Skip to main content
NCBI home page
Springer logoLink to Springer
. 2026 Jan 21;35(2):36. doi: 10.1007/s10646-025-03020-5

Widespread exposure to neonicotinoid insecticide in bobcats (Lynx rufus), fishers (Pekania pennanti), and river otters (Lontra canadensis) in North Dakota, USA

Eric S Michel 1, William F Jensen 2, Charlie S Bahnson 2, Stephanie A Tucker 2, Samantha Courtney 3, Jonathan A Jenks 4, Justin Zyskowski 5, Jonathan G Lundgren 6,✉
PMCID: PMC12823708  PMID: 41565858

Abstract

Neonicotinoids are the most widely used class of insecticides and are marketed as being safe for wildlife because of their specificity to the nervous system of invertebrates. However, recent work has found widespread environmental exposure to neonicotinoids for several species, as well as an adverse effect on behavior and survival of captive white-tailed deer (Odocoileus virginianus) and other non-target species. Our objective was to measure imidacloprid in the spleens of three wild mesocarnivore species in North Dakota: bobcats (Lynx rufus), fishers (Pekania pennanti) and river otters (Lontra canadensis). We detected imidacloprid in 13% (n = 112), 15% (n = 100), and 35% (n = 100) of bobcats, fishers, and river otters with respective mean concentrations of 3.25, 4.07, and 3.33 ng/g. Mean imidacloprid concentrations found in spleens of mesocarnivores were nearly 10 times greater than the mean level of imidacloprid in spleens of captive white-tailed deer fawns that died during neonicotinoid experiments and were 5 times greater than in free-ranging deer (0.60 ng/g) in North Dakota. The probability of detecting imidacloprid was related to Julian date and peaked around 18 July for otters. The detection probability in these species increased as they increased their distance from water. Imidacloprid concentration also varied by minor watershed for river otters. Detection probability and imidacloprid concentration was not related to age, sex, Julian date, distance to water, or year for bobcats or fishers. Our findings confirm the presence of imidacloprid in vertebrate species unlikely to have direct contact with treated grain crops. This suggests that the perpetuation or maintenance of imidacloprid within food webs potentially extends well beyond the seasonal agricultural influx. Further research is needed to understand these dynamics, and their impact to animal and environmental health.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10646-025-03020-5.

Keywords: Bobcat, Fisher, Imidacloprid, Mammal, Neonicotinoids, River otter

Neonicotinoids are a class of broad-spectrum insecticide primarily used as seed treatments for major agricultural crops, but are also used in agricultural sprays, household pest management, and parasite control for domesticated animals (Goulson 2013). Neonicotinoids were first developed in the 1990s (Tokumoto et al., 2013), gained popularity from 2003 to 2011 (Douglas and Tooker 2015), and are now the most widely used class of pesticides in the world (Jeschke et al. 2011). There is a lack of reliable information on what types of pesticides are used in the USA for seed treatments and/or field applications (Hitaj et al. 2020). As of 2015, the majority of the U.S. corn crop was treated with clothianidin and thiamethoxam, respectively, while an estimated 317,500 kg of imidacloprid was used annually for seed treatments on various field crops (US EPA, 2017b). The use of these neonicotinoids steadily increased in North Dakota between 2001 and 2014 and has likely continued to grow (Dixon et al. 2021). Given the widespread use of neonicotinoids, a better understanding of the impacts to non-target wildlife species is warranted.

Neonicotinoids are water soluble and have been detected in water bodies in the United States, Canada, Australia, Europe, South Africa, and Asia (Morissey et al., 2015; Maloney et al. 2017; Marsay et al. 2025). Studies of lakes, streams, and rivers in Iowa, Minnesota, and Saskatchewan, Canada, detected one or more neonicotinoids in 91–100% of samples (Hladik et al. 2014; Main et al. 2014; Berens et al. 2021). Neonicotinoids may persist in soil for years at accumulating rates (Goulson 2013). These environmental sources may affect trophic pathways through bi- and tri-trophic interactions (Bredeson et al. 2015; Fabure et al. 2025). Collectively, these studies present several potential exposure routes for wildlife.

In addition to their documented detrimental effects on beneficial terrestrial and aquatic insects (Mogren and Lundgren 2016; Calvo-Agudo et al. 2019; Schepker et al., 2020; Fritsch et al. 2025; Mamy et al. 2025), neonicotinoids adversely affect non-target vertebrates, including white-tailed deer (Berheim et al. 2019), rats (Rattus norvegicus), mice (Mus musculus), rabbits (Sylvilagus sp.), red-legged partridges (Alectoris rufa), Nile tilapia (Oreochromis niloticus), Medaka (Oryzias latipes), and black-spotted pond frogs (Rana nigromaculata) (Gibbons et al. 2014). However, the exposure rates and health effects of these chemicals on other mammalian taxa are poorly understood. Research on the effects of neonicotinoids in mammals has predominantly focused on rodents (Costas-Ferreira and Faro 2021) with some work having been done on other mammals such as wild boar (Sus scrofa) and roe deer (Capreolus capreolus) (Kaczynski et al. 2021). At present, little is known about bioaccumulation of neonicotinoids within mammals; however, Berheim et al. (2019) documented relationships between spleen concentration and survival and health of fawns and activity of adult female and fawn white-tailed deer. To our knowledge, no information is available on potential neonicotinoid exposure in mesocarnivores that are associated with habitats embedded within major grain-producing regions of North Dakota.

Three mesocarnivores of the region with different biologies that may affect their exposure to neonicotinoids are river otters, bobcats, and fishers. Otters feed primarily on fish and crayfish in North Dakota and elsewhere (Kruuk 2007; Stearns and Serfass 2011). Stearns and Serfass (2011) found that minnows and carp (Cyprinidae), catfish (Ictaluridae), suckers (Catostomidae), and sunfish (Centrarchidae) were the most predominant fish species found in otter scats from North Dakota. Additional taxa found in the scats included insects, birds, amphibians, mammals, and freshwater mussels (Stearns and Serfass 2011). Bobcats are strict carnivores that generally prefer forests with dense understory (Hansen 2007; Seabloom 2020) but are associated with a wide variety of habitats that can provide sufficient prey, stalking cover, den sites, and resting areas free of disturbance (Hansen 2007). One study from North Dakota found that 48.6% of bobcat stomachs contained cottontails (Sylvilagus sp.), 44.4% small mammals, and 15.3% deer (Odocoileus sp.) (Trevor et al., 1989; Thomasma et al. 1994). Fishers are typically known to select mixed hardwoods with confer cover > 50% (Thomasma et al. 1994) and dense lowland conifer forest (Powell 1993) but have shown greater habitat plasticity in North Dakota where they occur primarily along deciduous, riparian forest patches (Triska et al. 2020). Fishers are generalist and opportunistic predators that consume a variety of medium and small mammals including snowshoe hare (Lepus americanus), porcupine (Erethizon dorsatum), red squirrels (Tamiasciurus hudsonicus), as well as passerine birds, deer carrion, and vegetation (Powell 1993; Martin 1994). A high proportion of the fisher’s diet is small mammals (Powell 1993; McNeil et al. 2017; Kirby et al. 2018).

Our objectives were to assess exposure rates of one neonicotinoid, imidacloprid, in spleen samples collected from bobcats (Lynx rufus), fishers (Pekania pennanti) and river otters (Lontra canadensis; hereafter, otter) in these areas. Given the apparent ubiquity of neonicotinoids, we hypothesized that the compound would be detected in these species. We further hypothesized that otters would have higher concentrations of imidacloprid than more terrestrial species (fishers and bobcats) due to the high detection rates of neonicotinoids documented in surface water. We also assessed whether the influence of age, sex, date, and year affected the incidence and severity of imidacloprid exposure.

Materials and methods

Study area

North Dakota is the least forested state in the USA (< 2% of cover type is classified as forest) and was historically comprised of primarily tall, mixed, and short grass prairie on an east-to-west gradient. Tall grass prairie is restricted to the Red River Valley, and short grass prairie is restricted to the southwestern portion of the state (Seabloom 2020). North Dakota also encompasses five drainage basins: Devils Lake (a closed basin), James River, Missouri River, Mouse River, and Red River (Fig. 1). Within these basins, the major rivers denoted by their name are fed by numerous tributaries. Since European settlement in the late 19th Century, much of the native prairie has been converted to cropland. Currently, undisturbed native grasslands compose < 25% of North Dakotas’ surface area (ND GFP, 2025). However, the level of land use conversion varies across the state. Within the tall grass prairie in the Red River basin, < 2% remains in native prairie, while > 93% of the western badlands, drained by the Little Missouri River, remains in native vegetation (ND GFP, 2025). Converted land is primarily used for crop production that is grown under monoculture conditions. The majority (65.06%) of North Dakota’s land is planted with one of three crops: wheat (Triticum aestivum; 2.69 million hectares (Mha), soybeans (Glycine spp.; 2.33 Mha), and corn (Zea mays; 0.79 Mha; USDA 2020). An average of 0.09 Mha of sugarbeets (Beta vulgaris) were planted in North Dakota in 2020; primarily in the northern Red River basin, a primary region of beet production for the nation (USDA NASS, 2020). The rest of the managed land in North Dakota is rangelands (25% of the state’s land). More than 90% of crop ground is treated with neonicotinoid insecticides (seed treatments) (Dixon et al. 2021).

Fig. 1.

Fig. 1

Open in a new tab

Distribution of bobcat, fisher and otter spleen samples collected in North Dakota (2017 to 2020). Boundaries of five drainage basins encompassed within the state denoted by bolded lines. Cross-hatched area closed to fisher trapping. The stippled area indicates the Park River watershed, which had the highest imidacloprid concentrations found in river otters

Sample collection

Regulated hunting and/or trapping for bobcats, fishers, and otters began in North Dakota in 1971, 2011, and 2017, respectively. Harvest is legal statewide with the exception of Bottineau and Rolette counties, which are closed to fisher hunting and trapping (Fig. 1). For this study, most otters and fishers were harvested in the eastern part of North Dakota, while most bobcats were harvested in the Southwest (Table 1). Bobcats were collected in 2016 (n = 2), 2017 (30), 2018 (52), and 2019 (27). Fishers were collected in 2014 (n = 22), 2015 (40), 2016 (2), 2018 (29), and 2019 (7). River otters were collected in 2016 (n = 1), 2017 (38), 2018 (32), 2019 (27), 2020 (2). Animals were collected at various distances from water, ranging from 0 to 12.01 km from water sources. More specifically, animals were collected within < 0.1 km (n = 81), 0.1 ≥ < 0.5 km (83), 0.5 ≥ < 1.0 km (61), or ≥ 1.0 km (87) from the nearest surface water. After they were skinned, the head and carcass of harvested animals were submitted to the North Dakota Game and Fish Department (NDGFD), as required by regulation (ND GFP, 2021). These carcasses, along with any other reported mortalities, were frozen and stored at -4° C until January-February of each year when they were thawed for further examination.

Table 1.

Summary of bobcat, fisher and otter sex and age (years) distribution of samples. Ages were determined via cementum analysis (Matson’s Laboratory, Manhattan, MT)

Ages (years)
Species Sex 1 2 3 > 4 Unknown Total
Bobcat Female 25 16 4 3 0 48
Male 32 15 10 7 0 64
Fisher Female 20 16 9 8 0 54
Male 21 18 4 3 0 46
Otter Female 7 10 4 7 13 41
Male 10 11 6 16 16 59
Open in a new tab

Cause of death (e.g., roadkill, hunter harvest, trapper harvest), date collected, location (Section, Township, Range / Latitude, Longitude), and sex was recorded. In addition, a canine tooth was collected from each carnivore for aging via cementum annuli (aged to the year; Matson’s Lab, Manhattan, MT) and a spleen sample was archived at -4 °C at the NDGFD Wildlife Health Laboratory. Spleens were selected for study following a recent study of white-tailed deer which revealed that neonicotinoid contents of the spleens, compared to other tissues, were most reflective of the health and fitness effects of consuming imidacloprid (Berheim et al. 2019). During late-summer and early-fall of 2020, spleens were thawed and subsections were collected. Instruments and surface areas were cleaned with ethanol between samples to minimize cross-contamination. Samples were refrozen, then sent to the Ecdysis Foundation in Estelline, South Dakota for testing in September 2020. We also analyzed 45 samples (including non-detections and detections) from all species via liquid chromatography-tandem mass spectrometry (LC-MS/MS) to validate our enzyme-linked immune-sorbent assay (ELISA) sampling.

ELISA testing

Imidacloprid levels were determined for spleens by first mincing 0.5–0.75 g of tissue using a sterilized scalpel and placing them into a polypropylene microcentrifuge tube. Water was added to the tube at a ratio of 1 mL:1 g tissue sample. Each mixture was vortexed, heated in an 80° C water bath for 10 min, then frozen at -20° C. Frozen mixtures were thawed and centrifuged at 21,130 g for 1 min and the supernatant was extracted and placed into separate microcentrifuge tubes. These were vortexed and 25 µL was extracted and placed into separate microcentrifuge tubes. These samples were each mixed with 25 µL of water, vortexed for 5 s, and centrifuged for 2 s in preparation for the ELISA assay.

Direct competitive ELISAs were run using a commercially available imidacloprid test kit (Product #500800; Abraxis LLC, Warminster, PA). Specified cross reactivities of the antibody in this kit was 121% for clothianidin, and < 0.1% for thiamethoxam (the other dominant neonicotinoids used in this region). Each sample (50 µL) was loaded into a microtiter plate that was precoated with an anti-imidacloprid primary antibody. An additional HRP conjugate antibody (50 µL) was added to each well. Plates were covered and gently agitated on an orbital shaker for 30 s and were incubated at room temperature for 60 min. Each well was washed 3 times using phosphate buffered saline, and 150 µL of 3,3’,5,5’-Tetramethylbenzidine (TMB) was added to each well. Wells were incubated for 20–25 min at room temperature, and a 100 µL of sulfuric acid was used to stop the reaction. All samples were read at 450 nm using a microplate reader (uQuant, Biotek Instruments, Winooski, VT).

Each plate had at least two standard curves of purified imidacloprid (Product number: 37894; SIGMA-ALDRICH, St. Louis, MO, USA). All spleen samples return a baseline signal due to the sample matrix. To account for this baseline absorbance, samples from the five spleens with the lowest absorbance values were homogenized into a single solution and used as a baseline for quantifying imidacloprid quantities. Furthermore, imidacloprid was mixed with baseline spleens for all controls on each plate. Stock solutions of imidacloprid were created at 0.0, 0.03, 0.06, 0.13, 0.25, 0.5, 1.0, and 2.0 ng/g of tissue. The standard curve on the ELISA plate contained 25 µL control organ in solution with 25 µL of the stock solution of imidacloprid, which encompassed eight wells with concentrations that comprised one standard curve. Our ELISAs produced consistent responses; the standard error of the mean produced in our control series was 3.43 ± 0.41% of the mean. Our limit of detection was calculated by measuring the average zero value based on our plate-specific standard curves, calculating 2.5 times the standard deviation of this value, and subtracting this correction from zero values (Berheim et al. 2019). With our baseline criteria, the detection limit for our assays were 1.67, 1.19, and 1.37 ng/g of spleen for fishers, river otters, and bobcats, respectively.

LC-MS/MS

To confirm ELISA results, fifteen samples from each species were selected for additional testing using LC-MS/MS, as has been recommended (Gross et al. 2022). Subsections from the archived spleens of these animals were collected as described above, and screened for imidacloprid via LC-MS/MS.

Tissues were prepared for LC-MS/MS analysis by sectioning 0.5 g visually free of connective and vascular structure and placing it in a 7 mL Precellys bead beater tissue homogenization tube (Bertin Technologies, Montigny-le-Bretonneux, France), containing acetonitrile (2 mL). A control tube containing 1 ng/mL of d6 imidacloprid in acetonitrile was created. The tissues were homogenized by running the following program on a Precellys Evolution Homogenizer: Speed – 5800 rpm, Cycle 2 × 20 s with a 20 s pause between cycles. The homogenized tissues were centrifuged at 4000 rpm for 10 min to pellet solid material in the bottom of the tube. A portion of the clear top layer (250 µL) from each tube was transferred to a vial for LC-MS/MS analysis. If there was visible material in the top layer, the solution was filtered through a 0.22 μm membrane before placement in the analytical vial.

Analysis was performed with a liquid chromatographic-tandem mass spectrometer. An AB Sciex 6500 + ESI-MS/MS (Framingham, MA) interfaced to the Shimadzu LC (Shimadzu Corp, Kyoto, Japan) was used. Chromatography was achieved using the Kinetex F5: 2.6 μm (100 × 3 mm) column (Phenomenex, Torrance CA) under the following conditions: the column temperature was maintained at 40° with a column heater. Injections were 10 µL in volume. The mobile phase consisted of 0.1% formic acid in Milli-Q water (mobile phase A) and 0.1% formic acid in acetonitrile (mobile phase B) at a flow rate of 0.5 mL/min. Imidacloprid and its metabolites (6-chloronicotinc acid, 5-hydroxy imidacloprid, desnitro-imidacloprid, imidacloprid-olefin, imidacloprid urea, and 2-chloro-1,3-thiazole-5-carboxylic acid) were optimized for sensitivity and selectivity through the creation of specific Multiple Reaction Monitoring (MRM) ion pairs that were specific to each compound.

Data analysis from the LC-MS/MS was performed by confirming the identity of each compound analyzed by retention time of the peak and the MRM ion transitions specific to each compound. The ratio of the peak area of the quantifier (or largest) MRM transition of each compound analyzed and that of the d6 imidacloprid internal standard was calculated and fit to a calibration curve to determine the final concentration of each component in the original tissue. In an effort to compare the contribution of the individual concentration for compound analyzed to the single result obtained by ELISA, the individual concentrations of each component determined by LC-MS/MS were converted by the binding capacity of each component in the ELISA. These converted results were summed and used in the comparison to the ELISA data.

Statistical analysis

Our first goal was to assess if imidacloprid was present in spleens of these three species and if detection and concentration varied among species. We therefore assessed if imidacloprid detection probability varied amongst species with a logistic regression model and binomial distributions with the GLM function. We also assessed if imidacloprid concentration varied by species using Analysis of Variance (ANOVA) and made subsequent species comparisons using a Tukey’s Post Hoc Test.

Our second goal was to assess if neonicotinoid detection probability differed by demographics (age and sex), distance to water, date (Julian date and its quadratic to represent timing of agricultural planting) and year for each species. We used a logistic regression model and binomial distributions with the GLM function to assess these relationships. Given otters are semiaquatic, we also assessed if detection probability differed by watershed class: major (defined drainage basins; Fig. 1) or minor (rivers and streams within the defined basins) watersheds. We included all samples (detections = 1, non-detections = 0) to assess detection probability.

Our last goal was to assess if neonicotinoid concentrations varied by demographics (age and sex), date (Julian date and its quadratic), and distance to water. We used simple linear models to assess these relationships. We also assessed if neonicotinoid concentrations varied by major or minor watersheds (otters only) and if neonicotinoid concentrations varied by year using an Analysis of Variance (ANOVA). We further investigated relationships from significant ANOVAs using Tukey’s Post Hoc Tests. Finally, we used a simple linear model to compare concentrations detected by ELISA and LC-MS/MS and included non-detections for the ELISA and LC-MS/MS as 0. We log transformed imidacloprid concentrations to achieve normality and confirmed normality by visually evaluating if model residuals were normally distributed (linear models and ANOVAs) or distributed binomially (GLMs) using the check distribution function in the random Forest package (Liaw 2002). We con ducted all analyses in Program R (R Core Team 2019) and considered relationships statistically significant at α ≤ 0.05.

Results

We tested samples from 112 bobcats, 100 fishers, and 100 otters (Table 1). Detection rates varied by species and were greatest for otters (prevalence = 35%; β = 1.24, SE. 0.34, P < 0.001) but were not different between fishers (prevalence = 15%; β = 0.13, SE = 0.39, P = 0.74) and bobcats (prevalence = 13%; β = -1.87, SE = 0.28, P < 0.001). Imidacloprid concentrations varied by species (F2, 62 = 7.88, P < 0.001) with fishers having greater concentrations than bobcats (P = 0.004) and otters (P = 0.001). There was no difference in imidacloprid concentrations between bobcats and otters (P = 0.98). Mean concentration levels and associated standard deviations of imidacloprid for bobcats, fishers, and otters were 3.25 ± 0.52, 4.07 ± 0.60, and 3.33 ± 0.79 ng/g, respectively (Table 2).

Table 2.

Mean and standard deviation of Imidacloprid concentrations (ng/g of spleen) for otter collected from minor watersheds in North Dakota, USA from 2017 to 2020. “LOD” is limit of detection. N refers to the number of otters collected in each watershed

Minor Watershed Mean Imidacloprid Concentration (ng/g of spleen) SD Imidacloprid Concentration N
Elm River < LOD 0.00 5
Devils Lake 0.91 1.42 6
Forest River 1.80 1.74 5
Goose River 0.98 1.48 9
James River < LOD . 1
Maple River 1.04 1.34 10
Park River 2.99 2.23 7
Pembina River < LOD 0.00 6
Red River < LOD 0.00 12
Sheyenne River 1.75 1.86 24
Souris River < LOD 0.00 2
Turtle River 0.68 1.51 5
Wild Rice River 2.06 1.74 8
Open in a new tab

Detection probability was positively related to distance to water (β = 0.82, SE = 0.26, P = 0.002, n = 100) and Julian date (β = -0.03, SE = 0.01, P = 0.004, n = 100; Fig. 2) but was negatively related to Julian date quadratic function (β = < -0.01, SE = < 0.01, P = 0.001, n = 100) with detection probability peaking around 18 July for otters (Table 3). Age, sex, major and minor watershed, and year were not related to detection probability for otters (P ≥ 0.22; Table 3). Detection probability was also not related to age, sex, date, distance to water, or year for bobcats (P ≥ 0.25; Supplemental Table 1) or fishers (P ≥ 0.42; Supplemental Table 2).

Fig. 2.

Fig. 2

Open in a new tab

Relationship between Julian date and imidacloprid detection probability for river otters collected from 2017 to 2020 in North Dakota, USA. The data distributions for river otters that tested positive and negative for imidacloprid were distinguished

Table 3.

Logistic regression model results assessing Imidacloprid detection probability for multiple variables in river otters collected in North Dakota, USA from 2015 to 2020

Model Variable β SE P n
Demographic Intercept -0.43 0.34 0.20 100
Age 0.06 0.08 0.43
SexMale -0.53 0.43 0.22
Major Watershed Intercept -0.69 0.87 0.42 100
James River -15.87 2399.54 0.99
Mouse River -15.87 1696.73 0.99
Red River 0.13 0.89 0.89
Minor Watershed Intercept -0.69 0.87 0.42 100
Elm River -17.87 2917.00 0.99
Forest River 1.10 1.26 0.38
Goose River 0.00 1.12 1.000
James River -17.87 6523.00 0.99
Maple River 0.29 1.08 0.79
Park River 1.61 1.20 0.18
Pembina River -17.87 2663.00 0.99
Red River -17.87 1883.00 0.99
Sheyenne River 0.69 0.96 0.47
Souris River -17.87 4612.00 0.99
Turtle River -0.69 1.41 0.62
Wild Rice River 1.20 1.13 0.29
Distance to Water Intercept -1.22 0.28 0.000 100
Distance 0.83 0.26 0.002
Date Intercept -2.61 1.21 0.03 100
Julian Date 0.04 0.01 0.004
Julian Quadratic 0.00 0.00 0.001
Year Intercept 16.57 2399.54 0.99 100
Year2017 -17.34 2399.54 0.99
Year2018 -17.21 2399.54 0.99
Year2019 -16.94 2399.54 0.99
Year2020 -33.13 2938.83 0.99
Open in a new tab

Imidacloprid concentrations in otters varied by minor watershed class (F7, 27 = 2.46, P = 0.04, n = 35), with Park River concentrations being greater than Maple River concentrations (P = 0.03, Difference = 0.045; Table 4). Imidacloprid concentrations did not vary by major watershed class (F1, 33 = 1.35, P = 0.25, n = 35). Imidacloprid concentration was not related age, sex, date, dates quadratic function, or distance to water for otters (P ≥ 0.44, n = 35; Supplemental Table 3), bobcats (P ≥ 0.33, n = 35; Supplemental Table 4), or fishers (P ≥ 0.33, n = 15; Supplemental Table 5). Imidacloprid concentrations did not differ by year for otters (F3, 31 = 0.32, P = 0.81, n = 35), bobcats (F2, 12 = 1.01, P = 0.39, n = 15), or fishers (F3, 11 = 1.26, P = 0.34, n = 15).

Table 4.

Results from the tukey’s post hoc analysis comparing Imidacloprid concentrations for otters collected from minor watersheds in North Dakota, USA from 2017 to 2020

Minor Watershed Comparison Difference P
Forest River-Devils Lake 0.07 1.000
Goose River-Devils Lake 0.07 1.000
Maple River-Devils Lake -0.06 1.000
Park River-Devils Lake 0.40 0.239
Sheyenne River-Devils Lake 0.23 0.766
Turtle River-Devils Lake 0.21 0.982
Wild Rice River-Devils Lake 0.18 0.942
Goose River-Forest River 0.00 1.000
Maple River-Forest River -0.12 0.987
Park River-Forest River 0.33 0.295
Sheyenne River-Forest River 0.16 0.893
Turtle River-Forest River 0.14 0.998
Wild Rice River-Forest River 0.11 0.991
Maple River-Goose River -0.13 0.986
Park River-Goose River 0.33 0.300
Sheyenne River-Goose River 0.16 0.897
Turtle River-Goose River 0.14 0.998
Wild Rice River-Goose River 0.11 0.992
Park River-Maple River 0.45 0.026
Sheyenne River-Maple River 0.28 0.204
Turtle River-Maple River 0.27 0.904
Wild Rice River-Maple River 0.24 0.586
Sheyenne River-Park River -0.17 0.699
Turtle River-Park River -0.18 0.985
Wild Rice River-Park River -0.22 0.626
Turtle River-Sheyenne River -0.02 1.000
Wild Rice River-Sheyenne River -0.05 1.000
Wild Rice River-Turtle River -0.03 1.000
Open in a new tab

Imidacloprid concentrations generated by ELISA and LC-MS/MS were well correlated, confirming that ELISA results could be reliably and conservatively used to assess imidacloprid status of spleen samples (Supplemental Fig. 1). LC-MS/MS analysis detected imidacloprid in 12 samples that were below the detection limit for the ELISA; two samples were below the detection limit for both analytical methods. There were no false positives produced by the ELISA method.

Discussion

Our results support our primary hypothesis that imidacloprid could be readily detected in three mesocarnivore species across North Dakota. Furthermore, exposure rates were twice as high in otters compared to fishers or bobcats, while mean concentrations detected in exposed animals were greatest for fishers. We hypothesized that otters would have the highest concentrations of imidacloprid, which was not the case. The foraging behavior, diets, and physiology of these different groups likely contributed to their relative exposure rates and concentrations within animals. Considering that neonicotinoids are water soluble (Morrissey et al. 2015), and that they are apparently non-toxic to most fish species (Gibbons et al. 2014), these chemicals might be present in fish and in turn result in chronic exposure to species like otters that rely heavily on fish for their diet. It seems noteworthy that Sacramento minnows (Cyprinidae) and Sacramento suckers (Catostomidae), members of two families of fish known to be heavily consumed by otters (Knudsen and Hale 1968; Stearns and Serfass 2011), were species found to have the highest concentrations of pesticides in fish sampled from the Russian River watershed in California (Jabusch et al. 2018). Bobcats feeding on small mammals probably ingest these prey whole, thus also consuming stomach contents. Trevor et al. (1989) also reported that voles (Microtus sp.) made up 80% of the small mammals found in bobcat stomachs. Analysis of fecal pellets of prairie voles (Microtus ochrogaster) showed that a high proportion of insects may be consumed, particularly in late summer and fall. Perhaps insects contaminated with neonicotinoids are being consumed by voles and in turn bobcats. As with bobcats, fishers may be consuming this small prey whole, including their stomach contents.

The ELISA antibodies used in this study amplified clothianidin and imidacloprid with similar efficiencies, which may have obscured the precision of our ELISA data for a particular insecticide. Clothianidin was the most frequently detected neonicotinoid in studies of surface waters in Minnesota, Iowa, and Saskatchewan and was typically found at higher concentrations compared to imidacloprid (Hladik et al. 2014; Main et al. 2014; Berens et al. 2021). The LC-MS/MS that quantified imidacloprid produced higher quantities of imidacloprid than our ELISA results; this suggests that our ELISA results are a conservative estimate of imidacloprid content of these animals’ spleens, but we cannot rule out that some of our ELISA absorbance values may have been clothianidin.

Location where the predators were collected influenced that amount of imidacloprid found in the spleens of otters. There was a significantly higher occurrence in minor watersheds versus drainage basins, and higher detection rates in some minor watersheds compared to others (e.g., Park River). One notable difference among these watersheds was mean annual stream flow. The annual mean stream flow for the Park River watershed at Grafton is 1,994 L/s, whereas other watersheds are larger (Pembina River at Neche 7,362 L/s; Red River at Fargo 22,710 L/s; Red River at Drayton 138,724 L/s) (https://www.swc.nd.gov/info_edu/map_data_resources/publishedmaps/; Accessed 7 February 2022). Flow rates were not available for the Elm River. Regardless, this suggests that increased flow may dilute imidacloprid concentrations, potentially reducing exposure. Additionally, concentrations of imidacloprid in otters were greater with increased distance from the main stem or channel of the river system. This may suggest that imidacloprid concentrations in otters were greatest in those animals living and feeding in smaller meandering streams and wetlands. Although our data are not directly comparable, the areas most impacted appear to closely mimic hotspots reported by Dixon et al. (2021) where annual total insecticide use on major crops was estimated and mapped for North Dakota.

The period of time that imidacloprid remains detectable in spleens is unknown, but assumed to be relatively short, given that imidacloprid is water soluble. The lack of association between concentration and predator age further suggests the detection of imidacloprid represents intra-annual exposure. However, other work suggests the environmental load of imidacloprid may be increasing annually. Berhiem et al. (2019) documented a 0.11 ng/g increase in imidacloprid per year in spleens of white-tailed deer collected from 2009 to 2017 in North Dakota; however, the amount used in agriculture within the state has also increased (Dixon et al. 2021).

The source of exposure to these species is also unknown. Elsewhere in North America, the highest imidacloprid concentrations in surface water were detected in May and June with steady decreases throughout the season, coinciding with planting of grain crops (Hladik et al. 2014; Main et al. 2014; Berens et al. 2021). Imidacloprid concentrations in otters was related to date, with a peak around July 1st, which corresponded to seasonal input of neonicotinoids into non-target plants (Pecenka and Lundgren 2015) (Fig. 2). This finding supports the notion that exposure in otters is linked to contaminated water resulting from early-season agricultural input. Date was not related to imidacloprid concentrations in bobcats or fishers, but the distance from water sources was significant in these species, with otters collected farther from waterways having higher levels of imidacloprid. Otter home range is 5 km from water sources (Hanrahan et al. 2019), which begs the question of how important this spatial relationship is in otter ecology. It is possible that otters that leave water sources may be more inclined to consume food sources associated with neighboring croplands. Investigating smaller waterways may be an additional important determinant of neonicotinoid exposure worthy of future study.

The role of trophic transfer and biomagnification of neonicotinoids also warrants further investigation. Among other routes, exposure to wildlife species may occur via ingestion of treated seeds or plants, contaminated soil or dust, or contaminated invertebrates and vertebrate prey items (reviewed in Roy and Chen 2023; Fabure et al., 2025). Kraus et al. (2021) documented bioaccumulation and larval-to-adult transfer of pesticides in emerging aquatic insects, providing evidence of pesticide transfer from contaminated water to terrestrial food webs. It is therefore plausible that imidacloprid exposure in three primarily carnivorous species documented here may be a result of pesticide flux that continues to the terminus of such food webs.

The health effects of sublethal imidacloprid exposure on fishers, bobcats, and river otters are also unknown. However, exposure to imidacloprid decreased fecundity and body mass in mice and affected neonate survival in white-tailed deer (Burke et al., 2018; Berheim et al. 2019). Mean imidacloprid concentrations found in spleens of mesocarnivores were nearly 10 times greater than the mean level of imidacloprid in spleens of captive white-tailed deer fawns (0.18 ng/g spleen) that died during neonicotinoid experiments and were 5 times greater than in free-ranging deer (0.60 ng/g) in North Dakota (Berheim et al. 2019). Neonicotinoids, including imidacloprid, have been widely adopted in agriculturally dominated landscapes such as North Dakota. Our results provide further evidence that this shift has led to common, frequent, and widespread exposure to many, and perhaps, nearly all animal species that cohabit such landscapes. A better understanding of the potential environmental health costs of these compounds would help to assess the long-term impacts of these chemicals on mammals inhabiting areas of high agricultural activity.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (302.6KB, docx)

Acknowledgements

Special thanks to NDGFD staff including M. Becker, J. Mortenson, J. Miller, H. Pritchert, M. Kietzman, M. Ryckman, S. Courtney, C.E. Penner, and D. Santillanez. Elizabeth Adee and Daniel Pecenka helped analyze the samples.

Author contributions

These authors contributed equally to this work.

Data availability

Data from this study is available from the Open Science Framework [https://doi.org/10.17605/OSF.IO/CWB8G] (https:/doi.org/10.17605/OSF.IO/CWB8G).

Declarations

Conflict of interest

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Berens MB, Capel PD, Arnold WA (2021) Neonicotinoid insecticides in surface water, groundwater, and wastewater across land use gradients and potential effects. Environ Toxicol Chem 40:1017–1033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Berheim E, Jenks JA, Lundgren JG, Michel ES, Grove DM, Jensen WF (2019) Effects of neonicotinoid insecticides on physiology and reproductive characteristics of captive female and Fawn white-tailed deer. Sci Rep 9:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bredeson MM, Reese RN, Lundgren JG (2015) The effects of insecticide dose and herbivore density on tri-trophic effects of Thiamethoxam in a system involving wheat, aphids, and lady beetles. Crop Prot 69:70–76 [Google Scholar]
  4. Burke AP, Niibori Y, Terayama H, Ito M, Pidgeon C, Arsenault J, Camarero PR, Cummins CL, Mateo R, Sakabe K, Hampson DR (2018) Mammalian susceptibility to a neonicotinoid insecticide after fetal and early postnatal exposure. Sci Rep. 8: 1–13
  5. Calvo-Agudo M, Gonzalez-Cabrera J, Pico Y, Calatyud-Vernich P, Urbaneja A, Dicke M, Tena A (2019) Neonicotinoids in excretion products of phloem-feeding insects kill beneficial insects. Proc Natl Acad Sci USA 116:16817–16822 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Costas-Ferreira C, Faro LRF (2021) Neurotoxic effects of neonicotinoids on mammals: what is there beyond the activation of nicotinic acetylcholine receptors? A systematic review. Int J Mol Sci 22:8413 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dixon DJ, Zheng H, Otto CRV (2021) Land conversion and pesticide use degrade forage areas for honeybees in america’s beekeeping epicenter. PLoS ONE 16:e0251043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Douglas MR, Tooker JF (2015) Large-scale deployment of seed treatments has driven rapid increase in use of neonicotinoid insecticides and preemptive pest management in U.S. Field crops. Environ Sci Technol 49:5088–5097 [DOI] [PubMed] [Google Scholar]
  9. Faburé J, Hedde M, Le Perchec S, Pesce S, Sucre E, Fritsch C (2025) Role of trophic interactions in transfer and cascading impacts of plant protection products on biodiversity: a literature review. Environ Sci Pollut Res 32:2993–3031 [Google Scholar]
  10. Fritsch C, Berny P, Crouzet O, Le Perchec S, Coeurdassier M (2025) Wildlife ecotoxicology of plant protection products: knowns and unknowns about the impacts of currently used pesticides on terrestrial vertebrate biodiversity. Environ Sci Pollut Res 32:2893–2955 [Google Scholar]
  11. Gibbons D, Morrissey C, Mineau P (2014) A review of the direct and indirect effects of neonicotinoids and fipronil on vertebrate wildlife. Environ Sci Pollut Res 22:103–118 [Google Scholar]
  12. Goulson D (2013) An overview of the environmental risks posed by neonicotinoid insecticides. J Appl Ecol 50:977–987 [Google Scholar]
  13. Gross MS, Woodward EE, Hladik ML (2022) Evaluation of ELISA for the analysis of Imidacloprid in biological matrices: cross-reactivities, matrix interferences, and comparison to LC-MS/MS. Chemosphere 286:e131746 [Google Scholar]
  14. Hanrahan AT, Rutter AU, Nielsen CK, Schauber EM (2019) Spatial ecology of river otters in a human-modified landscape. J Mammal 100:1327–1339 [Google Scholar]
  15. Hansen K (2007) Bobcat: master of survival. Oxford University Press, p 212
  16. Hitaj C, Smith DJ, Code A, Wechsler S, Esker PD, Douglas MR (2020) Sowing uncertainty: what we do and don’t know about the planting of pesticide-treated seed. BioSci 70:390–403 [Google Scholar]
  17. Hladik ML, Kolpin DW, Kuivila KM (2014) Widespread occurrence of neonicotinoid insecticides in streams in a high corn and soybean producing region, USA. Environ Pollut 193:189–196 [DOI] [PubMed] [Google Scholar]
  18. Jabusch T, Sutton R, Davis J (2018) Task 2: Contaminants of emerging concerns in sport fish from the Russian River watershed. Pages 19–65 in Maruya KA, Mehinto AC, Lao W, Sutton R, Jabusch T, Sun J, Lin D, Davis J, Fadness F, eds. Pilot Monitoring of Constituents of Emerging Concerns (CECs) in the Russian River Watershed (Region 1) Southern California Coastal Water Research Project, Technical Report 1020
  19. Jeschke P, Nauen R, Schindler M, Elbert A (2011) Overview of the status and global strategy for neonicotinoids. J Agric Food Chem 59:2897–2908 [DOI] [PubMed] [Google Scholar]
  20. Kaczynski P, Lozowicka B, Perkowski M, Zon W, Hrynko I, Rutkowska E, Skibko Z (2021) Impact of board-spectrum pesticides used in the agriculture and forestry sector on the pesticide profile in wild boar, roe deer and deer and risk assessment for venison consumers. Sci Total Environ 784:147215 [DOI] [PubMed] [Google Scholar]
  21. Kirby R, Freeh C, Gilbert JH, Olson JF, Pauli JN (2018) Poor body condition and diet in harvest population of fishers. Wildl Biol 2018: wlb.00334
  22. Knudsen GJ, Hale JB (1968) Food habits of otters in the great lakes region. J Wildl Manage 32:89–93 [Google Scholar]
  23. Kraus JM, Kuivila KM, Hladik ML, Shook N, Mushet DM, Dowdy K, Harrington R (2021) Cross-ecosystem fluxes of pesticides from prairie wetlands mediated by aquatic insect emergence: implications for terrestrial insectivores. Environ Toxicol Chem 40:2282–2296 [DOI] [PubMed] [Google Scholar]
  24. Kruuk H (2007) Otters: ecology, behaviour, and conservation. Oxford University Press, p 265
  25. Liaw, A., Wiener, M. 2002. Classification and regression by random Forest. R News 2/3: 18-22
  26. Main AR, Headley JV, Peru KM, Michel NL, Cessna AJ, Morrissey CA (2014) Widespread use and frequent detection of neonicotinoid insecticides in wetlands of canada’s prairie pothole region. PLoS ONE 9:e101400 [Google Scholar]
  27. Maloney E, Morrissey CA, Headley J, Peru K (2017) Cumulative toxicity of neonicotinoid insecticide mixtures to Chironomus Dilutus under acute exposure scenarios. Environ Toxicol Chem 36:3091–3101 [DOI] [PubMed] [Google Scholar]
  28. Mamy L, Pesce S, Sanchez W, Aviron S, Bedos C, Berny P, Bertrand C, Betoulle S, Charles S, Chaumot A, Coeurdassier M, Coutellec M-A, Crouzet O, Fabure J, Fritsch C, Gonzalez P, Hedde M, Leboulanger C, Margoum C, Mougin C, Munaron D, Nelieu S, Pelosi C, Rault M, Sucre E, Thomas M, Tournebize J, Leenhardt S (2025) Impacts of neonicotinoids on biodiversity: a critical review. Environ Sci Pollut Res 32:2794–2829 [Google Scholar]
  29. Martin SK (1994) Feeding ecology of American Martins and fishers. In: Buskirk SW, Harestad AS, Raphael MG, Powell RA (eds) Martens, sables, and fishers: biology and conservation. Cornell University Press, Ithaca, NY, USA, pp 297–315 [Google Scholar]
  30. McNeil DJ Jr, Nicks ACA, Wester JC, Larkin JL, Lovallo MJ (2017) Diets of fishers (Pekania pennanti) and evidence of intraspecific consumption in Pennsylvania. Am Midl Nat 117:200–210 [Google Scholar]
  31. Mogren CL, Lundgren JG (2016) Neonicotinoid-contaminated pollinator strips adjacent to cropland reduce honeybee nutritional status. Sci Rep 6:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Morrissey CA, Mineau P, Devries JH, Sanchez-Bayo F, Liess M, Cavallaro MC, Liber K (2015) Neonicotinoid contamination of global surface waters and associated risk to aquatic invertebrates: a review. Environ Int 74:291–303 [DOI] [PubMed] [Google Scholar]
  33. ND GFP (2021) North Dakota 2021–2022 hunting and trapping guide. North Dakota Game and Fish Department, Bismarck, p 51 [Google Scholar]
  34. ND GFP(2025) North Dakota state wildlife action plan 2025. North Dakota Game and Fish Department, Bismarck, ND, p 327 [Google Scholar]
  35. Pecenka JR, Lundgren JG (2015) Non-target effects of Clothianidin on monarch butterflies. Sci Nat 105:19 [Google Scholar]
  36. Powell RA (1993) The fisher: life history, ecology, and behavior. University of Minnesota Press, p 217
  37. R Core Team (2019) R: A language and environment for statistical computing. R foundation for statistical computing. Vienna, Austria, URL: https://www.R-project.org/. Accessed < 16 June 2021&gt
  38. Roy CL, Chen D (2023) High prevalence of neonicotinoids in sharp-tailed grouse and greater prairie-chickens across an agricultural gradient during spring and fall. 2023. Sci Total Environ. 856: 159120
  39. Schepker TJ, Webb EB, Tillitt D, LaGrange T (2020) Neonicotinoid insecticide concentrations in agricultural wetlands and associations with aquatic invertebrate communities. Agric Ecosyst Environ 287:106678 [Google Scholar]
  40. Seabloom RW (2020) The mammals of North Dakota, 2nd edn. North Dakota State University, Fargo, p 480 [Google Scholar]
  41. Stearns CR, Serfass TL (2011) Food habits and fish prey size selection of a newly colonizing population of river otters (Lontra canadensis) in Eastern North Dakota. Am Midl Nat 165:169–184 [Google Scholar]
  42. Thomasma LE, Drummer TD, Peterson RO (1994) Modeling habitat selection by fishers. In: Buskirk SW, Harestad AS, Raphael MG, Powell RA (eds) Martens, sables, and fishers: biology and conservation. Cornell University Press, Ithaca, NY, USA, pp 316–325 [Google Scholar]
  43. Tokumoto J, Danjo M, Kobayashi Y, Kinoshita K, Omotehara T, Tatsumi A, Hashiguchi M, Sekijima T, Kamizoyama H, Yokoyama T, Kitagawa H, Hoshi N (2013) Effects of exposure of Clothianidin on the reproductive system of male quails. J Vet Med Sci 75:755–760 [DOI] [PubMed] [Google Scholar]
  44. Trevor JT, Seabloom RW, Allen SH (1989) Food habits in relation to sex and age of bobcats from Southwestern North Dakota. Prairie Nat 21:163–168 [Google Scholar]
  45. Triska MD, Loughry SL, Serfass TL (2020) Fisher (Pekania pennanti) distribution in riparian forest patches of Northeastern North dakota: habitat plasticity or a short-term aberration? Can Wildl Biol Manag 9:68–80 [Google Scholar]
  46. United States Department of Agriculture National Agricultural Statistics Service (2020) NASS-Quick Stats. USDA National Agricultural Statistics Service. 2020 [cited 2021 October 22]. Available from: https://data.nal.usda.gov/dataset/nass-quick-stats
  47. United States Environmental Protection Agency (2017b) Office of Chemical Safety and Pollution Prevention. Preliminary Terrestrial Risk Assessment to Support the Registration Review of Imidacloprid. [cited 2024 March 22] Available from: https://www.regulations.gov/document/EPA-HQ-OPP-2008-0844-0140
  48. United States Environmental Protection Agency Office (2017a) Chemical Safety and Pollution Prevention. Preliminary Bee Risk Assessment to Support the Registration Review of Clothianidin and Thiamethoxam. [cited 2024 March 22] Available from: https://www.regulations.gov/document/EPA-HQ-OPP-2011-0865-0173
  49. van den Marsay M, Nimako C, Ikenaka Y, Ishizuka M, Wepender V (2025) Effects of Imidacloprid on Afrotropical aquatic ecosystems: a South African microcosm study. Int Environ Assess Manag vjaf125 10.1093/inteam/vjaf125

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (302.6KB, docx)

Data Availability Statement

Data from this study is available from the Open Science Framework [https://doi.org/10.17605/OSF.IO/CWB8G] (https:/doi.org/10.17605/OSF.IO/CWB8G).

Articles from Ecotoxicology (London, England) are provided here courtesy of Springer

RESOURCES