djamila Djeddour
Kwasi Sackey Yobo
Stephen J. Novak
John Wilson
Hernan Norambuena
Mark Schwarzlander
Carol A. Ellison
Hugh Gourlay
Rex Friesen
Robert Bourchier
John Ireson
Jim Cullen
Alan Dowdy
Sarah Simons
Andrew W. Sheppard
Sarah Dodd
Lindsay Smith
Kenneth Puliafico
Peter McEvoy
Graeme Bourdot
Ghislaine Cortat
Urs Schaffner
Andr Gassmann
Heinz Mller-Schrer
Andrew Norton
Stephen Hight
Hariet Hinz
Louise Morin
Robin Adair
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
Alejandro Sosa
George Roderick
Luke Skinner
Richard Shaw
Veronica Manrique
Rodrigo Diaz
Nod Kay
Lars Baker
Jenny Cory
Gary Taylor
Ruth. A Hufbauer
Matthew Purcell
Paul Pratt
Fritzi Grevstad
Fernando McKay
Joanne Daly
Christophe Girod
Joaquim Sanabria
David Thompson
Phil Tipping
Vanessa Carney
Isabelle Mandon-Dalger
Rosemarie de Clerck-Floate
Ronny Groenteman
Peter Toth
Imogen Bassett
Geoffrey Hurrell
Helen Harman
Michael Cripps
Julie Coetzee
Mireille Jourdan
Julien Vendeville
Jean-Yves Meyer
Meimei Xiang
Yaowei Kang
Bertie Hennecke
Gitta Grosskopf
Alan Wood
Susan Turner
John Scott
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
Livy Williams
Quentin Paynter
Thomas Le Bourgeois
Guadalupe Traversa
Lincoln Smith
Tracy Jonhson
Andrea Davalos
Rstem Hayat
Rachel McFadyen
Naomi Diplock
Massimo Cristofaro
Jack DeLoach
Maurizio Vurro
Helen Spafford-Jacob
Eva Kohlschmid
Jeff Littlefield
Mickael Pitcairn
Christopher Dunlap
Ernest Delfosse
Michael J. Grodowitz
Steeve Schawann
Richard Smart
Ryan Brudvig
Jennifer Andreas
Janita Baguant
Martin Hill
Judith Hough-Goldstein
Anthony King
Costas Zachariades
Ted Center
Lorraine Strathie
Lisa Buccellato
Charles Wikler
Sharlene Sing
Kashif Zeehan
Meriem Boutiti-Zouaoui
Fiona Impson
Shon Shooler
Rangaswamy Muniappan
Myriam Barat
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
Carien Kleinjan
Gadi Reddy
David Simelane
Unknown
Gaetano Campobasso
Oona Delgado
Hala Alloub
Andrew McConnachie
Alana Den Breeyen
Loke T. Kok
Jianqing Ding
Pierre Gotteland
Julie Nachtrieb
Marcelo Diniz Vitorino
Julio Medal
Ricardo Segura
Alice Yeates
Ashwini Mohan Jadhav
Faith Kalibbala
Tatyana Skupko
Tamara Kolomyets
Luisa Carvalheiro
Darren Kriticos
Abuelgasim Elzein
Marie-Claude Bon
Janine Vitou
Brian Rector
Mic Julien
Ren Sforza
XII International Symposium on Biological Control of Weeds
744
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
31b
32
33
34
35
36
37
38
39
You might also like
- Proceedings of the XII International Symposium on Biological Control of Weeds First Edition Mic H. Julien download100% (1)Proceedings of the XII International Symposium on Biological Control of Weeds First Edition Mic H. Julien download61 pages
- Proceedings of the XII International Symposium on Biological Control of Weeds First Edition Mic H. Julien - Get the ebook instantly with just one clickNo ratings yetProceedings of the XII International Symposium on Biological Control of Weeds First Edition Mic H. Julien - Get the ebook instantly with just one click45 pages
- Instant ebooks textbook Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan download all chapters100% (5)Instant ebooks textbook Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan download all chapters41 pages
- Instant download Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan pdf all chapter100% (4)Instant download Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan pdf all chapter60 pages
- Instant download Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan pdf all chapter100% (6)Instant download Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan pdf all chapter51 pages
- Proceedings of the 5th International Symposium on Biological Control of ArthropodsFrom EverandProceedings of the 5th International Symposium on Biological Control of ArthropodsNo ratings yet
- Control of Pests and Weeds by Natural Enemies An Introduction to Biological Control 1st Edition Roy Van Driesche - The latest ebook is available, download it todayNo ratings yetControl of Pests and Weeds by Natural Enemies An Introduction to Biological Control 1st Edition Roy Van Driesche - The latest ebook is available, download it today48 pages
- (Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 download100% (2)(Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 download56 pages
- (Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 - Discover the ebook with all chapters in just a few seconds100% (1)(Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 - Discover the ebook with all chapters in just a few seconds52 pages
- Biological Weed Control 18-11-2024 SECTYION BNo ratings yetBiological Weed Control 18-11-2024 SECTYION B24 pages
- Get (Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 PDF ebook with Full Chapters Now100% (1)Get (Ebook) Biological Control of Tropical Weeds using Arthropods by Rangaswamy Muniappan, Gadi V. P. Reddy, Anantanarayanan Raman ISBN 9780521877916, 0521877911 PDF ebook with Full Chapters Now76 pages
- Biological Control A Global Perspective First Edition C Vincent - Download the full ebook version right now100% (1)Biological Control A Global Perspective First Edition C Vincent - Download the full ebook version right now53 pages
- Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan instant download100% (1)Biological Control of Tropical Weeds using Arthropods 1st Edition Rangaswamy Muniappan instant download57 pages
- Inundative Biocontrol. in This Type of Biological Control,: A Joint Alberta, Manitoba, Saskatchewan FactsheetNo ratings yetInundative Biocontrol. in This Type of Biological Control,: A Joint Alberta, Manitoba, Saskatchewan Factsheet6 pages
- The Status of Biological Control and Recommendations For Improving Uptake For The FutureNo ratings yetThe Status of Biological Control and Recommendations For Improving Uptake For The Future13 pages
- [Ebooks PDF] download (Ebook) Trophic and Guild Interactions in Biological Control (Progress in Biological Control) by Jacques Brodeur, Guy Boivin ISBN 9781402047664, 1402047665 full chapters100% (4)[Ebooks PDF] download (Ebook) Trophic and Guild Interactions in Biological Control (Progress in Biological Control) by Jacques Brodeur, Guy Boivin ISBN 9781402047664, 1402047665 full chapters71 pages
- Bam Module 2 Concept of Biological Control of Agriculture PestNo ratings yetBam Module 2 Concept of Biological Control of Agriculture Pest20 pages
- Evaluating Indirect Ecological Effects of Biological Control First Edition E. Wajnberg (Editor) All Chapter Instant Download100% (10)Evaluating Indirect Ecological Effects of Biological Control First Edition E. Wajnberg (Editor) All Chapter Instant Download84 pages
- Trophic and Guild Interactions in Biological Control Progress in Biological Control 1st Edition Jacques Brodeur download100% (1)Trophic and Guild Interactions in Biological Control Progress in Biological Control 1st Edition Jacques Brodeur download55 pages
- Evaluating Indirect Ecological Effects of Biological Control First Edition E. Wajnberg (Editor) 2024 Scribd Download100% (5)Evaluating Indirect Ecological Effects of Biological Control First Edition E. Wajnberg (Editor) 2024 Scribd Download71 pages
- 2.success Stories in Biological ControlNo ratings yet2.success Stories in Biological Control14 pages
- Mycoherbicides: Fungal Plant Pathogens For Biological Control of WeedsNo ratings yetMycoherbicides: Fungal Plant Pathogens For Biological Control of Weeds3 pages
- Weed Ecology in Natural and Agricultural Systems First Edition B D Booth - Read the ebook online or download it to own the full content100% (1)Weed Ecology in Natural and Agricultural Systems First Edition B D Booth - Read the ebook online or download it to own the full content55 pages
- Get Weed Ecology in Natural and Agricultural Systems First Edition B D Booth free all chapters100% (3)Get Weed Ecology in Natural and Agricultural Systems First Edition B D Booth free all chapters86 pages
- The GMO Deception: What You Need to Know about the Food, Corporations, and Government Agencies Putting Our Families and Our Environment at RiskFrom EverandThe GMO Deception: What You Need to Know about the Food, Corporations, and Government Agencies Putting Our Families and Our Environment at RiskNo ratings yet
- Pengaruh Suhu Rendaman Dan Media Tumbuh Terhadap Daya Kecambah Benih Dan Pertumbuhan Tanaman Indigofera ZollingerianaNo ratings yetPengaruh Suhu Rendaman Dan Media Tumbuh Terhadap Daya Kecambah Benih Dan Pertumbuhan Tanaman Indigofera Zollingeriana17 pages
- KRM Public School: Periodic Assessment (April - 2021)No ratings yetKRM Public School: Periodic Assessment (April - 2021)4 pages
- Micropropagation of Calla Lily (Zantedeschia Rehmannii) : Olia OrticulturaeNo ratings yetMicropropagation of Calla Lily (Zantedeschia Rehmannii) : Olia Orticulturae6 pages
- Morphology of Flowering Plants &animal Tissues-Q.pNo ratings yetMorphology of Flowering Plants &animal Tissues-Q.p8 pages
- Chapter 2. Vegetative Morphology of PlantsNo ratings yetChapter 2. Vegetative Morphology of Plants23 pages
- S4 Biology Notes-Kugonza H. Arthur Ndejje Senior Secondary School Growth and Development 2020No ratings yetS4 Biology Notes-Kugonza H. Arthur Ndejje Senior Secondary School Growth and Development 20206 pages
- This Activity Introduces The Relationship Between Flower Structures and PollinationNo ratings yetThis Activity Introduces The Relationship Between Flower Structures and Pollination6 pages
- Contes et Légendes des Fées et des Princesses 2022nd Edition Mylestone - The latest ebook edition with all chapters is now availableNo ratings yetContes et Légendes des Fées et des Princesses 2022nd Edition Mylestone - The latest ebook edition with all chapters is now available80 pages
