Biological Conservation 169 (2014) 128–135

Contents lists available at ScienceDirect

Biological Conservation
journal homepage: www.elsevier.com/locate/biocon

The identity of crop pollinators helps target conservation for improved

ecosystem services
M.P.D. Garratt a,⇑, D.J. Coston a, C.L. Truslove a, M.G. Lappage b, C. Polce b, R. Dean a, J.C. Biesmeijer b,c,
S.G. Potts a
a
Centre for Agri-Environmental Research, School of Agriculture, Policy and Development, University of Reading, UK
b
Faculty of Biological Sciences, University of Leeds, Leeds, UK
c
Naturalis Biodiversity Center, Leiden, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Insect pollinated mass flowering crops are becoming more widespread and there is a need to understand
Received 5 August 2013 which insects are primarily responsible for the pollination of these crops so conservation measures can be
Received in revised form 30 October 2013 appropriately targeted in the face of pollinator declines. This study used field surveys in conjunction with
Accepted 3 November 2013
cage manipulations to identify the relative contributions of different pollinator taxa to the pollination of
two widespread flowering crops, field beans and oilseed rape. Flower visiting pollinator communities
observed in the field were distinct for each crop; while field beans were visited primarily by a few bum-
Keywords:
blebee species, multiple pollinator taxa visited oilseed, and the composition of this pollinator community
Crop pollination
Field beans
was highly variable spatially and temporally. Neither pollinator community, however, appears to be
Oilseed rape meeting the demands of crops in our study regions. Cage manipulations showed that multiple taxa
Ecosystem service can effectively pollinate both oilseed and field beans, but bumblebees are particularly effective bean poll-
Crop pollinators inators. Combining field observations and cage manipulations demonstrated that the pollination
Pollinator conservation demands of these two mass flowering crops are highly contrasting, one would benefit from management
Bumblebees to increase the abundance of some key taxa, whilst for the other, boosting overall pollinator abundance
and diversity would be more appropriate. Our findings highlight the need for crop specific mitigation
strategies that are targeted at conserving specific pollinator taxa (or group of taxa) that are both active
and capable of crop pollination in order to reduce pollination deficits and meet the demands of future
crop production.
Ó 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY license.

1. Introduction (DEFRA, 2012) and with global coverage currently standing at

33.6 million ha (FAOStat, 2013), oilseed is rapidly becoming a crop
Insect pollinators are important for the production of many of global significance.
fruit, vegetable and field crops (Klein et al., 2007) and their contri- Pollination of oilseed rape (Brassica napus) occurs through a
bution to global agriculture has been valued at €153bn annually combination of wind and insect vectors with considerable autog-
(Gallai et al., 2009). Like many European countries, insect pollina- amy apparent (Delaplane and Mayer, 2000). Field and cage studies
tion underpins some key sectors of UK agriculture, particularly the have shown positive effects of insect pollination on pod set and
top and soft fruit industries but with increasing areas of insect seed set (Jauker and Wolters, 2008; Manning and Wallis, 2005;
pollinated field crops such as field beans and oilseed rape being Morandin and Winston, 2005; Stanley et al., 2013; Williams
grown, the current valuation of UK insect pollination services of et al., 1987), with associated benefits to the yield and quality of
£430 million per annum is set to increase (Smith et al., 2011). production (Bommarco et al., 2012). These benefits are dependent
Driven by increasing demand for biofuels, the area of oilseed rape on variety and the genetic origin of the oilseed, with some varieties
in the UK has increased by over 150,000 ha in the past decade showing increased yield responses to insects (Hudewenz et al.,
2013; Steffan-Dewenter, 2003). Overall the contribution of insect
pollination to oilseed production has been estimated to be around
18% of total yield (Bommarco et al., 2012).
Oilseed rape is visited by a variety of pollinating insects world-
wide, including honey bees, solitary bees and hoverflies (Ali et al.,
⇑ Corresponding author. Tel.: +44 118 3766149. 2011; Arthur et al., 2010). In the UK, bumblebees and honey bees
E-mail address: [email protected] (M.P.D. Garratt). were found to be active flower visitors in oilseed fields (Hayter

0006-3207 Ó 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY license.
http://dx.doi.org/10.1016/j.biocon.2013.11.001
 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135 129

and Cresswell, 2006) as well as Andrena spp., Osmia spp. and Lasio- the composition of pollinator communities (Kennedy et al., 2013),
glossum spp. solitary bee species (Woodcock et al., 2013). The pol- we used Corine Land Cover map (European Environment Agency,
lination efficiency of different taxa for oilseed has been shown to 2010) to characterise the local landscape and chose sites along a
vary, Osmia bicornis increased pod set when compared to hoverflies gradient of semi-natural habitat. Field bean fields varied between
(Jauker et al., 2012) and considerable variation between bee spe- 0% and 46% semi-natural at a 2 km radius and oilseed varied be-
cies in Pakistan was found (Ali et al., 2011). Furthermore, pollinator tween 0-37%. This ensured that the pollinator surveys in each of
behaviour on oilseed flowers in terms of stigma contact and time our crops would provide a good reflection of the variation in polli-
spent foraging varies (Woodcock et al., 2013), and the amount of nator communities that might be expected in the wider landscape.
pollen carried by different oilseed flower visiting insects depends In each field, two 150 m crop tramlines were selected at least 50 m
on taxa (Stanley et al., 2013). Given the variety of wild insects that from the field edge. At 50 m intervals along each tramline, three
visit oilseed flowers, and their potential impact on crop production, crop watch areas were established measuring 2 m by 1 m in bean
our understanding of the actual contribution of different taxa to fields and 2 m by 2 m in oilseed rape. At each location, 15 min crop
crop pollination in the wider landscape of the UK remains limited. watches were carried out three times during the season, at early,
With a total production area of 168,000 ha in the UK in 2010 mid and late flowering. All floral insect visits, as well as the num-
(DEFRA, 2012) and 2.3 million ha grown worldwide (FAOStat, ber of open flowers, within the crop watch area were recorded.
2013), another important insect pollinated field crop is the field bean Flower visitors were divided into five taxa: honey bees, bumble-
(Vicia faba). The positive effects of insect visits on the pollination of bees, solitary bees, hoverflies and others (which included other
field beans has long been appreciated (Free, 1993), with associated Diptera, Lepidoptera, Hymenoptera and Coleoptera). Where possi-
increases in pod set, beans per pod and pod weight; positive impacts ble, pollinators were identified to species, and in beans, whether
on pod distribution on the plant have also been observed the visitor was nectar raiding or carrying out legitimate visits
(Aouar-Sadli et al., 2008; Benachour et al., 2007; Kendall and Smith, was recorded. Surveys were carried out only when temperatures
1975). It has been suggested that only long-tongued bumblebees can were in excess of 15 °C and with no more than light wind. Flower-
access nectar due to the floral anatomy (Free, 1993) but legitimate ing occurred throughout May for field beans and from the end of
visitation by honey bees (Kendall and Smith, 1975) and solitary April to the end of May for oilseed. Field bean surveys were under-
bee species have been observed, although raiding behaviour by some taken in 2011 in Berkshire on winter sown field beans, variety Wiz-
bumblebee species and honey bees is common (Aouar-Sadli et al., ard. Oilseed surveys were carried out in 2012 in Yorkshire on the
2008; Benachour et al., 2007; Tasei, 1976). Pollinator communities restored hybrid varieties Excalibur and DK Expower.
visiting bean flowers in the field have been characterised more re-
cently in North Africa (Aouar-Sadli et al., 2008; Benachour et al., 2.2. Effect of different pollinators on field bean and oilseed rape
2007), and in 1976 in France, honey bees, bumblebees and several pollination
solitary bee species were observed visiting field beans with varying
proportions of legitimate and raiding visits (Tasei, 1976). A system- To enable manipulation of both flowering crops and pollinators,
atic survey of field bean visitors and their relative contribution to flight cages were constructed at the University of Reading and Uni-
pollination in the UK has not been undertaken. versity of Leeds experimental farms, using 2.4 by 2.1 m frames cov-
In the UK, there is increasing demand for insect pollination ser- ered in polyethylene mesh with a gauge size of 1.33 mm. In
vices, particularly as field crops reliant on wild pollinators, like oil- separate flight cages, four potential crop pollinators were estab-
seed rape, become more widespread (Breeze et al., 2011). With the lished: honey bees (Apis mellifera), bumblebees (Bombus terres-
continued decline of potential insect crop pollinators, both wild tris-audax – a UK subspecies), a solitary mason bee (O. bicornis)
(Biesmeijer et al., 2006; Carvalheiro et al., 2013; Potts et al., and a hoverfly (Episyrphus balteatus). These pollinators were cho-
2010a) and managed (Potts et al., 2010b), possible associated im- sen because they are commercially available and represent four
pacts on production are a concern. If pollination services are to distinct flower visiting insect guilds which may be effective crop
be sustainably managed to maintain crop productivity in the face pollinators. Pollinators were provided with nesting and forage re-
of increasing demand and continued pollinator decline, it is essen- sources within the cage when not involved in experiments, thus
tial that we identify those pollinators key to production of our encouraging natural foraging behaviour for the period of experi-
most widely grown insect pollinated crops and quantify whether mentation. Apis mellifera, through the use of a double entrance
their activity in the field is adequate. Only then can pollination ser- hive, was also given access both to the flight cage and the outside.
vice management strategies be targeted at appropriate species in To compare the effects on pollination of our four pollinator spe-
order to stabilise and improve crop production. cies, flowering oilseed rape and bean plants were placed in a ran-
The aims of the present study were to use field surveys to identify domised block in flight cages with pollinators for a controlled
insect pollinators which are floral visitors of two important UK flow- number of visits per flower. Oilseed rape (cultivar: Heros) and field
ering crops, field bean and oilseed, as well as establishing their rela- bean (cultivar: Clipper) plants were grown individually in pots.
tive level of activity in the field. Then, by using cage manipulation Experimental plants were planted in multiple temporal cohorts
experiments, measure the crop pollination effectiveness of poten- to ensure plants at the correct phenological stage were available
tially important insect pollinators, thereby identifying those taxa that for pollinator treatments and to enable repeated experimentation
are currently primarily responsible for crop production and whether through time. During pollinator exposures, cages contained either
their activity in the field is meeting the demands of the crop. This is 3 bean plants, or 10 oilseed plants, of which 5 were experimental.
essential information to underpin pollination service management Within the cage, a focal plant was selected at random and all flow-
strategies for safeguarding crop production in the future. er visits to that plant were recorded until a threshold number of
visits was reached. By incorporating the total number of flowers
within the cage, pollinator visitation rates to experimental plants
2. Materials and methods could be manipulated by controlling the length of time plants were
inside cages. Experimental visitation rates used were 1 (low) and 3
2.1. Pollinator communities of field bean and oilseed rape (high) visits on average per flower for oilseed, and 1 (low), 2 (med-
ium) and 4 (high) visits on average per flower for field bean. Fol-
For each crop, pollinator surveys were carried out in eight fields lowing exposure to pollinators, all flowers in anthesis on each of
at least 2 km apart. Acknowledging that landscape structure affects the experimental plants were marked with cable ties. Due to
130 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135

potential effects of plant phenology on responses to pollination, cohort (1–8) and replicate within cohort (1–3) as random effects.
only oilseed plants which had any of the first 30 flowers on the Due to non-normal data, seed weight was log transformed prior
main stem open and field beans, in flower up to node 11, were used to analysis. Pod set represents the proportion of flowers exposed
for experiments. to pollinators that set pods. A generalised linear mixed effects
The availability of plant cohorts at the appropriate phenological model with a binomial error structure and the same fixed and ran-
stage, in conjunction with active pollinators within flight cages dom effects was used to analyse pod set.
meant that two bean cohorts per year were involved in the study To compare hand pollinated and pollinator excluded treatments
and from these, nine bumblebee, seven honey bee, five mason with insect pollinator treatments, mixed effects models were used
bee and six hoverfly replicates of high, medium and low visitation again for each of the yield parameters for both beans and oilseed.
rates were possible. Eight oilseed cohorts were utilised, from In this case, pollination treatment only was included as a fixed ef-
which nine bumblebee, eight honey bee, eight mason bee and six fect and the separate visit number replicates were included in the
hoverfly replicates of high and low visitation rates were carried model as an additional random effect nested in replicate. All anal-
out. ysis was carried out using R version 2.14.1.
In addition to insect pollinator treatments, for each crop cohort
a series of additional treatments were set up. Ten plants from each
cohort were randomly selected and assigned, in groups of 5, either
3. Results
a hand pollination or pollinator excluded treatment. For oilseed
hand pollinated plants, the first 30 flowers to develop on the pri-
3.1. Pollinator communities of field bean and oilseed rape
mary stem were supplementary pollinated, with pollen from 5 do-
nor plants. For beans, hand pollination on all flowers on two or
All pollinator taxa were observed visiting beans on at least one
three consecutive nodes, between nodes 1 and 11 on one stem of
occasion. Of those bumblebees that were positively identified, 54%
each plant was done using pollen from two donor plants. For the
of legitimate visits were made by B. terrestris/lucorum, 19% by B.
pollinator exclusion treatments, the five randomly selected plants
hortorum, 17% by B. lapidarius, 8% by B. pascuorum, 1% by B. hypno-
from each cohort were stored in isolation cages for the duration of
rum and less than 1% by B. pratorum. In addition to these legitimate
flowering.
visits, a number of bee species were recorded raiding floral nectar
Before and after pollinator exposure, plants were stored, by co-
through the back of the flower. Eighty-three percent of visits by B.
hort, in randomised blocks within isolation cages and allowed to
pratorum were raids, 50% of B. hypnorum, 44% of B. terrestris/luco-
mature and ripen naturally. Hand pollinated and pollinator ex-
rum, 29% of B. lapidarius, 10% of B. pascuorum, 2% of B. hortorum
cluded plants were stored with their respective cohorts. At harvest,
and 23% of visits by honey bees were raids. Legitimate visits per
the number of bean pods per node and for oilseed, the number of
flower per minute by bumblebees was significantly higher than
set and failed pods from experimentally manipulated flowers
all other pollinator taxa (F4-115 = 16.61, P = <0.0001) (Fig. 1). There
(those marked with cable ties), was noted. Field bean pods then re-
was no significant effect of survey round (F2-106 = 0.081, P = 0.92),
ceived further drying for 48 h in an 80 °C oven. The number of
field site (F7-108 = 2.07, P = 0.053) or a pollinator:round interaction
beans per pod was recorded and beans were weighed to the near-
(F8-98 = 0.078, P = 1.0) on insect visitors.
est 0.001 g. For oilseed, five randomly selected experimentally
All study taxa were observed visiting oilseed flowers. There
manipulated pods from each plant were collected. The number
was no significant difference in the visitation rates of different
and weight, again to the nearest 0.001 g, of all seeds in those pods
pollinator taxa in oilseed fields (F4-98 = 1.17, P = 0.33) (Fig. 2).
was recorded.
There was a significant effect of site (F7-98 = 2.96, P = 0.0074)
and a pollinator:survey round interaction (F8-98 = 2.50,
2.3. Analysis
P = 0.016). Overall, there was no significant effect of survey round
on visits per flower per minute (F2-98 = 2.76, P = 0.068). A high
An average visitation rate (visit per flower per minute) was cal-
number of non-syrphid flies were also observed on flowers,
culated across the 6 crop watch locations of each field for each sur-
although movement between flowers during observations was
vey round. Analysis of variance was used to analyse the influence
very rare. The pollination efficiency of these flies is not known
of pollinator taxa, survey round, site and the pollinator:survey
and they were not subject to cage manipulations as part of this
round interaction on visitation rate. Models were then simplified
study, they were therefore excluded from the analysis. Further re-
until only pollinator taxa and any other significant effects re-
search is necessary to understand the contribution of other dip-
mained. If there was a significant effect of pollinator taxa on visita-
tera groups on the pollination of oilseed.
tion rate then a Tukey honest significant difference test was used
to determine significant differences between pollinator taxa. Visi-
tation data was log + 1 transformed to improve normality prior
to analysis.
Linear mixed effects models were used to analyse pollinator and
visit number effects on bean pods per node, beans per pod, bean
weight and pod weight. Pollinator, visit number (L, M, H) and their
interaction were included in the model as fixed effects; Year (2011,
2012), University (Reading, Leeds) and replicate within cohort (1–
4) were random effects. Models were then simplified to include
only significant fixed effects. To improve normality, pods per node
was log + 1 transformed prior to analysis. Plants which produced
no pods on the treatment nodes were removed from the bean
number, bean weight and pod weight analysis. Linear mixed effects
models were also used to investigate pollinator and visit number
effects on seeds per pod, seed weight and pod weight for oilseed. Fig. 1. Visits/flower/minute shown by some potentially important pollinator taxa
Pollinator, visit number (L, H) and their interaction were included legitimately visiting field beans across eight field sites in Berkshire, UK.
in the model as fixed effects with University (Reading, Leeds), Mean ± S.E.M.
 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135 131

pollinators. Lower visitation rates also resulted in fewer seeds

per pod. There was no significant pollinator:visit number interac-
tion on seeds per pod (F3-287 = 1.91, P = 0.13) (Fig. 4). There was a
significant pollinator:visit number interaction effect on seed
weight but no significant direct effects of pollinator or visit number
(Table 2). Pod weight was significantly affected by both pollinator
and visit number, again with hoverflies and low visitation rates
showing the lowest pod weights. Pollination by mason bees also
resulted in greater pod weights than pollination by honey bees
(Table 2). No significant pollinator:visit number interaction was
found. Significantly fewer pods set under low visit numbers but
no significant effect of pollinator or a pollinator:visit number inter-
action on pod set was found (Table 2). There was a significant effect
Fig. 2. Visits/flower/minute shown by some potentially important pollinator taxa of control treatments on seed number (F5-358 = 22.72, P < 0.0001)
visiting oilseed rape across eight field sites in Yorkshire, UK. Mean ± S.E.M. with pollinator excluded and hoverfly treatments setting fewer
seeds than all other treatments and pollinator exclusion also
resulting in fewer seeds than hoverfly pollination (Fig. 4). The same
3.2. Effect of different pollinators on field bean pollination
pattern was seen for pod weight, although pod weight following
pollination by mason bees was also significantly greater than hon-
There was a significant effect of pollinator on pod set
ey bee and hand pollination treatments (Table 2). No such effect
(F3-229 = 11.87, P < 0.0001) of beans, with hoverflies setting signifi-
was seen for seed weight. Pod set was affected by treatment, with
cantly fewer pods per node than bumblebees, honey bees and
pollinator excluded treatments significantly lower than all other
mason bees. Pod set by bumblebees was also significantly greater
treatments (Table 2).
than pod set by honey bees (Fig. 3). There was no significant effect
of visit number (F2-227 = 1.39, P = 0.25) or a pollinator:visit number
interaction (F6-221 = 1.21, P = 0.30) on pod set. There was no signif- 4. Discussion
icant effect of pollinator, visit number or a pollinator:visit number
interaction on beans per pod (Table 1). Similarly, no significant ef- 4.1. Field bean pollination
fect of pollinator, visit number or their interaction on bean weight
or pod weight was found (Table 1). There was a significant effect of The vast majority of pollinators carrying out legitimate flower
control treatments (i.e. hand pollination and pollinator exclusion) visits in field bean fields in Berkshire were bumblebees. This pat-
on pods set per node (F5-244 = 13.84, P < 0.001) with pollinator ex- tern was common throughout the season and across sites, as indi-
cluded treatments setting fewer pods than bumblebees, honey cated by the absence of significant survey round and site effects.
bees and mason bees. Hand pollination treatments also resulted The clear prominence of bumblebees visiting beans is perhaps con-
in significantly greater pods set than hoverfly pollination (Fig. 3). sistent with the morphology of bean flowers limiting access to nec-
There was no significant effect of control treatments on beans tar for smaller solitary bee species and honey bees, and supports
per pod, bean weight or pod weight (Table 1). conclusions made by Free (1993). Insect visitation improved pod
set in beans, and bumblebees, honey bees and mason bees have
the capacity to improve pod set by between 60% and 69%. The ab-
3.3. Effect of different pollinators on oilseed rape pollination sence of a significant visit number effect suggests that good pod set
is achieved with visitation rates as low as an average of one visit
There was a significant effect of pollinator (F3-290 = 7.74, per flower. Bumblebees did increase pod set above that of honey
P = 0.0008) and visit number (F1-290 = 7.55, P = 0.0064) on oilseed bees indicating that they may be particularly effective field bean
seeds per pod, with hoverflies showing fewer seeds than other pollinators, although such a difference between bumblebees and

Fig. 3. Pods per node on field beans following visitation by four different pollinators at three visitation rates per flower (1 = [ ], 2 = [ ], 4 = [ ] visits). Pods per node following
pollinator exclusion and hand pollination also shown, Mean ± S.E.M. Treatments with different letters are significantly different according to a linear mixed effects model.
132 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135

Table 1
Yield measures of field bean following pollination by four different pollinators at three visitation rates per flower (L = 1, M = 2, H = 4 visits). Yield following pollinator exclusion
and hand pollination treatments also shown, Mean ± S.E.M. F and P values for main effects shown, following mixed effects models including pollinators and visit numbers and
models including pollinators and control treatments. Means with different letters are significantly different (P < 0.05).

Pollinator Visit number Beans per pod Pod weight (g) Bean weight (g)
Bumblebee L 2.67 ± 0.16 2.12 ± 0.24 0.78 ± 0.06
M 2.73 ± 0.18 a 1.20 ± 0.22 a 0.71 ± 0.04 a
H 2.44 ± 0.21 1.83 ± 0.29 0.65 ± 0.09
Honeybee L 2.55 ± 0.17 2.08 ± 0.23 0.79 ± 0.06
M 2.29 ± 0.15 a 1.99 ± 0.10 a 0.91 ± 0.07 a
H 2.52 ± 0.16 2.00 ± 0.23 0.78 ± 0.05
Hoverfly L 2.02 ± 0.42 1.74 ± 0.35 0.73 ± 0.15
M 2.70 ± 0.25 a 2.32 ± 0.26 a 0.77 ± 0.10 a
H 2.74 ± 0.27 2.07 ± 0.22 0.79 ± 0.09
Mason bee L 2.56 ± 0.40 1.74 ± 0.08 0.71 ± 0.09
M 2.51 ± 0.16 a 1.46 ± 0.17 a 0.55 ± 0.04 a
H 2.73 ± 0.23 1.81 ± 0.22 0.65 ± 0.04
Pollinators and visit number Pollinator F3-198 = 2.09, P = 0.10 F3-198 = 0.88, P = 0.45 F3-196 = 1.17, P = 0.32
Visit number F2-196 = 0.50, P = 0.60 F2-196 = 0.06, P = 0.94 F2-199 = 1.69, P = 0.19
Pollinator:visit number F6-190 = 0.79, P = 0.58 F6-190 = 0.68, P = 0.67 F6-190 = 1.10, P = 0.37
Bumblebee 2.62 ± 0.15 a 2.00 ± 0.19 a 0.72 ± 0.04 a
Honeybee 2.44 ± 0.12 a 2.00 ± 0.16 a 0.83 ± 0.05 a
Hoverfly 2.58 ± 0.15 a 2.02 ± 0.19 a 0.75 ± 0.09 a
Mason bee 2.54 ± 0.04 a 1.64 ± 0.13 a 0.64 ± 0.05 a
Pollinator excluded 2.27 ± 0.20 a 1.97 ± 0.20 a 0.83 ± 0.07 a
Hand pollination 2.82 ± 0.29 a 1.69 ± 0.30 a 0.62 ± 0.14 a
Pollinators and controls F5-202 = 1.20, P = 0.31 F5-202 = 0.49, P = 0.78 F5-202 = 1.54, P = 0.18

Fig. 4. Seeds per pod of oilseed rape following pollination by four different pollinators at two visitation rates per flower ([ ] = 1, [ ] = 3 visits). Seed numbers following
pollinator exclusion and hand pollination also shown, Mean ± S.E.M. Treatments with different letters are significantly different according to a linear mixed effects model.

honey bees were not seen by Kendall and Smith (1975). The inabil- (Free, 1993), but 58% of flowers visited is by no means saturation
ity of the hoverfly, E. balteatus to pollinate beans is unsurprising gi- and production could therefore be vulnerable to bumblebee de-
ven their small size and lack of robustness to carry out legitimate cline or low visitation in poor weather years.
visits. Positive impacts of insect visitation on bean quality in terms
of size, reported in earlier studies (Aouar-Sadli et al., 2008; Benac- 4.2. Oilseed rape pollination
hour et al., 2007), was not apparent in this work.
We showed bumblebees are key bean pollinators and this is a Pollinator surveys showed that oilseed rape flowers are visited
product of their high activity in the field and good pollination effi- by a more diverse pollinator community than field beans and there
ciency. Our field surveys showed an average visitation rate of was no significant difference in visitation rates between any polli-
0.0004 flowers per minute. As bean flowers remain open to bee vis- nator taxa. There was a significant effect of site on visitation and a
its for 3 days (Osborne et al., 1997) and assuming 8 h of pollinator significant pollinator:survey round interaction. This temporal and
foraging per day in good weather, this would mean that, on aver- spatial variation points to seasonal and local landscape effects on
age, 58% of flowers could expect at least one visit. Not all flowers crop visitors. The open and accessible nature of oilseed flowers
on bean plants will set pods regardless of levels of pollination, means they are visited by a diverse pollinator community, one that
and this depends on node location and flower numbers per node is more responsive to seasonal and local factors, particularly when
 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135 133

Table 2
Yield measures of oilseed rape following pollination by four different pollinators at two visitation rates per flower (L = 1, H = 3 visits). Yield following pollinator exclusion and
hand pollination treatments also shown, Mean ± S.E.M. F, Z and P values for main effects shown, following mixed effects models including pollinators and visit numbers and
models including pollinators and control treatments. Means with different letters are significantly different (P < 0.05).

Pollinator Visit number Seed weight (mg) Pod weight (g) Pod set%
Bumblebee L 0.044 ± 0.001 0.102 ± 0.007 95.91 ± 1.74
H 0.042 ± 0.002 a 0.102 ± 0.005 ab 97.96 ± 0.68 a
Honeybee L 0.043 ± 0.002 0.096 ± 0.007 94.89 ± 1.57
H 0.041 ± 0.001 a 0.101 ± 0.004 b 97.07 ± 1.92 a
Hoverfly L 0.044 ± 0.003 0.087 ± 0.009 97.04 ± 1.09
H 0.043 ± 0.003 a 0.097 ± 0.008 c 97.41 ± 1.67 a
Mason bee L 0.044 ± 0.002 0.107 ± 0.007 97.48 ± 1.35
H 0.048 ± 0.002 a 0.114 ± 0.009 a 97.88 ± 0.95 a
Pollinators and visit number Pollinator F6-287 = 1.51, P = 0.21 F3-290 = 5.99, P < 0.001 Z < 1.25, P > 0.21
Visit number F6-287 = 0.17, P = 0.68 F1-290 = 4.79, P = 0.03 Z = 3.19, P < 0.01
Pollinator:visit number F3-287 = 3.09, P = 0.03 F3-287 = 0.74, P = 0.53 Z < 1.01, P > 0.31
Bumblebee 0.043 ± 0.001 a 0.101 ± 0.006 abc 96.94 ± 1.11 a
Honeybee 0.042 ± 0.001 a 0.098 ± 0.005 b 96.03 ± 1.38 a
Hoverfly 0.044 ± 0.003 a 0.092 ± 0.008 c 97.27 ± 1.18 a
Mason bee 0.046 ± 0.002 a 0.110 ± 0.008 a 97.68 ± 1.02 a
Pollinators excluded 0.044 ± 0.002 a 0.067 ± 0.007 d 83.33 ± 3.43 b
Hand pollination 0.042 ± 0.002 a 0.094 ± 0.005 b 95.98 ± 1.54 a
Pollinators and controls F5-358 = 1.20, P = 0.31 F5-358 = 18.73, P < 0.001 Z > 5.83, P < 0.001

compared to the relatively specialised and mobile bumblebees pollinator taxa to improve oilseed pollination, and spatial and tem-
seen in high numbers on beans. This diversity of insect visitors to poral variation in field activity of these taxa, demonstrates that the
oilseed has also been seen in other studies and on similar crops, pollination ecology of oilseed is contrasting to that of field beans.
some showing impacts of seasonality and local landscape (Ali Field beans are reliant on a few key pollinators whilst oilseed is
et al., 2011; Arthur et al., 2010; Hayter and Cresswell, 2006; Rader serviced by a more diverse and variable pollinator community. Pol-
et al., 2012; Woodcock et al., 2013). Many non-syrphid flies were linator management strategies to maintain or improve production
observed on oilseed flowers and although their contribution to pol- in each of these crops must therefore be targeted accordingly.
lination was not tested in this study, it is important that their po- Management to support field bean pollinators should be aimed
tential contribution is quantified in future research, if the at maintaining or increasing bumblebee abundance. Despite the
pollination ecology of oilseed is to be fully understood. proven ability of honey bees and mason bees to pollinate beans,
In common with previous studies, this research highlights the very low activity in the field would suggest resources would be
improved pollination of oilseed flowers following insect visitation better targeted at bumblebees. The establishment of additional flo-
(Bommarco et al., 2012; Jauker et al., 2012; Stanley et al., 2013), ral resources within agricultural landscapes can increase the local
but also highlighted is that very distinct taxa can improve pollina- abundance and diversity of bumblebees (Pywell et al., 2011,
tion when compared with pollinator excluded treatments. Im- 2006). Such measures could be implemented to stabilise bumble-
proved oilseed pollination by bumblebees, honey bees and bee populations or even boost them, improving crop pollination,
mason bees when compared to hoverflies was also apparent, with particularly if flower choice is targeted specifically at those bum-
increased seeds per pod. The number of seeds per pod after expo- blebee species showing potential as good field bean pollinators,
sure to these three pollinators was not significantly different from namely the long tongued species (Carvell et al., 2011). Our study
hand pollination treatments, suggesting that these three pollina- shows the long tongued Bombus hortorum could be a highly effec-
tors are also achieving maximum pollination after as few as three tive bean pollinator due to its high activity in the field and low
visits on average per flower. flower raiding activity. While improving local floral resources can
Using field visitation rates for our potential pollinators and help bumblebee populations in the long-term, maximising field
assuming oilseed flowers are receptive for 3 days (Bell and bean pollination may require planting species that do not co-flow-
Cresswell, 1998), our data demonstrates that in 2012, only 3.4% er with beans, or cutting flower margins during bean flowering so
of oilseed flowers could expect a visit from a pollinator. Given encouraging bumblebees onto the crop. The context of any man-
the positive effects of insect visitation on pollination of oilseed, this agement option in terms of local landscape and agricultural system
indicates that insect pollination service in our study fields could be however, must be considered to maximise its effects (Scheper
severely limited, particularly when 3 visits is better than 1 with re- et al., 2013). Utilisation of commercially available pollinators, as
gards to maximising pollination. This has potential negative impli- seen for some tree crops and in protected cultivation, could be
cations for the yield and quality of UK oilseed (Bommarco et al., adopted. The low unit area value of field beans and high cost of
2012) and needs to be addressed through appropriate manage- commercially produced bumblebees, however, would most likely
ment of insect pollinator communities. preclude this as a viable option, thus local and landscape scale hab-
itat manipulation to conserve bumblebees would be more cost
4.3. Conserving pollinators for improved ecosystem services effective.
Considering the influences of season and local landscape on oil-
Driven by habitat loss and falling floral abundance and diver- seed flower visitors, management to support general pollinator
sity, Europe and the US have seen significant declines in many diversity would provide stability in oilseed pollination services in
bumblebee species (Goulson et al., 2008). Given that six species the face of ongoing landscape and environmental change.
of bumblebee were recorded visiting beans in the present study, Furthermore, management to increase general pollinator abun-
four in significant numbers, declines in any of these species has dance could address the sub-optimal pollinator activity observed
implications for field bean pollination. The ability of distinct in this study. Management of meadows or buffer strips under
134 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135

certain agri-environment schemes have been shown to increase 2006. Parallel declines in pollinators and insect-pollinated plants in Britain and
the Netherlands. Science 313, 351–354.
pollinator diversity and abundance, with associated improvement
Bommarco, R., Marini, L., Vaissiere, B.E., 2012. Insect pollination enhances seed
in pollination service, albeit for non-crop species (Albrecht et al., yield, quality, and market value in oilseed rape. Oecologia 169, 1025–1032.
2007), and sown flower strips not only support bumblebees but Breeze, T.D., Bailey, A.P., Balcombe, K.G., Potts, S.G., 2011. Pollination services in the
also hoverflies (Haenke et al., 2009). Furthermore, natural and UK: how important are honeybees? Agriculture, Ecosystems and Environment
142, 137–143.
semi-natural habitats can benefit pollinator diversity and the sta- Brittain, C., Kremen, C., Klein, A.M., 2013a. Biodiversity buffers pollination from
bility of pollination service (Garibaldi et al., 2011), and these hab- changes in environmental conditions. Global Change Biology 19, 540–547.
itats should be maintained within agricultural landscapes to Brittain, C., Williams, N., Kremen, C., Klein, A.-M., 2013b. Synergistic effects of non-
Apis bees and honey bees for pollination services. Proceedings of the Royal
ensure a diverse and abundant pollinator community for oilseed, Society Biological Sciences Series B 280, 1–7.
although the extent and location of these habitats should be opti- Brosi, B.J., Armsworth, P.R., Daily, G.C., 2008. Optimal design of agricultural
mised (Brosi et al., 2008). To maximise cost benefit of any pollina- landscapes for pollination services. Conservation Letters 1, 27–36.
Carvalheiro, L.G., Kunin, W.E., Keil, P., Aguirre-Gutiérrez, J., Ellis, W.N., Fox, R.,
tor management strategy, the value of pollinator diversity to crop Groom, Q., Hennekens, S., Van Landuyt, W., Maes, D., Van de Meutter, F., Michez,
production, through both synergistic effects and buffering of land- D., Rasmont, P., Ode, B., Potts, S.G., Reemer, M., Roberts, S.P.M., Schaminée, J.,
scape and environmental change, needs to be understood. This re- WallisDeVries, M.F., Biesmeijer, J.C., 2013. Species richness declines and biotic
homogenisation have slowed down for NW-European pollinators and plants.
search has begun (Brittain et al., 2013a,b; Greenleaf and Kremen, Ecology Letters 16, 870–878.
2006; Hoehn et al., 2008) and the present study further highlights Carvell, C., Osborne, J.L., Bourke, A.F.G., Freeman, S.N., Pywell, R.F., Heard, M.S., 2011.
the potential benefits of diverse pollinator communities for crop Bumble bee species’ responses to a targeted conservation measure depend on
landscape context and habitat quality. Ecological Applications 21, 1760–1771.
pollination and in oilseed in particular.
DEFRA, 2012. Agriculture in the United Kingdom (Chapter 5), <http://
www.defra.gov.uk/statistics/files/defra-stats-foodfarm-crosscutting-auk-2011-
chapter05-foodchain-120612.xls>.
5. Conclusion
Delaplane, K.S., Mayer, N.F., 2000. Crop Pollination by Bees. CABI Publishing,
Wallingford.
The proliferation of field grown insect pollinated crops puts European Environment Agency, 2010. Corine Land Cover 2000 raster dataset.
new pressures on wild insect pollination services and it would ap- Version 13.
FAOStat, 2013. Crops, <http://faostat.fao.org/site/567/DesktopDefault.aspx?PageID=
pear, certainly for oilseed that these demands are not currently 567#ancor>.
being met. The importance of insect pollination for crop production Free, J.B., 1993. Insect pollination of crops. In: Leguminosae: Vicia. Academic Press,
is clear but the specific demands of a crop, considering both London (Chapter 34).
Gallai, N., Salles, J.M., Settele, J., Vaissiere, B.E., 2009. Economic valuation of the
pollinator activity in fields and the pollination efficiency of those vulnerability of world agriculture confronted with pollinator decline. Ecological
pollinators, is crop specific, thus pollination service management Economics 68, 810–821.
strategies must be targeted. Some crops, such as oilseed, will ben- Garibaldi, L.A., Steffan-Dewenter, I., Kremen, C., Morales, J.M., Bommarco, R.,
Cunningham, S.A., Carvalheiro, L.G., Chacoff, N.P., Dudenhöffer, J.H., Greenleaf,
efit from management to increase general abundance and diversity S.S., Holzschuh, A., Isaacs, R., Krewenka, K., Mandelik, Y., Mayfield, M.M.,
of pollinator populations so pollination services can be provided in Morandin, L.A., Potts, S.G., Ricketts, T.H., Szentgyörgyi, H., Viana, B.F., Westphal,
different landscapes and in changing environments. By contrast, C., Winfree, R., Klein, A.M., 2011. Stability of pollination services decreases with
isolation from natural areas despite honey bee visits. Ecology Letters 14, 1062–
other crops, including field beans, will benefit from more tailored 1072.
mitigation strategies to increase the abundance of the more func- Goulson, D., Lye, G.C., Darvill, B., 2008. Decline and conservation of bumble bees, In:
tionally important taxa through targeted management of local Annual Review of Entomology. Annual Reviews, Palo Alto. pp. 191–208.
Greenleaf, S.S., Kremen, C., 2006. Wild bees enhance honey bees’ pollination of
landscapes.
hybrid sunflower. Proceedings of the National Academy of Sciences of the
United States of America 103, 13890–13895.
Acknowledgements Haenke, S., Scheid, B., Schaefer, M., Tscharntke, T., Thies, C., 2009. Increasing syrphid
fly diversity and density in sown flower strips within simple vs. complex
landscapes. Journal of Applied Ecology 46, 1106–1114.
This Insect Pollinators Initiative research was funded jointly by Hayter, K.E., Cresswell, J.E., 2006. The influence of pollinator abundance on the
a grant from BBSRC, Defra, NERC, the Scottish Government and the dynamics and efficiency of pollination in agricultural Brassica napus:
implications for landscape-scale gene dispersal. Journal of Applied Ecology 43,
Wellcome Trust, under the Living with Environmental Change
1196–1202.
Partnership. We would also like to thank all the research staff Hoehn, P., Tscharntke, T., Tylianakis, J.M., Steffan-Dewenter, I., 2008. Functional
involved in the work and all the farmers and landowners who en- group diversity of bee pollinators increases crop yield. Proceedings of the Royal
Society B-Biological Sciences 275, 2283–2291.
abled us to carry out research on their land.
Hudewenz, A., Pufal, G., Bogeholz, A.-L., Klein, A.-M., 2013. Cross-pollination
benefits differ among oilseed rape varieties. The Journal of Agricultural Science
References FirstView, 1–9.
Jauker, F., Bondarenko, B., Becker, H.C., Steffan-Dewenter, I., 2012. Pollination
efficiency of wild bees and hoverflies provided to oilseed rape. Agricultural and
Albrecht, M., Duelli, P., MÜLler, C., Kleijn, D., Schmid, B., 2007. The Swiss agri-
Forest Entomology 14, 81–87.
environment scheme enhances pollinator diversity and plant reproductive
Jauker, F., Wolters, V., 2008. Hover flies are efficient pollinators of oilseed rape.
success in nearby intensively managed farmland. Journal of Applied Ecology 44,
Oecologia 156, 819–823.
813–822.
Kendall, D.A., Smith, B.D., 1975. The pollinating efficiency of honeybee and
Ali, M., Saeed, S., Sajjad, A., Whittington, A., 2011. In search of the best pollinators
bumblebee visits to field bean flowers (Vicia faba L.). Journal of Applied
for canola (Brassica napus L.) production in Pakistan. Applied Entomology and
Ecology 12, 709–717.
Zoology 46, 353–361.
Kennedy, C.M., Lonsdorf, E., Neel, M.C., Williams, N.M., Ricketts, T.H., Winfree, R.,
Aouar-Sadli, M., Louadi, K., Doumandji, S.-E., 2008. Pollination of the broad bean
Bommarco, R., Brittain, C., Burley, A.L., Cariveau, D., Carvalheiro, L.G., Chacoff,
(Vicia faba L. var. major) (Fabaceae) by wild bees and honey bees
(Hymenoptera: Apoidea) and its impact on the seed production in the Tizi- N.P., Cunningham, S.A., Danforth, B.N., Dudenhoffer, J.H., Elle, E., Gaines, H.R.,
Ouzou area (Algeria). African Journal of Agricultural Research 3, 266–272. Garibaldi, L.A., Gratton, C., Holzschuh, A., Isaacs, R., Javorek, S.K., Jha, S., Klein,
Arthur, A.D., Li, J., Henry, S., Cunningham, S.A., 2010. Influence of woody vegetation A.M., Krewenka, K., Mandelik, Y., Mayfield, M.M., Morandin, L., Neame, L.A.,
on pollinator densities in oilseed Brassica fields in an Australian temperate Otieno, M., Park, M., Potts, S.G., Rundlof, M., Saez, A., Steffan-Dewenter, I., Taki,
landscape. Basic and Applied Ecology 11, 406–414. H., Viana, B.F., Westphal, C., Wilson, J.K., Greenleaf, S.S., Kremen, C., 2013. A
Bell, S.A., Cresswell, J.E., 1998. The phenology of gender in homogamous flowers: global quantitative synthesis of local and landscape effects on wild bee
temporal change in the residual sex function of flowers of Oil-seed Rape pollinators in agroecosystems. Ecology Letters 16, 584–599.
(Brassica napus). Functional Ecology 12, 298–306. Klein, A.-M., Vaissière, B.E., Cane, J.H., Steffan-Dewenter, I., Cunningham, S.A.,
Benachour, K., Louadi, K., Terzo, M., 2007. Role of wild and honey bees Kremen, C., Tscharntke, T., 2007. Importance of pollinators in changing
(Hymenoptera: Apoidea) in the pollination of Vicia faba L. var. major landscapes for world crops. Proceedings of the Royal Society B: Biological
(Fabaceae) in Constantine area (Algeria). Annales De La Societe Sciences 274, 303–313.
Entomologique De France 43, 213–219. Manning, R., Wallis, I.R., 2005. Seed yields in canola (Brassica napus cv. Karoo)
Biesmeijer, J.C., Roberts, S.P.M., Reemer, M., Ohlemüller, R., Edwards, M., Peeters, T., depend on the distance of plants from honeybee apiaries. Australian Journal of
Schaffers, A.P., Potts, S.G., Kleukers, R., Thomas, C.D., Settele, J., Kunin, W.E., Experimental Agriculture 45, 1307–1313.
 M.P.D. Garratt et al. / Biological Conservation 169 (2014) 128–135 135

Morandin, L.A., Winston, M.L., 2005. Wild bee abundance and seed production in Scheper, J., Holzschuh, A., Kuussaari, M., Potts, S.G., Rundlöf, M., Smith, H.G., Kleijn,
conventional, organic, and genetically modified canola. Ecological Applications D., 2013. Environmental factors driving the effectiveness of European agri-
15, 871–881. environmental measures in mitigating pollinator loss – a meta-analysis.
Osborne, J.L., Awmack, C.S., Clark, S.J., Williams, I.H., Mills, V.C., 1997. Nectar and Ecology Letters 16, 912–920.
flower production in Vicia faba L. (field bean) at ambient and elevated carbon Smith, P., Ashmore, M., Black, H., Burgess, P., Evans, C., Hails, R., Potts, S.G., Quine, T.,
dioxide. Apidologie 28, 43–55. Thomson, A., 2011. UK National Ecosystem Assessment. In: Regulating Services.
Potts, S.G., Biesmeijer, J.C., Kremen, C., Neumann, P., Schweiger, O., Kunin, W.E., UNEP-WCMC, Cambridge (Chapter 14).
2010a. Global pollinator declines: trends, impacts and drivers. Trends in Stanley, D., Gunning, D., Stout, J., 2013. Pollinators and pollination of oilseed rape
Ecology and Evolution 25, 345–353. crops (Brassica napus L.) in Ireland: ecological and economic incentives for
Potts, S.G., Roberts, S.P.M., Dean, R., Marris, G., Brown, M.A., Jones, R., Neumann, P., pollinator conservation. Journal of Insect Conservation, 1–9.
Settele, J., 2010b. Declines of managed honey bees and beekeepers in Europe. Steffan-Dewenter, I., 2003. Seed set of male-sterile and male-fertile oilseed
Journal of Apicultural Research 49, 15–22. rape (Brassica napus) in relation to pollinator density. Apidologie 34, 227–
Pywell, R.F., Meek, W.R., Hulmes, L., Hulmes, S., James, K.L., Nowakowski, M., 235.
Carvell, C., 2011. Management to enhance pollen and nectar resources for Tasei, J.N., 1976. Pollinators of winter field beans Vicia-faba-equina and
bumblebees and butterflies within intensively farmed landscapes. Journal of pollination of male sterile plants for hybrid seed production. Apidologie 7,
Insect Conservation 15, 853–864. 1–38.
Pywell, R.F., Warman, E.A., Hulmes, L., Hulmes, S., Nuttall, P., Sparks, T.H., Critchley, Williams, I.H., Martin, A.P., White, R.P., 1987. The effect of insect pollination on
C.N.R., Sherwood, A., 2006. Effectiveness of new agri-environment schemes in plant development and seed production in winter oilseed rape (Brassica-napus
providing foraging resources for bumblebees in intensively farmed landscapes. L). Journal of Agricultural Science 109, 135–139.
Biological Conservation 129, 192–206. Woodcock, B.A., Edwards, M., Redhead, J., Meek, W.R., Nuttall, P., Falk, S.,
Rader, R., Howlett, B.G., Cunningham, S.A., Westcott, D.A., Edwards, W., 2012. Nowakowski, M., Pywell, R.F., 2013. Crop flower visitation by honeybees,
Spatial and temporal variation in pollinator effectiveness: do unmanaged bumblebees and solitary bees: behavioural differences and diversity
insects provide consistent pollination services to mass flowering crops? Journal responses to landscape. Agriculture, Ecosystems and Environment 171,
of Applied Ecology 49, 126–134. 1–8.