See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/368372074

Ecological effects of mosquito control with Bti: evidence for shifts in the
trophic structure of soil- and ground-based food webs

Article in Aquatic Sciences · February 2023

DOI: 10.1007/s00027-023-00944-0

CITATIONS READS

5 151

5 authors, including:

Brendan G. Mckie Astrid Rita Taylor

Swedish University of Agricultural Sciences Swedish University of Agricultural Sciences
151 PUBLICATIONS 7,456 CITATIONS 42 PUBLICATIONS 1,828 CITATIONS

SEE PROFILE SEE PROFILE

André Frainer Willem Goedkoop

Norwegian Institute for Nature Research Swedish University of Agricultural Sciences
51 PUBLICATIONS 1,902 CITATIONS 120 PUBLICATIONS 4,237 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Brendan G. Mckie on 06 March 2023.

The user has requested enhancement of the downloaded file.

Aquatic Sciences (2023) 85:47
https://doi.org/10.1007/s00027-023-00944-0 Aquatic Sciences
RESEARCH ARTICLE

Ecological effects of mosquito control with Bti: evidence for shifts

in the trophic structure of soil‑ and ground‑based food webs
Brendan G. McKie1 · Astrid Taylor2 · Tobias Nilsson1 · André Frainer3,4 · Willem Goedkoop1

Received: 10 June 2022 / Accepted: 24 January 2023 / Published online: 8 February 2023
© The Author(s) 2023, corrected publication 2023

Abstract
The microbial control agent Bacillus thuringiensis var. israelensis (Bti) has been successfully used worldwide to reduce
abundances of biting Nematocera (Diptera), often with little direct impact on non-target organisms observed. However, the
potential for additional indirect effects on other ecosystem properties, including on trophic linkages within food webs, is
poorly known. We investigated the effects of multiple-year mosquito control treatments using the Bti product V ­ ectoBac®-G
on the stable isotope composition of epigeal and soil-based consumers inhabiting replicate floodplains along the River Daläl-
ven, Sweden. We observed significant changes in the isotopic composition of detritivores feeding at the base of floodplain
food webs. Enchytraeid worms were characterised by 3.5% higher δ13C values in treated floodplains, suggesting increased
consumption of δ13C-enriched food. The overall range of community-wide δ15N values was 56% greater in the treated flood-
plains, whilst δ15N values of oribatid mites were elevated by 97%. These results suggest extra fractionation in the transfer of
nitrogen through floodplain food chains. We conjecture that the ecological mechanisms driving these food web shifts are (1)
the mass mortality of high δ13C A. sticticus larvae, which leaves high concentrations of dead mosquito biomass deposited on
soils at local scales, after the floodwaters have receded and (2) incorporation of the very high δ13C-enriched corn particles
comprising the bulk of the V ­ ectoBac®-G product into floodplain food webs. Our results suggest that repeated applications
of Bti might have wider, still largely unknown implications for nutrient and energy cycles within floodplain ecosystems.

Keywords Aedes sticticus Bacillus thuringiensis var. israelensis (Bti) · Floodplain · Isotopic niche metrics · Soil
microarthropods Spiders · Polyunsaturated fatty acids

Introduction et al. 2020; Land et al. 2019; Schäfer and Lundström 2014).
Unlike most other aquatic invertebrates, Nematoceran Dip-
The microbial control agent Bacillus thuringiensis var. tera have an alkaline gut environment, within which Bti
israelensis (Bti) has been successfully used worldwide to produces crystal aggregations containing toxins disrupting
reduce emergence of adult mosquitoes (Culicidae) and other the gut wall (Boisvert and Boisvert 2000). This results in
nuisance blood-feeding Nematocera (Diptera) from lakes, mortality of the target organisms in the larval stage, without
rivers and floodplains (Boisvert and Boisvert 2000; Brühl the often severe environmental impacts of chemical insec-
ticides (Becker et al. 2010; Boisvert and Boisvert 2000).
However, impacts of Bti on potentially vulnerable non-target
* Brendan G. McKie
[email protected] organisms (NTOs), primarily comprising other Nematocera
such as the ecologically diverse non-biting midges (Chirono-
1
Department of Aquatic Sciences and Assessment, Swedish midae), have been observed in some field studies (Allgeier
University of Agricultural Sciences, 75007 Uppsala, et al. 2019b; Hershey et al. 1998; Vaughan et al. 2008),
Sweden
though not others (Boisvert and Boisvert 2000; Lundstrom
2
Department of Ecology, Swedish University of Agricultural et al. 2010). Additionally, the potential for indirect effects
Sciences, 75007 Uppsala, Sweden
on other ecosystem properties, including trophic linkages
3
Faculty of Biosciences, Fisheries and Economics, UiT The and the flow of carbon and nutrients through food webs, is
Arctic University of Norway, 9037 Tromsø, Norway
increasingly raised as a possible risk associated with the use
4
Norwegian Institute for Nature Research (NINA), Fram of Bti (Brühl et al. 2020; Hershey et al. 1998; Poulin et al.
Centre, 9296 Tromsø, Norway

13
Vol.:(0123456789)
47 Page 2 of 15 B. G. McKie et al.

2010), but has been little assessed in field research (Land the final years of the twentieth century, this region suffered
et al. 2019). increasingly severe outbreaks of the floodplain mosquito
Indirect effects on food webs might arise when Bti- Aedes sticticus Meigen (Schäfer and Lundström 2014). A
induced changes in the density and/or behaviour of one or Bti-based control program initiated in 2002 has been highly
more organism groups result in alterations of food web struc- effective at reducing abundances of flying A. sticticus by
ture (e.g. number of trophic levels), and/or function, e.g. pro- 90–100%, with little evidence of negative outcomes for
cessing of different resources or biomass production (Land NTOs including Chironomidae (Lundstrom et al. 2010).
et al. 2019; Truchy et al. 2015). In the case of Bti, both Nevertheless, increases in abundance of Bti itself in treated
target organisms and the most vulnerable NTOs complete floodplain soils (Schneider et al. 2017), along with substan-
their larval development in aquatic or semi-aquatic habitats tial increases in density and taxonomic richness of hetero-
before emerging into the terrestrial environment as winged trophic protozoans (Östman et al. 2008), point to potential
adults, sometimes at very high densities (Armitage et al. changes in the structure and function of soil and epigeal
1994; Becker et al. 2010; Carlson et al. 2016). Adult aquatic food webs.
insects are often notably rich in polyunsaturated fatty acids We used stable isotopes to characterise shifts in the soil-
which are essential for animal growth and development, and and ground-based food webs of floodplain habitats of the
comprise an important resource for both vertebrate (birds, lower Dalälven River catchment, following more than a
bats, amphibians) and invertebrate (spiders, insects) preda- decade of Bti application (as granular ­VectoBac®-G) for
tors (Müller-Navarra et al. 2000; Ramberg et al. 2020). mosquito control. We studied the isotopic composition of
Accordingly, the massive (e.g. 90–100%, Schäfer and Lund- various groups of ground- and soil-based detritivores, rep-
ström 2014) reductions in the biomass of adult mosquitoes resenting the probable first scavengers of dead mosquitos in
following Bti treatment dramatically reduces a potentially floodplain soils and/or grazers of microbial decomposers of
abundant and nutrient-rich food source for terrestrial organ- mosquito remains, and of key invertebrate predators (spiders
isms. Indeed, 66% reductions in the biomass of emerging and beetles). As a C4 plant, the corn comprising the bulk of
Nematoceran Dipterans (mosquitoes and chironomids) the ­VectoBac®-G product has a strongly contrasting isotopic
from French floodplains treated with Bti were associated composition compared with the C3 plants that comprise the
with reduced breeding success and fitness of house martins vast majority of terrestrial flora in Sweden (Pyankov et al.
(Delichon urbicum L.) (Poulin et al. 2010). 2010). Based on this, we expected that any incorporation of
Control of mosquitoes using Bti has particularly strong these corn particles into the ground-based food webs might
potential to affect ground-based (epigeal) and soil-based alter the C isotope values of consumers. We hypothesised
food webs. Extreme densities of mosquitoes and other that:
mass-emerging aquatic insects regularly exceed the feeding
capacities of their main predators (Yang et al. 2010). As 1) Detritivore δ13C values are expected to be higher in the
such, a large portion of aquatic insect productivity ends up treated than control floodplains, in line with the deposi-
deposited on riparian or floodplain soils as dead adults (Hoe- tion of C4-plant-derived particles used in delivery of
kman et al. 2012). Deposition of dead adult chironomids has Bti.
been linked with an increase in the number of soil-based 2) Consumer δ15N isotopes will be enriched in the treated
detrivores including Collembola and mites (Hoekman et al. floodplains, if the mass accumulation of dead larvae
2011), and to increased soil nutrient availability (Gratton acts as a significant nutrient subsidy in the treated flood-
et al. 2017). Additional indirect and poorly studied effects plains, lengthening food chains (Hoekman et al. 2012).
on soil food webs might arise from the vector by which the
Bti spores are delivered. For example, the widely used for- We further analysed adult A. sticticus for their isotopic
mulation ­VectoBac®-G (Valent BioSciences, USA) consists composition and polyunsaturated fatty acid (PUFA) con-
of pellets derived from the husks of corn (Zea mays) cobs tent, to gain insight into their value as a food resource for
treated with a dried culture of Bti, attached with the help higher consumers (Goedkoop et al. 2007; Müller-Navarra
of corn oil (Lundstrom et al. 2010). During mosquito con- et al. 2000).
trol, these pellets are typically distributed by helicopter over
water bodies or inundated floodplains, where they are either
immediately consumed by biota or, alternatively, enter the Materials and methods
pool of organic detritus available to microorganisms and
detritivores. Study sites and mosquito control
In this study, we focus on the effects of mosquito control
using Bti on epigeal and soil food webs of floodplains in The lower River Dalälven region is characterised by a flat
the lower River Dalälven catchment, Sweden. Beginning in alluvial landscape, encompassing Scandinavia’s only true

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 3 of 15 47

river Delta, and interspersed with extensive wet meadow The ten floodplains were divided into five treated–refer-
habitats that are strongly shaped by recurring floods (Holm- ence pairs based on their proximity to the main river, soil
stedt and Linderheim 2019). High river discharges and sub- types and vegetation composition (Tables S1-2, Fig. 1). This
sequent flooding can occur several times over the course of a paired design was used in our statistical models to account
growing season. Conservation values for the region are high, for background variation amongst the floodplains that might
due to a rich biodiversity of plant and animal life, patches of have affected our analyses of mosquito control effects. Infor-
ancient forest, and the presence of many endangered species mation on soil type was obtained from the Swedish Geologi-
and habitat types. Much of the region has received European cal Survey (SGU, www.s​ gu.s​ e). Vegetation composition was
Union Natura 2000 designation, and is protected in national assessed onsite during field sampling, with the percent cover
parks or nature reserves. of different grasses, herbs and woody vegetation recorded in
Since 2002, the Biological Mosquito Control Project four 5-m2 quadrats per floodplain. Pairs 1–3 were dominated
(BMCP) (www.​mygg.​se) has repeatedly treated several of by grasses and herbs, with pairs 1–2 also characterised by
the floodplain areas to limit the mass emergence of the nui- significant growth of Salix (willow) (Table S1). Pairs 4–5
sance mosquito A. sticticus using Bti. Aedes sticticus is a are characterised by significant Sphagnum cover, which
relatively large-bodied and aggressive mosquito that attacks was lacking in pairs 1–3. Overall, neither the elevation nor
both humans and livestock, and which has had severe socio- ground cover of the vegetation categories differed signifi-
economic effects in the lower River Dalälven region, includ- cantly between treated and floodplain categories, except for
ing on tourism (Schäfer and Lundström 2014). A. sticticus Salix cover which was slightly greater in treated floodplains
lays its eggs in wet-meadow and alder swamp habitats, overall (Table S1).
where their eggs can lay dormant for many years. The eggs
typically hatch when a dry period in spring or summer is Field sampling
followed by a flood with water temperature exceeding 8–10
degrees (Schäfer and Lundström 2009). After the floodwater We collected organisms for stable isotope analysis during
recedes (often after only a few days), the larvae continue July 22–24, 2014, 2 weeks after the most recent mosquito
their development in remaining temporary standing water control campaign (undertaken July 9–10 after the second
habitats. The exact diet of A. sticticus larvae in Sweden has mass emergence of A. sticticus during 2014; www.​mygg.​
not been characterised. However, larval floodplain mos- se). At the time of collection, the flood waters had receded
quitoes elsewhere consume mostly particulate detritus and and the floodplains were relatively dry. We focussed on key,
microbes, with algae important only where environmental predominantly saprophagous, invertebrate groups likely
conditions (light and nutrient levels) favour significant pri- to respond to the deposition of dead mosquitoes and/corn
mary production (Merritt et al. 1992). particles in Vectobac G: (1) Annelid worms, including
We sampled epigeal and soil invertebrates from five earthworms (Oligochaeta: Lumbricidae) and pot worms
treated and five control floodplains that are included in the (Oligochaeta: Enchytraeidae), and (2) soil microarthropod
BMCP long-term monitoring program. All are located in detritivores, comprising springtails (Collembola) and moss
and around Färnebofjärden National Park, in close proximity mites (Acari: Oribatida). The two oligochaete families con-
to the main Dalälven river stem (Fig. 1, Tables S1-2). All sume decaying plant material and other detritus and asso-
sampled floodplains are dominated by “wet meadow” habi- ciated microflora (fungi, bacteria, actinomycetes, algae),
tats, with soils water saturated for much of the growing sea- whilst Lumbricidae also feed on living plant material and
son and characterised by a cover of herbaceous sedges and microfauna (Curry and Schmidt 2007; Didden 1993). The
grasses (Table 1). Treated floodplains have been subjected to microarthropods Collembola and Oribatida similarly encom-
mosquito control starting in 2002, using the granular formu- pass species that predominantly consume decaying vegeta-
lation of VectoBac-G®, applied at a dosage of 13–15 kg per tion and associated microflora, but also include predacious
hectare (Schäfer and Lundström 2014). The continuous area and carrion-consuming scavenger species (Petersen 2002;
of treated habitat encompassing our sampling sites ranged Schneider et al. 2004). We further sampled ground-hunting
from 1.1 to 11.3 k­ m2 (Table S2). Prior to 2014 (the year of predators, comprising wolf spiders (Lycosidae), fishing spi-
our study), treatments had been applied during May–August ders (Pisauridae: Dolomedes fimbriatus) and rove beetles
in most years, and most often 2–3 times per year, excepting (Staphylinidae), to assess whether shifts in isotopic compo-
2004 and 2013 (no applications) and 2009 and 2012 (four sition of detritivores and herbivores can be traced at higher
applications) (Schäfer and Lundström 2014, supplemented trophic levels. Finally, when available we also collected
with information for 2013 from www.​mygg.​se). Reference semi-aquatic Diptera (Tipulidae, Chironomidae), which feed
floodplains have never been treated, and all are situated more on particulate detritus and algae in moist floodplain habitats
than 1.75 km from the nearest treated area (Table S2). (Armitage et al. 1994), and predacious soil mites (Acari:
Mesostigmata).

13
47 Page 4 of 15 B. G. McKie et al.

Fig. 1  Map showing the

location in Sweden of the
Färnebofjärden in the lower
River Dalälven showing the
paired (by numbers) treated (T)
and reference floodplains (R).
See Table S1 for detailed infor-
mation on floodplains and pairs

Table 1  Means of Layman Niche metric Reference Treatment Probability refer- Probability ref-
isotopic niche metrics for ence < treated erence > treated
the reference and control
floodplains, with the probability NR 1.04 2.38 96.55%
of divergence between
CR 2.44 1.86 78.88%
floodplain types estimated based
on probability distributions TA 1.36 2.26 76.98%
CD 0.82 1.01 78.57%
MNND 0.75 0.96 67.95%
SDNND 0.49 0.22 32.05%

Wolf spiders and rove beetles were collected by sweep- beetle group in each floodplain. Captured spiders and beetles
netting vegetation near to ground level, whilst D. fimbriatus were kept alive in individual jars, cold and in the dark, until
was hand collected by visually searching for individuals. We further processing in the laboratory.
aimed to collect at least five individuals for each spider and

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 5 of 15 47

All soil invertebrates, except Collembola, were col- australis) were cut from the water surface down to a depth of
lected from two types of habitats in the floodplains: 10 cm. Both aquatic and terrestrial plant samples were both
raised tussocks formed of grasses and sedges, and damper placed directly into 11 L buckets filled with 2.5L distilled
depressions forming in between the tussocks. The dryer water. The plants were gently brushed in the distilled water
tussocks were targeted for collection of earthworms and to remove biofilm. The water (1.5–2.5 dL) containing bio-
oribatid and predacious mites, whilst pot worms and semi- film was then filtered through pre-combusted Whatman glass
aquatic Diptera were collected from the damp depres- fiber filters (GF/F, effective pore size = 0.7 µm) to capture
sions. We sampled two tussocks and two damp depres- biofilm-associated algae and other microbes. The filters were
sions per floodplain. Tussocks were harder to locate in the then inspected to remove any larger particles, and placed in
Sphagnum-dominated pairs 4 and 5 (Table S1). In these 4-ml tubes for transport back to the laboratory in a cool box.
floodplains we nevertheless divided our sampling between We successfully collected targeted consumer groups
raised, drier areas of ground and lower, wetter depressions. from all floodplains, except for lumbriculid earthworms
To sample soil in the tussocks, we used a spade to first and tipulid dipterans (Table S3). We also obtained four of
cut off the grassy top, and then dug a c. 800–1000 c­ m 3 five targeted resources from all floodplains (Table S3). The
(18 × 18 x c. 2–3 cm depth) soil core. Damp depressions exception was biofilm from the adjacent river waters: in
were sampled using an 8 (inner diameter) × 15 cm cylindri- two reference and two treated floodplains, there was neither
cal soil corer (754 ­cm3). Larger earthworms were sorted Phragmites nor other suitable water plants available for epi-
immediately from the tussock core in the field and retained phytic biofilm collection.
in jars with additional soil from the site. All soil core sam- Unfortunately, we were unable to obtain A. sticticus dur-
ples were then stored in plastic bags and returned in cool ing our two field sampling visits. There was no extensive
boxes to the laboratory for extraction of soil fauna on the mosquito emergence during our study visits, which were
same day. Biota from the tussock samples were extracted conducted in between flood events when standing water on
over Tullgren funnels for a total of 4 days (Tullgren 1971). the flood plains was minimal. However, the BMCP were
Soil cores from the damp depressions were extracted over able to provide us with 12 adult A. sticticus hatched from
3 days using a cold wet extraction technique, with the aid mature (pre-pupal) larvae. The larvae were collected from
of modified Baermann funnels (Didden et al. 1995). Ani- temporary standing water habitats at one of our sampling
mals were picked daily during extraction and stored cold areas (Ista) during an outbreak later in the season. Four of
in the dark. the reared adults were sent for analyses of stable isotopes,
Springtails (Collembola) were sampled by collecting a and the remainder were allocated for FA analyses. The iso-
larger soil volume (0.02 ­m3) with a spade from two tus- tope and PUFA composition of these individuals fully reflect
sock habitats per floodplain. In the field, this soil sample resources consumed during larval development in their natal
was placed in a bucket of river water, and individuals were floodplain habitats, as they were newly emerged and had not
picked as they floated to the surface. As we were unable to fed since collection as mature larvae. Finally, the BMCP
collect sufficient springtail biomass for isotopic analysis dur- provided us with a sample of the control product ­VectoBac®,
ing our first sampling visit, we collected additional spring- which we analysed for stable isotopes.
tails during a second sampling trip to the floodplains on
11–12 August 2014. There were no new flood or mosquito
control events between these two trips. Stable isotope and fatty acid analyses
We also collected plant, biofilm and soil samples for iso-
topic analysis, to characterise the basal resources available to Following field work and subsequent laboratory extraction,
earthworms and microdetrivores in the floodplains. Based on collected consumer and resource samples were separated
our vegetation inventories, Cirsium palustre (marsh thistle) into the taxonomic and resource groups described above.
and Carex spp. were selected to represent common terres- After 24–48 h of cold storage, during which invertebrates
trial litter types, occurring in all floodplains. We collected were not supplied with food, all samples were snap frozen in
leaves from multiple specimens of these plants across a 50 liquid nitrogen (− 196 °C) and stored in a freezer at − 40 °C.
­m2 area per floodplain, and stored them cold for return to At the time of freezing, all organisms except the staphyli-
the laboratory. During our second field visit, we additionally nid beetles were alive. After freeze drying, samples were
collected samples for analysis of soil isotopes (five soil cores weighed and packaged into 5 × 9 mm tin capsules. Larger
per floodplain) and biofilm from both aquatic and terrestrial animals and plants were homogenised prior to weighing to
plants. Biofilm from terrestrial plants was collected by cut- obtain a representative sub-sample. Target weights were
ting plants close to ground level, which were then placed in 1.3 mg for invertebrates, 3.5 mg for plants and 6.0 mg for
11L buckets filled with distilled water. To collect aquatic soil. For filters with biofilm, ten pieces of 4 mm discs were
biofilm, submersed fragments of aquatic reeds (Phragmites clipped out per filter with a hole puncher and stacked in tin

13
47 Page 6 of 15 B. G. McKie et al.

capsules. Three samples with blank filters were weighed and a probabilistic distribution (the posterior) representing the
sent for analysis, to obtain isotopic values for the filters range of possible estimates of the metric.
themselves. Three replicates of each soil and terrestrial plant First, we plotted convex hulls for treated and floodplain
type were also analysed from each floodplain. Unfortunately, communities to visually assess how much and in what direc-
we had insufficient material to repeat this process for the tion the communities diverge from one another in the iso-
biofilm samples. topic biplot. We then applied Jackson et al’s (2011) Bayesian
Stable isotopes have long been used to provide a picture framework to estimate the niche metrics proposed by Lay-
of an organism’s ecological niche, where enrichment in the man et al. (2012):
15
N isotope indicates a higher trophic position, and varia-
tion in the 13C isotope indicates whether the basal resource 1) Ranges of δ13C and δ15N in the niche space (two metrics:
has aquatic or terrestrial origins (e.g. Hoekman et al. 2012; CR and NR, respectively)
Kupilas et al. 2016). Stable isotopes have been recom- 2) Total area (TA) of the convex hull, acting as a measure
mended as a tool in evaluation of whole-ecosystem effects of of the total amount of niche space occupied by the com-
mosquito control using Bti (Boisvert and Boisvert 2000), but munity.
remain little utilised (but see Allgeier et al. 2019a). Our sam- 3) Mean distance to the centroid of points in two-dimen-
ples were analysed for stable isotopes by the Stable Isotope sional space (CD), which acts as a measure of species
Facility at the University of California, Davis, as described spread;
in the supplementary online file S1. 4) Mean nearest neighbor distance in two-dimensional
Fatty acids (FA) that are specific for certain algal groups space (MNDD), functioning as a measure of density of
and conservatively incorporated into the tissues of consum- species packing;
ers can be used as biomarkers, and certain polyunsaturated 5) Standard deviation of nearest neighbor distance
fatty acids (PUFA) are considered essential for the growth (SDNND), measuring evenness of species packing in
and development of consumers (Goedkoop et al. 2007; Mül- two-dimensional space;
ler-Navarra et al. 2000). We analysed our adult mosquitos
their fatty acid composition following the protocol of Lau Overlap in posterior probability distributions was used
et al. (2014), as described in the supplementary online file to assess the percent probability that each Layman metric
S1. differed between reference and treated floodplains (Jackson
et al. 2011). A probability value greater than 95% provides
Data analysis strong support for divergence. Uncertainty in each Lay-
man metric estimate is visualised as credible intervals. All
Stable isotope data were first evaluated visually, using iso- Layman metrics, convex hull estimates and probabilities of
topic biplots and niche space estimation. Isotopic biplots, divergence between floodplains were generated using the
produced with the R package siar (siar: stable isotope anal- SIBER package version 2.1.6 (Jackson et al. 2011) imple-
ysis in R, Parnell and Jackson (2013)), allow for a visual mented in R version 4.1.1 (R core team, Vienna Austria,
assessment of the similarity of potential food sources to http://​www.R-​proje​ct.​org/).
different groups of consumers. In the production of these We supported our descriptions of isotopic niche space
isotope biplots, data from the five control areas and from by analysing differences in isotope composition of consum-
the five Bti-treated areas, respectively, were merged. This ers and resources between treated and reference floodplains
resulted in two charts that summarise data for the repli- (Table S3) using two-sided, paired t tests (JMP version 11.0,
cate treated and reference floodplains. SAS institute). Two-sided t tests were followed by one sided
We used the Bayesian framework developed by Jack- t tests to further evaluate the direction (increase or decrease)
son et al. (2011) to characterise differences in the dimen- of change.
sions of niche space occupied by consumers (i.e. excluding
resources) between treated and reference floodplains. Use
of a Bayesian framework facilitates estimation of commu- Results
nity wide metrics that account for uncertainty in the isotopic
composition of particular community members by incor- Isotopic biplots
porating information on their variability. Here, we used
individual floodplains within Bti treatment categories as The most marked difference in the isotopic biplot between
replicates for estimating uncertainty. Consequently, for all floodplain types was the presence of corn-containing
niche-related metrics, one value is derived for treated and ­VectoBac®-G in the treated floodplains, which had far higher
another for untreated floodplain communities, together with (less negative) δ13C values than the in situ resources (Fig. 2).
Aquatic biofilm samples were characterised by lower (more

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 7 of 15 47

Fig. 2  Isotopic biplots (δ15N

and δ13C) for reference (upper
panel) and treated floodplains
(lower panel) showing basal
resources (colors), key taxa
of detritivores (small letters)
and predators (capital letters),
as well as samples of Aedes
sticticus and ­VectoBac®-G.
Resources plotted as mean val-
ues ± 1 SD. For ­VectoBac®-G,
means (± sd) are plotted for
three technical subsamples,
whilst for adult mosquitoes (A.
sticticus), data for four individu-
als are plotted. Both panels are
plotted using identical axes, to
facilitate comparison

negative) δ13C than the terrestrial plant samples in the refer-

ence but not in treated floodplains. Soil δ13C was also lower 8
than all terrestrial plant samples. The large error bar for
meadow algae δ13C reflects the wide range, from − 25.41 to 6
δ15N ‰

− 33.13, of the replicates obtained. The δ13C of the four A.

sticticus individuals analysed varied from − 27.3 to 28.71, 4
spanning the range of most other detritivores. Finally, detri-
tivore consumers, comprising worms, springtails, oribatid 2
mites and semi-aquatic dipterans, all had lower δ15N val-
ues in the biplot space than the predominantly predacious
−31 −30 −29 −28 −27 −26 −25
groups, including predacious mites, spiders and beetles
(Fig. 2). δ13C ‰

Bayesian Layman metrics and niche ellipses Fig. 3  Isotopic-based convex hulls for reference (black) and treated
(red) floodplain communities, pooling individual floodplains within
each category. Individual data points from reference floodplains are
The convex hulls for the treated and reference floodplain
plotted as circles in grey black, treated floodplains as triangles in red
communities (Fig. 3) overlapped and were similarly sized
(Table 1). However, the reference hull was more spread out
along the δ13C axis (Fig. 3a-b), with the treated community and SDNND) were less than 70% and 35%, respectively
hull more spread out along the δ15N (Fig. 3a). Isotopic niche (Table 1, Fig. 4e, f).
ellipses for individual floodplains are given in Supplemen-
tary Information, Fig. S1.
Amongst the Layman metrics, the range of δ15N in the Consumers and resources
niche space (NR) was 56% higher in treated compared
with reference floodplains, with the probability of diver- Values of δ15N were 97% higher for detrivorous oribatid
gence greater than 95% (Table 1, Fig. 4a). Probabilities of mites (Fig. 5a) in treated compared with reference flood-
divergence in the δ13C range (CR), total area of the con- plains (Table 2). A similar trend for the pisaurid spider
vex hull (TA) and mean distance to the centroid (CD) were Dolomedes fimbriatus (9% higher in treated compared
all > 75%, with trends for 31% lower CR, 39% higher TA and with untreated floodplains, Fig. 5a) was near significance
18% higher CD in treated relative to reference floodplains at the 5% level in the two-sided t test (Table 2). Values of
(Table 1, Fig. 4b, d). Probabilities of divergence in the mean δ13C were 3.5% higher in reference than treated floodplains
and standard deviation of nearest neighbor distance (MNND for Enchytraeidae (Fig. 5b, Table 2). There were several

13
47 Page 8 of 15 B. G. McKie et al.

Fig. 4  Bayesian Layman metric

modes plotted for each reference
and treated floodplains with
95% credible intervals: ranges
of a δ15N (NR) and b δ13C (CR)
in the niche space, c total area
(TA) of niche space occupied,
based on convex hulls; d species
spread, based on mean distance
to centroid (CD); e species
packing, based on mean nearest
neighbor distance (MNDD);
(f) evenness of species pacing,
based on standard deviation
of nearest neighbor distance
(SDNND). Black dots represent
the modal estimate, whilst shad-
ing in the density plot repre-
sents the 50% (dark gray), 75%
(light gray) and 95% (white)
credibility intervals. See Table 1
for probability of divergence
estimates between wetland types

additional trends for lower δ13C and/or higher δ15N in other field-collected fourth instar larvae (Table S3). The most
key consumer taxa in treated compared with reference flood- abundant FA were the monounsaturated oleic acid (18:1ω9)
plains (Fig. 5, Supplementary material Fig. S1), but these and hypogeic acid (16:1ω9), and saturated FA palmitic acid
were either not significant in two-sided t tests, or could not (16:0), which accounted for 19–27.4% of total FA. Amongst
be evaluated due to a lack of replication (Lumbricidae, Tip- PUFAs, eicosapentaenoic acid (20:5ω3, EPA) and arachi-
ulidae) across all floodplains (Table S3). donic acid (20:4ω6) and their precursors α-linolenic acid
Resources obtained from all sample sites (i.e. soil, (18:3ω3, ALA) and linoleic acid (18:2ω6, LA), respectively,
meadow biofilm scrapes, Carex sp. and Cirsium palustre, comprised between 2.4 and 4.9%. Finally, adult A. sticticus
Table S1) did not differ in either δ13C or δ15N values between also had detectable levels of docosahexaenoic acid (22:6ω3,
treated and reference floodplains (all t <|1.3|, p > 0.3). The DHA), i.e. 0.11% (Table S3).
lowest C:N molar mass ratios were observed for A. sticti-
cus (5.43 ± 0.51), and lake (9.44 ± 1.51) and meadow
(12.07 ± 2.95) biofilms (Fig. 6). ­VectoBac®-G had the high- Discussion
est C:N ratio (126.7 ± 19.86, Fig. 6). C:N ratios were higher
in treated floodplains for Carex sp. (27% higher; t = 4.43, Our isotopic analyses of key invertebrate consumers indi-
p = 0.011). No other differences in C:N ratios between flood- cate that more than a decade of mosquito control in the
plain types were detected for the remaining resource types lower Dalälven using a granular formulation of Bti is
(all t <|1.4|, p > 0.23). causing significant changes in the isotopic composition
of key organism groups feeding at the base of floodplain
Fatty acid composition of A. sticticus food webs. Enchytraeid worms were characterised by sig-
nificantly higher δ13C values in treated relative to reference
We detected 23 fatty acids (FA) with relative concentra- floodplains, and oribatid mites were significantly higher
tions exceeding 0.05% in adult A. sticticus reared from in δ15N values. Similar trends that were near significance

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 9 of 15 47

Fig. 5  Differences in mean ± 1

SE values for δ13C (a) and δ15N
(b) of key consumer taxa in
treated (orange) and reference
floodplains (blue). Significance
levels from two-sided paired
t tests of differences (Table 2)
indicated: *p < 0.05, •p < 0.1

Table 2  Results from paired t Taxon δ13C δ15N

tests for each taxon assessing
whether δ13C and δ15N values 2-side 1 side 1 side 2-side 1 side 1 side
differ between the treated and p >|t]p p>t p<t p >|t] p>t p<t
reference floodplains (two
Detritivores Enchytraeidae 0.011 Ns 0.005 Ns ns ns
sided tests) and whether these
values are elevated or lowered Chironomidae ns ns ns ns ns ns
in treated relative to reference Collembola ns ns ns ns ns ns
floodplains (one sided tests) Acari: Oribatida ns ns ns 0.021 0.011 ns
Predators Lycosidae ns ns ns ns ns ns
Dolomedes fimbriatus ns ns 0.094 0.069 0.034 ns
Acari: Mesostigmata ns ns ns ns ns ns
Staphylinidae ns ns 0.063 ns ns ns

Note that no tests were conducted for the earthworms Aporrectodea or Diptera Tipulidae, since in both
cases individuals were sampled from only 2 of the reference sites, and 3–4 of the treatment floodplains
All t tests conducted with 4 degrees of freedom
Significant tests with p < 0.05 highlighted in bold, other results near significance (p < 0.1) highlighted in
italic

13
47 Page 10 of 15 B. G. McKie et al.

Fig. 6  C:N molar mass ratios

of seven resources found in
the reference (blue/dark bars)
and treated (orange/light bars)
floodplains, ordered from lowest
to highest. Biofilm, soil and
plant samples were all collected
directly from the floodplains,
and the mean ± SE of these
resources is calculated at the
floodplain level. A. sticticus
were sampled prior to pupa-
tion from one of the treated
floodplains only (Ista mire),
whilst the V­ ectobac®-G sample
represents the product used for
mosquito control in the flood-
plains. The mean ± SE for the
A. sticticus and ­Vectobac®-G
are calculated based on three
and five analytical replicates.
Significant differences between
floodplain types based on paired
t tests indicated: *p < 0.05

were apparent in predacious staphylinid beetles and the concentrated in the remaining, temporary, standing water
raft spider D. fimbriatus. Finally, our Bayesian analyses habitats after the floodwater recedes (Fig. S2). This dead bio-
of isotopic niche metrics indicate a 96.5% and 79% prob- mass likely enters soil- and ground-based food webs, either
ability of divergence between floodplain types in the ranges as a result of scavenging from detritivores, or as a result of
of community δ15N and δ13C values. Variation in key soil microbial degradation (Gratton et al. 2017; Hoekman et al.
characteristics (soil types, depth, isotopic signature) and the 2012). Our results indicate that dead A. sticticus larvae rep-
composition of ground cover amongst our floodplains were resent a labile and relatively nutrient-rich resource, charac-
minimal, and did not differ systematically between flood- terised by a low C:N ratio and moderate concentrations of
plain types. Accordingly, we conclude these observed food long-chain PUFAs (Table S4). When made available, such
web shifts are a consequence of the repeated, large-scale high-quality food resources are preferentially consumed by
applications of V­ ectoBac®-G in the treated floodplains, i.e. soil detritivores, altering their isotopic composition (Scheu
3–5 times annually during all but 2 years over the period and Schaefer 1998; Tiunov and Scheu 2004). The dead A.
2002–2014. Ecological mechanisms potentially explaining sticticus larvae accumulating in floodplains might also be
these results include (1) incorporation of the high δ13C value attractive to predators and/or scavengers, including oribatid
corn particles, used as a vector in the ­VectoBac®-G product, mites and staphylinid beetles (Schneider et al. 2004; von
into floodplain food webs, and (2) the mass mortality of A. Berg et al. 2012).
sticticus, which results in high concentrations of dead larvae The ground corn particles used in V ­ ectoBac®-G also rep-
deposited on floodplain soils at local scales. resent an additional C input to the epigeal and soil-based
In the absence of mosquito control, the vast majority of food webs of the treated floodplains. ­VectoBac®-G has an
A. sticticus larvae will emerge as adults, and subsequently extremely high C:N ratio, similar to that of some of the
disperse in search of mates and (in the case of females) most nutrient-poor, naturally occuring, plant litter species
blood meals (Fig. S2). Those that escape predation can (Frainer et al. 2015; Handa et al. 2014). However, the per-
then be expected to die after swarming (for males) and ovi- centage of C in corn cob comprised of refractory lignin is
position (which occurs in moist soils), with their corpses relatively low, varying from c. 5–15% depending on grow-
being distributed over a wide geographic area. Dispersal ing conditions (Demirbaş 2005; Mourtzinis et al. 2014;
distances > 11 km have been documented for female A. Pointner et al. 2014). This is similar to the lignin content
sticticus (Mirta et al. 2011). In contrast, under a scenario of of those plant litter species categorised as “fast-decompos-
repeated Bti treatment events, adult abundances are reduced ing” by Handa et al. (2014), and is also at the lower end
by 90–100% in the lower River Dalälven (Schäfer and Lund- of the range observed previously for floodplain herbaceous
ström 2014), thus preventing wider dispersal of mosquito vegetation (9.6–29.4%, Britson et al. 2016). Furthermore,
biomass (Fig. S3). Instead, dead larval biomass remains the ­VectoBac®-G granules also incorporate an unknown

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 11 of 15 47

quantity of corn oil, used to aid attachment of Bti (Lund- The δ13C of newly emerged A. sticticus adults was highly
strom et al. 2010), which might enhance the resource quality variable amongst individuals, spanning the full range of ter-
of the particles. Thus, although nutrient-poor compared with restrial resources. Our PUFA analyses confirm A. sticticus
dead A. sticticus and meadow biofilms, the C characteris- consumes a mixed diet, with important long-chain PUFAs
ing the corn substrate used in ­VectoBac®-G is likely to be synthesised by algae (e.g. EPA 20:5ω3) making up around
relatively labile for decomposing organisms inhabiting a soil 5% of total FA, and the bacterial indicator FA vaccenic acid
matrix that is otherwise rich in highly refractory (Curry and (18:1ω7) making up 2%. This is in line with values reported
Schmidt 2007). by Sushchik et al. (2013) for mosquito adults in floodplains
in the Novosibirsk region in Russia, and those for detritivo-
Carbon characteristics of resources and consumers rous benthic invertebrates from boreal lakes, such as the
isopod Asellus aquaticus (Lau et al. 2012). Finally, doco-
Freshwater biofilms generally have lower δ13C values than sahexaenoic acid (DHA) is common in aquatic invertebrate
terrestrial plants, reflecting the higher presence of light 12C predators (e.g. Lau et al. (2012)), and suggests that A. sticti-
isotopes taken up during algal photosynthesis (Marty and cus also preys on smaller invertebrates. Overall, these results
Planas 2008). However, our biofilm samples were not as suggest that much of the C ingested by the mosquitoes in
13
C-depleted as expected, especially in the treated flood- their temporary floodplain habitats is terrestrial in origin,
plains, given that clean algal samples often have δ13C val- obtained through filtration of bacteria colonising particu-
ues < − 30 (Marty and Planas 2008). It is, thus, likely that late C (Merritt et al. 1992). However, the PUFA and isotope
our aquatic biofilm samples also included particulate C of results indicate this diet is patchily supplemented with fresh-
terrestrial origin. This “contamination” is a common prob- water algal productivity, likely reflecting an uneven deposi-
lem in field-based aquatic isotope studies (Kupilas et al. tion of algal cells over the floodplains during flooding events
2016; Marty and Planas 2008). Interestingly, soil δ13C values (Hein et al. 2003; Tockner et al. 1999).
were on average 4% more negative than those of terrestrial
plants. This likely reflects the long-term influence of flood- Consumer isotopes: comparing treated
ing on the soils of the wet meadows. Each time the floodwa- and reference floodplains
ters recede, carbon derived from aquatic primary producers
and thus depleted in 13C is deposited on the floodplain (e.g. Transitions between trophic levels typically involve a δ15N
as stranded algae or submerged aquatic plants), thus reduc- increase of about 3‰ (from 2.2 to 3.4 ‰) (Vander Zanden
ing soil δ13C (Hein et al. 2003; Tockner et al. 1999). Finally, and Rasmussen 2001). Accordingly, our isotopic analyses
our analyses confirm that the corn used as a substrate in the suggest our sampled invertebrates spanned at least three
Bti control agent V­ ectoBac®-G consitutes a unique C-source trophic levels. The overall increase in the range of δ15N
for the River Dalälven floodplains, with δ13C values that are values characterising consumers in the treated floodplains
substantially higher than any other analysed resource. suggests that repeated applications of V ­ ectoBac®-G have
Carbon isotopes of most consumers were similar to or increased the amount of fractionation involved in nutrient
lighter than the δ13C values of aquatic biofilm, and did not transfer through floodplain food webs, and hence increased
closely match with any of the terrestrial resources. This pro- food chain lengths. The marked increase in δ15N values of
vides further support for the importance of C derived from oribatid mites suggests that some of this extra fractionation
aquatic primary producers in these floodplain food webs, occurs at the base of these food webs, possibly mediated by
and points towards the capacity of organisms to select for protozoans and other micro/meiofauna that feed on bacteria
C of aquatic origin. Aquatic derived C is typically more and decaying organic matter (Griffiths 1994). The use of Bti
palatable (i.e. characterised by more labile C compounds, has led to an increase in single-celled protozoans in soils of
lower amounts of secondary plant compounds and higher Dalälven floodplains (Östman et al. 2008). This result was
nutrient content) than that arising from terrestrial leaf lit- attributed to reduced feeding by mosquito larvae on these
ter. Indeed, Lau et al. (2014) showed that autochthonous organisms (Östman et al. 2008), but might also reflect the
resources are the main driver of secondary production even ready availability of dead mosquito larvae for protozoans to
in dystrophic lakes, as algae supply consumers with fatty colonise. We, thus, hypothesise that the observed additional
acids that are essential for consumer growth and reproduc- N-fractionation for the oribatid mites in treated floodplains
tion. The higher palatability of aquatic-derived resources to reflects an increased transfer of N from decaying mosquito
terrestrial consumers is further seen in the low C:N ratios of larvae to higher trophic levels, mediated at least in part
lake biofilms. As expected, dead A. sticticus also had a low through protozoans.
C:N ratio, which is likely to make it attractive to decompos- Enchytraeid worms were significantly higher in δ13C
ers and scavengers alike. in treated relative to reference floodplains. We sampled
Enchytraeidae from the moist depressions in between grassy

13
47 Page 12 of 15 B. G. McKie et al.

tussocks characterising our floodplains. These are the habi- per treatment event, all in a relatively labile, albeit nutrient
tats where standing water is likely to persist longest after the poor, form. These figures accumulate with repeated appli-
floodwaters have receded, and hence are where both dead cations in 1 year. After ten applications, 75 kg of C and
mosquito larvae and the corn particles comprising the bulk 0.75 kg of N will have been added per hectare (though the
of the ­VectoBac®-G product are most likely to accumulate. final fate of these particles will depend on the influence of
Enchytraeidae are one of the main detritivore groups feeding subsequent winds, rainfall and flooding). Estimates of mos-
on particulate C and its associated microbes in moist soils quito biomass in treated and reference floodplains are more
(Didden 1993), and are capable of searching out patches of challenging, since we lack data on mosquito production per
more labile C within a soil matrix otherwise characterised hectare. In one of the most extreme mass emergence years
by highly refractory C (Curry and Schmidt 2007). Accord- (2000), some 192,000 individual female mosquitoes were
ingly, the increase in enchytraeid δ13C values in the treated caught from June–August in carbon dioxide traps cover-
floodplains might be especially attributable to consumption ing 99 ha of floodplains area. According to our analyses,
of the corn particles, given these are substantially richer in an adult A. sticticus contains on average 0.416 mg C and
13
C values than any other floodplain resource. The trend 0.04 mg N. Upscaling these figures and assuming a sex
for an overall reduced δ13C range in consumers of treated ratio of 1:1 (Lounibos and Escher 2008), we estimate that
floodplains suggests a shift from consumption of a broader the mosquito biomass captured in the traps in 2000 repre-
diversity of C sources towards the readily available C deliv- sented approximately 0.64 gC/ha and 0.14 gN/ha. This is, of
ered in the form of the ­VectoBac®-G corn particles, and/ course, an underestimate of emerging mosquitos, given that
or the abundant C introduced locally by the mass morality ­CO2 traps only capture a fraction of all flying adults and are
of larval mosquitoes. However, more research is required particularly inefficient at attracting males (McPhatter and
to confirm the evidence of divergence in not only the δ13C Gerry 2017). If we assume that similar-size populations of
range but also total isotopic niche area (77% probability) and mosquitoes instead had been killed as larvae, this C and N
species spread (79% probability) between floodplain types would remain concentrated in temporary pools in treated
observed in our study. floodplains. There are clearly multiple sources of uncertainty
in these calculations, including the lack of data on abun-
Quantities of ­VectoBac®‑G compared with possible dances of males, and uncertainties in the sampling efficiency
abundances of dead mosquito larvae of the carbon dioxide traps. Nevertheless, these figures serve
to demonstrate the potential impact of dead mosquitoes on
We are unable to definitively distinguish whether the shifts soil- and ground-based food webs compared with that of
in the isotopic characteristics of soil- and ground-dwelling the distribution of the corn particles in ­VectoBac®-G. Ulti-
food webs between treated and untreated floodplains of the mately, it is likely that the two effects are partially linked,
lower Dalälven are primarily attributable to the labile C as consumers that feed on dead mosquito larvae will also
introduced with ­VectoBac®-G, or the prevention of A. sticti- consume carbon from any ­VectoBac®-G remaining in the
cus emergence and the subsequent accumulation of dead mosquito’s digestive tract.
larvae on floodplain soils. An approach for separating these
mechanisms is the use of isotopic mixing models for quan-
tifying the contribution of different resources in consumer Conclusion
biomass (Layman et al. 2012; Phillips et al. 2014). However,
isotopic mixing models are prone to strong biases when the The findings of our spatially replicated field study point
isotopic range of potential resources falls well outside the towards changes in the trophic structure of the soil- and
niche space of consumers (Phillips et al. 2014), as is clearly ground-based food webs in the floodplains of the lower
the case with the corn particles comprising Vectobac-G. Dalälven, associated with the long-term, repeated appli-
Nevertheless, the question of the impact of the corn particles cations (2002–2014) of the mosquito control agent
is non-trivial, since alternative Bti products are available that ­VectoBac®-G. In a previous aquatic mesocosm experiment,
do not use corn as the substrate. Here, we explore the poten- Bti application as a suspension resulted in an increase in
tial for both mechanisms to change the isotopic composition isotopic niche areas of predatory dragonflies (Allgeier et al.
of resources in floodplain soils using data on the C and N 2019a). In contrast, the effects of Bti on our terrestrial,
content of the corn granules and mosquitoes, respectively. floodplain communities were clearest at the base of the food
Our analyses reveal that V ­ ectoBac®-G consists on aver- web. This is exemplified by the increased δ13C and δ15N
age of approximately 50% C (500 g/kg) and 0.5% (5 g/kg) values of detrivorous enchytraeid worms and oribatid moss
N. At the typical application rate of 15 kg/ha (Schäfer and mites, respectively. Additionally, the increased δ15N range
Lundström 2014), the spreading of V ­ ectoBac®-G in the characterising treated floodplain consumer communities
lower Dalälven floodplains adds 7050 gC/ha, and 75 gN/ha indicates an increase in the number of trophic fractionation

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 13 of 15 47

steps involved in the processing of nitrogen. We conjecture isotopic characteristics of most other consumer groups (i.e.
that the increase in the overall δ15N range in the treated wet- apart from Enchytraeidae and Oribatida) was weak, and the
lands arises from an extra trophic step involved in decom- isotopic niche space occupied by our untreated and treated
position of the dead mosquito larvae, which is likely to be floodplain communities substantially overlapped. Thus, we
mediated not only by the organism groups studied here, but suggest our results are best characterised as providing evi-
also by other key organisms involved in the processing of dence that the food webs of our treated floodplains might be
detritus (e.g. protozoans, nematodes and microorganisms). in transition, with effects of mosquito control manifesting
In contrast, the increase in the δ13C of enchytraeid worms unevenly amongst organism groups, trophic levels and flood-
might be more likely to reflect integration of the more plains. Further research with a greater level of replication
palatable C originating from the corn particles compos- and after more years of ­VectoBac®-G treatment is, therefore,
ing the ­VectoBac®-G product, into soil- and ground-based required to establish how extensive and established the shifts
food webs. Further research is required to assess these observed here become, and to evaluate the potential wider
hypotheses. implications for nutrient and energy cycles within floodplain
The potential ecological significance of our observed ecosystems and linkages with adjacent habitats.
shifts in isotopic values of some key consumer groups can
Supplementary Information The online version contains supplemen-
be assessed by comparing with previous studies of terres- tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 00027-0​ 23-0​ 0944-0.
trial and other invertebrates living in close association with
aquatic ecosystems. Riparian spider δ15N values were 9% Acknowledgements We are grateful to Peter Carlson for assistance in
higher in an agricultural compared with forested catchment the field, and to Tryggve Persson for advice on collection techniques
of Enchytraeidae. We thank Martina Schäfer and Jan Lundström of the
in New Zealand (Collier et al. 2002), and river restoration Swedish Biological Mosquito Control Project for advice on study site
was associated with 7–11% increases in δ15N values of ripar- location and providing us with mosquito and Bti material for analysis.
ian spiders and beetles (Kupilas et al. 2020). More marked
70–400% shifts in δ15N values were observed in detritivo- Author contributions BGM, WG and AT conceived and designed the
study. AF and TN conducted field sampling and laboratory analysis.
rous millipedes, omnivorous beetles and predatory spiders The data were analysed by BGM and AF. The first draft was prepared
living adjacent to rivers with marine-salmon-derived nutri- by BGM and WG. All authors contributed to draft revision.
ent enrichment, relative to references (Hocking and Reim-
chen 2002; Rammell et al. 2021). Finally, 50–150% increases Funding Open access funding provided by Swedish University of Agri-
cultural Sciences. This research was funded through a grant from the
in δ15N values were observed in benthic macroinvertebrates Swedish Environmental Protection Agency (Grant Number NV-04371–
along gradients of increasing anthropogenic nutrient enrich- 14) to BGM and WG.
ment in Sweden (Bergfur et al. 2009; Frainer and McKie
2015), similar to observations elsewhere (Clapcott et al. Data availability The datasets used and/or analysed during the current
study are available as supplementary material.
2010; Smucker et al. 2018). Differences in δ13C values in
the same studies ranged less widely, from 0 to 9% (anthro- Code availability Not applicable.
pogenic nutrient enrichment, Bergfur et al. 2009; Frainer and
McKie 2015), 0–4% (marine derived nutrients, Hocking and Declarations
Reimchen 2002; Rammell et al. 2021) and 0–1% (river res-
Conflict of interest The authors declared that they have no conflict of
toration, Kupilas et al. 2020). Collier et al (2002) observed interest.
a more marked 28–31% difference in δ13C values between
an agricultural compared with forested catchment in New Consent to participate Not applicable.
Zealand, which is attributable to strong differences in tree
Consent for publication Not applicable.
cover that were not a factor in our study. Finally, Tiunov and
Scheu (2004) found that enrichment of soil with refractory C
Open Access This article is licensed under a Creative Commons Attri-
(glucose) increased earthworm δ13C values by 7%. bution 4.0 International License, which permits use, sharing, adapta-
Accordingly, the 97% increase in δ15N values observed tion, distribution and reproduction in any medium or format, as long
in oribatid detritivore mites and the 3.5% increase in δ13C as you give appropriate credit to the original author(s) and the source,
observed for enchytraeid worms, fall within the mid-range provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
of shifts observed previously in studies of key environmen- included in the article's Creative Commons licence, unless indicated
tal drivers on riparian and floodplain invertebrate consum- otherwise in a credit line to the material. If material is not included in
ers. Furthermore, isotopic shifts in the range observed here the article's Creative Commons licence and your intended use is not
have previously been linked to altered nutrient and energy permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
flows, including across habit boundaries (Hladyz et al. 2011; copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
Hocking and Reimchen 2002; Kupilas et al. 2016, 2020).
On the other hand, statistical support for divergence in the

13
47 Page 14 of 15 B. G. McKie et al.

References Gratton C, Hoekman D, Dreyer J, Jackson RD (2017) Increased dura-

tion of aquatic resource pulse alters community and ecosystem
responses in a subarctic plant community. Ecology 98:2860–2872.
Allgeier S, Friedrich A, Brühl CA (2019a) Mosquito control based on
https://​doi.​org/​10.​1002/​ecy.​1977
Bacillus thuringiensis israelensis (Bti) interrupts artificial wetland
Griffiths BS (1994) Microbial-feeding nematodes and protozoa in soil:
food chains. Sci Total Environ 686:1173–1184. https://d​ oi.o​ rg/1​ 0.​
their effectson microbial activity and nitrogen mineralization in
1016/j.​scito​tenv.​2019.​05.​358
decomposition hotspots and the rhizosphere. Plant Soil 164:25–33
Allgeier S, Kästel A, Brühl CA (2019b) Adverse effects of mosquito
Handa IT et al (2014) Consequences of biodiversity loss for litter
control using Bacillus thuringiensis var. israelensis: Reduced chi-
decomposition across biomes. Nature 509:218–221
ronomid abundances in mesocosm, semi-field and field studies.
Hein T, Baranyi C, Herndl GJ, Wanek W, Schiemer F (2003) Alloch-
Ecotoxicol Environ Saf 169:786–796. https://​doi.​org/​10.​1016/j.​
thonous and autochthonous particulate organic matter in flood-
ecoenv.​2018.​11.​050
plains of the River Danube: the importance of hydrological con-
Armitage P, Cranston PS, Pinder LCV (eds) (1994) Chironomidae:
nectivity. Freshw Biol 48:220–232
biology and ecology of non-biting midges. Chapman and Hall,
Hershey AE, Lima AR, Niemi GJ, Regal RR (1998) Effects of Bacil-
London
lus thuringiensis israelensis (Bti) and methoprene on nontarget
Becker N et al (2010) Mosquitoes and their control. Springer-Verlag,
macroinvertebrates in Minnesota wetlands. Ecol Appl 8:41–60
Berlin
Hladyz S, Åbjörnsson K, Giller PS, Woodward G (2011) Impacts of
Bergfur J, Johnson RK, Sandin L, Goedkoop W (2009) Effects of nutri-
an aggressive riparian invader on community structure and eco-
ent enrichment on C and N stable isotope ratios of invertebrates,
system functioning in stream food webs. J Appl Ecol 48:443–452
fish and their food resources in boreal streams. Hydrobiologia
Hocking MD, Reimchen TE (2002) Salmon-derived nitrogen in ter-
628:67–79. https://​doi.​org/​10.​1007/​s10750-​009-​9746-4
restrial invertebrates from coniferous forests of the Pacific North-
Boisvert M, Boisvert J (2000) Effects of Bacillus thuringiensis var.
west. BMC Ecol 2:4. https://​doi.​org/​10.​1186/​1472-​6785-2-4
israelensis on target and nontarget organisms: a review of labo-
Hoekman D, Bartrons M, Gratton C (2012) Ecosystem linkages
ratory and field experiments. Biocontrol Sci Tech 10:517–561
revealed by experimental lake-derived isotope signal in heathland
Britson A, Wardrop D, Drohan P (2016) Plant community composi-
food webs. Oecologia 170:735–743
tion as a driver of decomposition dynamics in riparian wetlands.
Hoekman D, Dreyer J, Jackson RD, Townsend PA, Gratton C (2011)
Wetlands Ecol Manage 24:335–346. https://​doi.​org/​10.​1007/​
Lake to land subsidies: experimental addition of aquatic insects
s11273-​015-​9459-6
increases terrestrial arthropod densities. Ecology 92:2063–2072
Brühl CA et al (2020) Environmental and socioeconomic effects of
Holmstedt S, Linderheim A (2019) Färnebofjärden: the national park
mosquito control in Europe using the biocide Bacillus thuring-
of lower Dalälven (Färnebofjärden : nationalparken vid Nedre
iensis subsp. israelensis (Bti). Sci Total Environ 724:137800.
Dalälven). Votum & Gullers, Örebro
https://​doi.​org/​10.​1016/j.​scito​tenv.​2020.​137800
Jackson AL, Inger R, Parnell AC, Bearhop S (2011) Comparing iso-
Carlson PE, McKie BG, Sandin L, Johnson RK (2016) Strong land-
topic niche widths among and within communities: SIBER–Stable
use effects on the dispersal patterns of adult stream insects:
Isotope Bayesian Ellipses in R. J Anim Ecol 80:595–602
implications for transfers of aquatic subsidies to terrestrial con-
Kupilas B, Friberg N, McKie BG, Jochmann MA, Lorenz AW, Hering
sumers. Freshw Biol 61:848–861
D (2016) River restoration and the trophic structure of benthic
Clapcott JE, Young RG, Goodwin EO, Leathwick JR (2010)
invertebrate communities across 16 European restoration projects.
APPLIED ISSUES: exploring the response of functional indi-
Hydrobiologia 769:105–120
cators of stream health to land-use gradients. Freshw Biol
Kupilas B, McKie BG, Januschke K, Friberg N, Hering D (2020) Stable
55:2181–2199. https://​d oi.​o rg/​1 0.​1 111/j.​1 365-​2 427.​2 010.​
isotope analysis indicates positive effects of river restoration on
02463.x
aquatic-terrestrial linkages. Ecol Ind 113:106242. https://​doi.​org/​
Collier KJ, Bury S, Gibbs M (2002) A stable isotope study of linkages
10.​1016/j.​ecoli​nd.​2020.​106242
between stream and terrestrial food webs through spider preda-
Land M, Bundschuh M, Hopkins RJ, Poulin B, McKie BG (2019)
tion. Freshw Biol 47:1651–1659
What are the effects of control of mosquitoes and other nema-
Curry JP, Schmidt O (2007) The feeding ecology of earthworms - a
toceran Diptera using the microbial agent Bacillus thuringiensis
review. Pedobiologia 50:453–477
israelensis (Bti) on aquatic and terrestrial ecosystems? A system-
Demirbaş A (2005) Estimating of structural composition of wood and
atic review protocol. Environ Evid 8:32. https://​doi.​org/​10.​1186/​
non-wood biomass samples. Energy Sources 27:761–767. https://​
s13750-​019-​0175-1
doi.​org/​10.​1080/​00908​31049​04509​71
Lau DC, Sundh I, Vrede T, Pickova J, Goedkoop W (2014) Autoch-
Didden W et al (1995) The relative efficiency of wet funnel techniques
thonous resources are the main driver of consumer production in
for the extraction of Enchytraeidae. Pedobiologia 39:52–57
dystrophic boreal lakes. Ecology 95:1506–1519
Didden WAM (1993) Ecology of terrestrial Enchytraeidae. Pedobio-
Lau DCP, Vrede T, Pickova J, Goedkoop W (2012) Fatty acid com-
logia 37:2–29
position of consumers in boreal lakes – variation across species,
Frainer A, Jabiol J, Gessner MO, Bruder A, Chauvet E, McKie BG
space and time. Freshw Biol 57:24–38. https://​doi.​org/​10.​1111/j.​
(2015) Stoichiometric imbalances between detritus and detri-
1365-​2427.​2011.​02690.x
tivores are related to shifts in ecosystem functioning. Oikos
Layman CA et al (2012) Applying stable isotopes to examine food-web
125:861–871
structure: an overview of analytical tools. Biol Rev 87:545–562.
Frainer A, McKie BG (2015) Shifts in the diversity and composition
https://​doi.​org/​10.​1111/j.​1469-​185X.​2011.​00208.x
of consumer traits constrain the effects of land use on stream eco-
Lounibos LP, Escher RL (2008) Sex ratios of mosquitoes from long-
system functioning. Adv Ecol Res 52:169–200
term censuses of Florida tree holes. J Am Mosq Control Assoc
Goedkoop W, Demandt M, Ahlgren G (2007) Interactions between
24:11–15. https://​doi.​org/​10.​2987/​5656.1
food quantity and quality (long-chain polyunsaturated fatty acid
Lundstrom JO, Schafer ML, Petersson E, Vinnersten TZP, Landin J,
concentrations) effects on growth and development of Chirono-
Brodin Y (2010) Production of wetland Chironomidae (Diptera)
mus riparius. Can J Fish Aquat Sci 64:425–436. https://​doi.​org/​
and the effects of using Bacillus thuringiensis israelensis for mos-
10.​1139/​f07-​016
quito control. Bull Entomol Res 100:117–125. https://​doi.​org/​10.​
1017/​s0007​48530​99901​37

13
Ecological effects of mosquito control with Bti: evidence for shifts in the trophic structure… Page 15 of 15 47

Marty J, Planas D (2008) Comparison of methods to determine algal Schneider K, Renker C, Scheu S, Maraun M (2004) Feeding biology of
δ13C in freshwater. Limnol Oceanogr Methods 6:51–63. https://​ oribatid mites : a minireview. Phytophaga 14:247–256
doi.​org/​10.​4319/​lom.​2008.6.​51 Schneider S, Tajrin T, Lundström JO, Hendriksen NB, Melin P, Sundh I
McPhatter L, Gerry AC (2017) Effect of CO2 concentration on mos- (2017) Do multi-year applications of Bacillus thuringiensis subsp.
quito collection rate using odor-baited suction traps. J Vector Ecol israelensis for control of mosquito larvae affect the abundance of
42:44–50. https://​doi.​org/​10.​1111/​jvec.​12238 B. Cereus group populations in riparian wetland soils? Microb
Merritt RW, Dadd RH, Walker ED (1992) Feeding behavior, natural Ecol 74:901–909. https://​doi.​org/​10.​1007/​s00248-​017-​1004-0
food, and nutritional relationships of larval mosquitoes. Ann Rev Smucker NJ, Kuhn A, Cruz-Quinones CJ, Serbst JR, Lake JL (2018)
Entomol 37:349–376 Stable isotopes of algae and macroinvertebrates in streams
Mirta B, Enrih M, Tomislav B (2011) The flight distances of floodwater respond to watershed urbanization, inform management goals,
mosquitoes (Aedes vexans, Ochlerotatus sticticus and Ochlero- and indicate food web relationships. Ecol Ind 90:295–304. https://​
tatus caspius) in Osijek, Eastern Croatia. Biologia 66:678–683. doi.​org/​10.​1016/j.​ecoli​nd.​2018.​03.​024
https://​doi.​org/​10.​2478/​s11756-​011-​0073-7 Sushchik NN, Yurchenko YA, Gladyshev MI, Belevich OE, Kalachova
Mourtzinis S et al (2014) Distribution of structural carbohydrates in GS, Kolmakova AA (2013) Comparison of fatty acid contents
corn plants across the Southeastern USA. BioEnergy Res 7:551– and composition in major lipid classes of larvae and adults of
558. https://​doi.​org/​10.​1007/​s12155-​014-​9429-5 mosquitoes (Diptera: Culicidae) from a steppe region. Insect Sci
Müller-Navarra DC, Brett MT, Liston AM, Goldman CR (2000) A 20:585–600. https://​doi.​org/​10.​1111/j.​1744-​7917.​2012.​01582.x
highly unsaturated fatty acid predicts carbon transfer between Tiunov AV, Scheu S (2004) Carbon availability controls the growth of
primary producers and consumers. Nature 403:74–77 detritivores (Lumbricidae) and their effect on nitrogen mineraliza-
Östman Ö, Lundström JO, Vinnersten TZP (2008) Effects of mosquito tion. Oecologia 138:83–90
larvae removal with Bacillus thuringiensis israelensis (Bti) on Tockner K, Pennetzdorfer D, Reiner N, Schiemer F, Ward J (1999)
natural protozoan communities. Hydrobiologia 607:231–235 Hydrological connectivity, and the exchange of organic matter and
Parnell A, Jackson A (2013) siar: Stable Isotope Analysis in R. R pack- nutrients in a dynamic river–floodplain system (Danube, Austria).
age version 4.2. https://​CRAN.R-​proje​ct.​org/​packa​ge=​siar Freshw Biol 41:521–535
Petersen H (2002) General aspects of collembolan ecology at the turn Truchy A, Angeler DG, Sponseller RA, Johnson RK, McKie BG
of the millennium. Pedobiologia 46:24–260 (2015) Linking biodiversity, ecosystem functioning and services,
Phillips DL et al (2014) Best practices for use of stable isotope mixing and ecological resilience: towards an integrative framework for
models in food-web studies. Can J Zool 92:823–835. https://​doi.​ improved management. Adv Ecol Res 53:55–96
org/​10.​1139/​cjz-​2014-​0127 Tullgren A (1971) En enkel apparat för automatisk vittjande av såll-
Pointner M, Kuttner P, Obrlik A, Jäger A, Kahr H (2014) Composition gods (A simple apparatur for automatic yield of samples) Entom.
of corncobs as a substrate for fermentation of biofuels. Agron Tidskr 72:91–100
Res 12:391–396 Vander Zanden JM, Rasmussen JB (2001) Variation in δ15N and δ13C
Poulin B, Lefebvre G, Paz L (2010) Red flag for green spray: adverse trophic fractionation: Implications for aquatic food web studies.
trophic effects of Bti on breeding birds. J Appl Ecol 47:884–889 Limnnol Oceanog 46:2061–2066
Pyankov VI, Ziegler H, Akhani H, Deigele C, LÜTtge U, (2010) Euro- Vaughan IP, Newberry C, Hall DJ, Liggett JS, Ormerod SJ (2008) Eval-
pean plants with C4 photosynthesis: geographical and taxonomic uating large-scale effects of Bacillus thuringiensis var. israelensis
distribution and relations to climate parameters. Bot J Linn Soc on non-biting midges (Chironomidae) in a eutrophic urban lake.
163:283–304. https://​doi.​org/​10.​1111/j.​1095-​8339.​2010.​01062.x Freshw Biol 53:2117–2128. https://​doi.​org/​10.​1111/j.​1365-​2427.​
Ramberg E et al (2020) The structure of riparian vegetation in agri- 2008.​02043.x
cultural landscapes influences spider communities and aquatic- von Berg K, Traugott M, Scheu S (2012) Scavenging and active preda-
terrestrial linkages. Water. https://​doi.​org/​10.​3390/​w1210​2855 tion in generalist predators: a mesocosm study employing DNA-
Rammell NF, Dennert AA-O, Ernst CA-O, Reynolds JA-O (2021) based gut content analysis. Pedobiologia 55:1–5. https://​doi.​org/​
Effects of spawning Pacific salmon on terrestrial invertebrates: 10.​1016/j.​pedobi.​2011.​07.​001
Insects near spawning habitat are isotopically enriched with Yang LH, Edwards KF, Byrnes JE, Bastow JL, Wright AN, Spence KO
nitrogen-15 but display no differences in body size. Ecol Evol (2010) A meta-analysis of resource pulse-consumer interactions.
11:2045–7758 Ecol Monogr 80:125–151
Schäfer M, Lundström JO (2014) Efficiency of Bti-based floodwater
mosquito control in Sweden–four examples. J Eur Mosquito Con- Publisher's Note Springer Nature remains neutral with regard to
trol Assoc 32:1–8 jurisdictional claims in published maps and institutional affiliations.
Schäfer ML, Lundström JO (2009) The present distribution and pre-
dicted geographic expansion of the floodwater mosquito Aedes
sticticus in Sweden. J Vector Ecol 34:141–147
Scheu S, Schaefer M (1998) Bottom-up control of the soil macrofauna
community in a beechwood on limestone: manipulation of food
resources. Ecology 79:1573–1585

13

View publication stats