Agricultural Use of Insecticides Alters Homeostatic Behaviors and Cognitive Ability in Lymnaea stagnalis

Lymnaea stagnalis is an ecologically important, stress‐sensitive, freshwater mollusk that is at risk for exposure to insecticides via agricultural practices. We provide insight into the impact insecticides have on L. stagnalis by comparing specific behaviors including feeding, locomotion, shell regeneration, and cognition between snails collected at two different sites: one contaminated by insecticides and one not. We hypothesized that each of the behaviors would be altered in the insecticide‐exposed snails and that similar alterations would be induced when control snails were exposed to the contaminated environment. We found no significant differences in locomotion, feeding, and shell regeneration of insecticide‐exposed L. stagnalis compared with nonexposed individuals. Significant changes in feeding and shell repair were observed in nonexposed snails inhabiting insecticide‐contaminated pond water. Most importantly, snails maintained and trained in insecticide‐contaminated pond water did not form configural learning, but this cognitive deficit was reversed when these snails were maintained in insecticide‐free pond water. Our findings conclude that insecticides have a primarily negative impact on this higher form of cognition in L. stagnalis. Environ Toxicol Chem 2023;42:2466–2477. © 2023 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.


INTRODUCTION
Neonicotinoids and diamides are some of the most popular and widely used insecticides globally, registered in more than 120 countries, for their effectiveness in controlling certain insect populations (Jeschke et al., 2011).These chemicals are applied systemically on agricultural sites (Klingelhofer et al., 2022;Zhang et al., 2023).In laboratory studies individually applied insecticides have toxic, damaging effects on neurons in the central nervous system (CNS) of mollusks, including Lymnaea stagnalis (Goulson, 2013;Sanchez-Bayo et al., 2016;Vehovszky et al., 2015).For example, two neonicotinoids, thiacloprid and thiamethoxam, alter neuronal activity of L. stagnalis neurons because of their effects on nicotinic acetylcholine receptors (nAChRs; Vehovszky et al., 2015).In insects the neonicotinoids appear to primarily target nAChRs (Jeschke & Nauen, 2008), while the diamides target the ryanodine receptors, which regulate intracellular calcium release (Cordova et al., 2006;Lahm et al., 2005;Troczka et al., 2015).In the present study, however, our purpose was not to examine at the neuronal level the effects that a "cocktail" (i.e., thiamethoxam, clothianidin, cyantraniliprole, and chlorantraniliprole) of neonicotinoids and diamides have on Lymnaea but rather to examine some important behavioral (including a form of higher learning and memory) traits of Lymnaea.
Neonicotinoids and diamides are highly water-soluble and have long half-lives, allowing them to leach into groundwater systems and accumulate in water and soils over time (Hladik et al., 2018).Previous research has determined that neonicotinoids have significant adverse effects on insect pollinators (Muth & Leonard, 2019;Rondeau et al., 2014;Tomizawa & Casida, 2005), aquatic vertebrates (Sánchez-Bayo et al., 2016), insects (Rodrigues et al., 2017), and even songbirds (Eng et al., 2017).These chemicals, even at low levels, pose ecological risks for aquatic ecosystems (Cui et al., 2017) and are banned in certain countries.However, there exists a knowledge gap on the possible impacts that a cocktail of those chemicals in a "natural" setting has on aquatic organisms such as L. stagnalis.
Saskatchewan, a province in western Canada, is characterized as a prairie landscape containing >40% of Canada's cultivated farmland (Saskatchewan.ca, n.d.) and is also home to >100 000 lakes and streams (Tourism Saskatchewan, 2017).Because of intensive agricultural practices (i.e., pest control and chemical fertilizers), many prairie water bodies are contaminated by chemical treatments, particularly neonicotinoids and diamides.Both neonicotinoids and diamides (thiamethoxam, clothianidin, cyantraniliprole, and chlorantraniliprole; i.e., the cocktail) have been found in one of our L. stagnalis collection sites, Spirit Creek.Little is known about what impact this cocktail may have on the behavior and physiology of L. stagnalis.Our initial study serves as a baseline for further research into the impacts of these chemicals on mollusks.
The great pond snail, L. stagnalis (Linnaeus, 1758), which inhabits northern Eurasia and North America, is a large, commonly found freshwater gastropod (Fodor et al., 2020).This snail is known for its behavioral sensitivity to naturally occurring environmental stressors including calcium levels, predator detection, thermal stress (Swinton, Swinton, & Lukowiak, 2019), and possibly insecticides.These stressors alter adaptive behaviors such as learning and memory (Lukowiak et al., 2014).In Saskatchewan, naturally occurring L. stagnalis may therefore encounter contamination caused by large-scale agricultural practices, particularly insecticides and fertilizers.These chemicals may contaminate surface waters via runoff, seepage, and snow melt (Main et al., 2016;Morrissey et al., 2014).
We investigated the effects of neonicotinoids and diamides on specific homeostatic behaviors as well as shell regeneration in L. stagnalis by comparing these behaviors between snails collected at two different sites, one containing a cocktail of neonicotinoids and diamides, Spirit Creek, and another, Whitesand Lake, that has not been contaminated by these insecticides.In addition, we subjected Spirit Creek snails to a configural learning training procedure in contaminated (i.e., the cocktail) and noncontaminated pond water.Although there are many similarities between the two sites (Table 1), Spirit Creek is contaminated with neonicotinoids and diamides (Table 2) from agricultural runoff, whereas Whitesand Lake is not.This allows us to directly compare behaviors between L. stagnalis collected at the two sites and to determine if placing snails from one site into the water collected from the other site (e.g., Spirit Creek snails in Whitesand Lake water) will induce specific changes in behavior (Figure 1).Unlike laboratory-based experiments with an addition of a specific contaminant, we will use the "naturally" occurring contaminants (i.e., the cocktail) in Spirit Creek to determine if there are any specific effects of this cocktail on the snails, including whether a higher form of learning, configural learning, is altered.Previously it has been demonstrated that freshly collected snails such as the Whitesand Lake snails are competent to form configural learning memory when they experience food (e.g., carrot slurry) together with crayfish effluent (Batabyal et al., 2021;Swinton, Swinton, Shymansky et al., 2019).Configural learning is a form of higher-order learning where the organism forms an association between two stimuli experienced together that is different from the simple sum of their components (Giurfa, 2003).Lymnaea form configural learning when they experience food (e.g., carrot odor) together with a predator scent (crayfish effluent; Swinton, Swinton, & Lukowiak, 2019).Following configural learning (i.e., carrot odor+crayfish effluent), the carrot odor no longer elicits a feeding response and instead elicits a fear state.Thus, altering this ability (i.e., the establishment of a fear state) in snails could have major consequences.
We studied the following homeostatic behaviors in L. stagnalis: shell growth, locomotion, feeding behavior, and configural learning (i.e., the cognitive assay).We chose behaviors such as locomotion and feeding because they are important for survival and shell regeneration, which also has survival benefits.Mollusks, including L. stagnalis, can repair the damage inflicted on their shell (Swinton, Swinton, & Lukowiak, 2019).Configural learning is a higher-order associative type of learning that shows the decision-making  abilities of the animals.We hypothesize that in Spirit Creek pond water locomotion, feeding, and shell repair will be significantly reduced in both Spirit Creek and Whitesand Lake snails.We further hypothesize that these homeostatic behaviors will recover when both populations of snails are placed in our laboratory-created artificial pond water.Further, we hypothesize that cognitive ability (i.e., learning and memory) will be significantly reduced in snails when they are experiencing Spirit Creek pond water.Finally, we hypothesize that Spirit Creek snails when they are trained and tested in insecticide-free laboratory-made pond water will recover their cognitive ability so that configural learning can occur.
The detailed questions addressed in the present study along with the experimental groups are illustrated in a schematic in Figure 1.

Study sites
Lymnaea stagnalis are found across Saskatchewan, in slowmoving streams, ponds, and lakes.The two collection sites used in the present study are Spirit Creek and Whitesand Lake, located approximately 50 km from one another within the Assiniboine watershed in Saskatchewan, Canada (Whitesand Lake, 51°46′12.45″N,103°21′14.16″W;Spirit Creek, 51°45′ 9.62″N, 102°58′12.79″W).This area is in the aspen parkland ecoregion that has a climate characteristic of long, cold winters and short, warm summers (Padbury et al., 1998).
Spirit Creek is a stream that begins near Rama, Saskatchewan, and flows eastward, ultimately draining into Good Spirit Lake.This site is protected by the Nature Conservancy of Canada because of its high importance for the breeding of waterfowl (Nature Conservancy of Canada, 2018).The collection site along the creek has low flow with high levels of vegetation and an abundant L. stagnalis population.We have not observed native crayfish predators (Faxonius virilis) in this creek.However, preliminary experiments show that the Spirit Creek L. stagnalis are predator-experienced to kairomones released from crayfish.That is, detecting the kairomones released by crayfish elicits a range of antipredator behaviors in the snails (Orr et al., 2007).This is not surprising because this species of crayfish is invading ponds in this watershed much farther to the west.Spirit Creek provides drainage for approximately 1000 km 2 within the Good Spirit Lake drainage basin (Spirit Creek Watershed Monitoring Committee, 2006).Spirit Creek is at high risk for contamination by natural fertilizers and pesticides because 55% of this drainage basin is cropland.Data provided by the Water Security Agency collected in the 2017 to 2019 summer seasons demonstrate that Spirit Creek has been contaminated by multiple insecticidescyantraniliprole, chlorantraniliprole, thiamethoxam, and clothianidin (Table 2)-that persist through the summer months.
The data in Table 2 were collected by the Saskatchewan Water Security Agency from the samples collected at Spirit Creek and are comparable to values across the Prairie Pothole Region of North America (Main et al., 2014).The analyses were performed at the Agriculture and Agri-Food Canada research station in Lethbridge, Alberta.The methods are as follows: Twelve neonicotinoid insecticides in water are analyzed simultaneously using aqueous solid-phase extraction and liquid chromatography-tandem mass spectrometry (LC-MS/MS; adapted from Xie et al., 2011).Analytical standards of imidacloprid, clothianidin, thiamethoxam, glothianidin, and acetamiprid in 200 μL of the water sample are injected into the LC-MS/MS; and the compounds are screened by retention time and at least two characteristic ions.Samples are quantified using a standard curve containing at least seven points, and the limit of quantification (LOQ) was the smallest point.The limits of detection (LODs) were as follows: thiamethoxam 2.7 ng L -1 , clothianidin 5.4 ng L -1 , and imidacloprid 3.2 ng L -1 ; for all other compounds the LOD is one-half the LOQ.
On the other hand, Whitesand Lake is a small lake (~5 km 2 ), located 9 km south of the village of Margo, Saskatchewan.It has a significant population of L. stagnalis occurring alongside crayfish and freshwater fish (Hughes et al., 2017;Prestie et al., 2019;Shymansky et al., 2017).Whitesand Lake (elevation 525 m) is the headwaters of a small stream that connects a series of small lakes.This stream (sometimes called the Whitesand River) ultimately flows into the Assiniboine River near Kamsack, Saskatchewan (51°33′54″N, 101°53′41″W).It is not directly connected to Spirit Creek, although that stream also ultimately flows into the Assiniboine River, which ultimately joins the Red River in Winnipeg, Manitoba.It is the Assiniboine River system that has been the "pathway" by which crayfish have invaded the ponds where we collect snails (Batabyal & Lukowiak, 2021).
Whitesand Lake L. stagnalis are predator-experienced (to crayfish) and exhibit the so-called smart cognitive phenotype, which means they show enhanced long-term memory ability (Hughes et al., 2017).During the time of the present study, data provided by the Saskatchewan Water Security Agency indicate that there is no presence of neonicotinoids in Whitesand Lake.

Collection and housing
Approximately 300 freshly collected, wild snails were used in the present study; they were collected from both Spirit Creek and Whitesand Lake (~150 snails per site) between June and August of 2019 and transported live to the laboratory along with many carboys (20 L) of pond water from each site.Clear carboys were filled approximately 10 m from shore at each collection site, and the number of carboys collected varied each day between two and eight per site.The collected water was kept in a low-light, room-temperature environment (~20-22 °C).Snails were kept in four treatment groups of separate aquaria (see Cross-housing experiments, below), and within each treatment group the snails were organized by size cohort into three separate aquaria equipped with commercial filters filled with bioballs and aerators.Snails were kept in these aquaria from the day of collection until the end of the study (~60 days).The length of the snail's shell was measured and organized into one of the three size cohorts, which are as follows: small = <1.4cm, medium = 1.5-2.9cm, and large = >3.0cm.These size cohorts are to some degree representative of the age of the snails, assuming maturity with increased shell size.All guidelines for animal welfare and research outlined by the University Animal Care Committee were followed extensively to ensure the well-being of all animals used in our studies.
In addition, a cohort of Spirit Creek snails was transported to the Lukowiak laboratory at the University of Calgary, where they were maintained either in Lukowiak-laboratory pond water (see below) or Spirit Creek pond water transported to Calgary along with the snails to perform the configural experiments in the same carboys used for collection as mentioned previously.

Cross-housing experiments
To study the impacts of agriculture-based levels of insecticides, we performed cross-housing experiments (see Figure 1 for detailed experimental details).That is, we placed Whitesand Lake snails into Spirit Creek water and vice versa.Thus, 30 Whitesand Lake L. stagnalis were placed in an aquarium containing insecticide-contaminated water collected from Spirit Creek (Whitesand Lake snails in Spirit Creek pond water), and 30 Spirit Creek snails were placed in an aquarium containing Whitesand Lake water (Spirit Creek snails in Whitesand Lake pond water).Two control aquaria containing 30 Whitesand Lake and 30 Spirit Creek L. stagnalis and pond water from their respective sites (i.e., Whitesand Lake snails in Whitesand Lake pond water and Spirit Creek snails in Spirit Creek pond water) were used as controls.The individuals in these four aquaria were utilized in the feeding and shell regeneration experiments.Snails in all four aquaria were fed a diet of romaine lettuce (ad libitum).The lettuce was thoroughly washed and purchased from multiple grocery stores over the course of our study.All of these snails were maintained on a summer 17:7-h light:dark schedule at room temperature (~20 °C).
In a similar manner, a cohort of Spirit Creek snails was divided into three groups for the configural learning study.One was housed in Lukowiak-laboratory pond water, another was housed in water from Spirit Creek, while the third group was initially maintained in Lukowiak-laboratory pond water (see below) for 5 days before being returned to Spirit Creek water for 5 days before the configural learning training procedure.These snails were also maintained on the same summer-light schedule, and the temperature in the Lukowiak was also maintained at 20 °C.To make Lukowiak-laboratory pond water, 0.25 g L −1 of Instant Ocean (Spectrum Brands) is dissolved in deionized water, supplemented with calcium sulfate dihydrate to create a standard calcium level of 80 mg L −1 .In addition, they were fed romaine lettuce ad libitum.

Locomotion
The ability of L. stagnalis to maneuver throughout its environment was examined for both the Spirit Creek and Whitesand Lake snails to determine if it was impaired by the presence of insecticides.Locomotion of L. stagnalis was determined by recording the distance traveled over a grid within a 3-min time interval.A 0.6-cm 2 grid was placed on the bottom of a tray filled with enough water from the snail's respective environment to evenly cover the bottom and allow for movement (~150 mL).The snail was placed at a 2.54-cm 2 start point in the middle of the grid, and the number of individual squares the snail crossed in any direction was recorded.Velocity was calculated by dividing the number of squares crossed by the time in seconds (centimeters per second).

Carrot slurry
The carrot slurry acts as a chemosensory cue and induces feeding in L. stagnalis.Thus, a carrot slurry is utilized in our feeding experiments.The carrot slurry was created by blending approximately one large (two to three medium or four to seven small; i.e., ~600 g) organic, peeled carrot with the ends discarded, with 475 mL of the appropriately collected pond water (either Spirit Creek or Whitesand Lake pond water).The mixture was strenuously blended for 3-5 min with a standard Hamilton Beach ® kitchen blender and strained through a fine sieve to eliminate any suspended solids or particles.Fresh organic carrots were obtained from various local grocery stores.

Rasping behavior
Feeding (i.e., rasping) in L. stagnalis is a motor behavior that consists of repetitive movements called rasps.During a rasp, the mouth of the snail opens, and its toothed radula scrapes over the food substrate, which is then lifted into the mouth that closes as the substrate is swallowed (Elliott & Benjamin, 1989).This sequence is then repeated.Snails can be observed to rasp in water or any liquid medium which has suspended food particles such as lettuce or carrot slurry.To measure rasping, L. stagnalis were placed individually in a Petri dish filled with enough water from the snail's respective area, Whitesand Lake or Spirit Creek, to cover the bottom and allow for movement.The snail was left to acclimate in the dish for 10 min.The dish was then placed on top of rubber stoppers atop a mirror to view the snail's mouth.The number of rasps was observed and recorded within a 2-min time interval.The snail was then transferred to another Petri dish containing the carrot slurry (outlined above) and given a 10-min acclimation period.After this acclimation period, the number of rasps was observed and recorded within a 2-min interval before returning the snails to their respective housing tanks.

Configural learning experiments
Configural learning is a form of higher-order learning where the organism forms an association between two stimuli experienced together that is different from the simple sum of their components (Giurfa, 2003).Lymnaea form configural learning when they experience food (e.g., carrot odor) together with a predator scent (crayfish effluent; Swinton, Swinton, Shymansky et al., 2019).Following configural learning (i.e., carrot odor+crayfish effluent), the carrot odor no longer elicits a feeding response and instead elicits a fear state.The carrot odor+crayfish effluent paradigm that was used in the previous configural learning experiments for Lymnaea involves a cohort of snails being housed together before and after the carrot odor+crayfish effluent pairing and snails all being in the same beaker during the carrot odor+crayfish effluent pairing.We followed the configural learning procedure outlined for freshly collected snails (Kagan & Lukowiak, 2019) for Spirit Creek snails in both Spirit Creek water and Lukowiak laboratory pond water.Briefly, the configural learning procedure is as follows: The spontaneous rasping rate of the snails in pond water was measured, and then 3 h later we measured their rasping rate to the carrot slurry (labeled as C pre on graphs).The next day (~18-20 h) snails were exposed simultaneously to both the carrot slurry made using crayfish effluent instead of pond water for 45 min.We term this the configural learning training session.Three hours after the configural learning training session, the rasping response to the carrot slurry was measured again (labeled as C 3 h in the graphs).

Shell regeneration
Crayfish are one of the primary predators of L. stagnalis, which use their smaller pereiopods to chip away the snail's shell and then attempt to extract the snail from its shell (Swinton, Swinton, & Lukowiak, 2019).Shell damage in L. stagnalis is typically caused by unsuccessful predation attempts and stimulates a physiological response to repair the area by shell regeneration.Ten medium to large (1.5 cm-3.0 cm) snails from each separate cross-housing aquaria environment (e.g., Whitesand Lake snails in Spirit Creek pond water) were used.The snail was labeled using superglue and waterproof paper, and the snail's size was measured.Using forceps, a small fracture was made in the mantle of the snail's shell on the side of its pneumostome.This fracture was measured, and photos of the fractures were taken for reference; however, a technological error deemed the files unusable.The fractures of each of the snails were measured 2, 5, and 8 days after the initial fracture to measure the progress of healing and snail condition.

Statistics
Locomotion results were assessed by performing a Mann-Whitney rank sum test between the two control groups, Spirit Creek snails and Whitesand Lake snails in their home pond water to determine any statistical significance between the two groups.The rasping and shell regeneration experiments made across the four different treatment groups (Whitesand Lake snails in their home pond water, Spirit Creek snails in their home pond water, Whitesand Lake snails in Spirit Creek pond water, and Spirit Creek snails in Whitesand Lake pond water) were analyzed using two-way repeated measures analyses of variance (ANOVAs), followed by a Tukey's post hoc test.The configural learning data across the three groups of Spirit Creek snails were analyzed using separate one-way repeated measures ANOVAs, followed by a Tukey's multiple comparison post hoc test.Significance was set at p < 0.05.All analyses were performed using Graphpad Prism 9.0.1 for Mac.

Locomotion
The results of the locomotion studies (Figure 2) showed no significant difference (Mann-Whitney U = 1060, p = 0.111) between the Whitesand Lake snails and the Spirit Creek snails after a Mann-Whitney rank sum test.The mean velocity of 44 Spirit Creek snails with a size ranging between 1.5 cm and 3.2 cm was found to be 0.038 cm s -1 , while for 59 Whitesand Lake snails, with sizes ranging between 1.5 and 3.2 cm, the mean velocity was 0.048 cm s -1 .Locomotion data for Spirit Creek snails in Whitesand Lake pond water and Whitesand Lake snails in Spirit Creek pond water were not determined, to reduce the number of snails used because of the lack of difference and significance between the control trials.
Feeding behavior.The two-way ANOVA showed a significant interaction effect of the different treatment groups and the two stimuli (pond water and carrot slurry; F 3,57 = 5.51, p = 0.002) on rasping behavior (Figure 3).In each of the four groups of snails there is a significant increase in rasping, from spontaneous rasping (in pond water) with exposure to the carrot slurry (Whitesand Lake snails, t = 4.55, p < 0.001; Spirit Creek snails, t = 7.36, p < 0.001; Whitesand Lake snails in Spirit Creek pond water, t = 8.02, p < 0.001; Spirit Creek snails in Whitesand Lake pond water, t = 6.06, p < 0.001; Figure 3A,B).Specifically, the Whitesand Lake snails exhibited a large range of activity.Published data (Sugai et al., 2006) show a variance in rasping during the carrot slurry trial for snails used in that Japanese study, from 0 to approximately 22 rasps per minute, which is similar to the results we obtained in the present study.Whitesand Lake snails in Spirit Creek pond water had a higher rasp rate in carrot slurry compared to the Whitesand Lake control group (t = 4.40, p < 0.001; Figure 3A).Spirit Creek snails from the control group as well as Spirit Creek snails in Whitesand Lake pond water had a similar rasping rate in carrot slurry (p > 0.05; Figure 3B).

Shell regeneration
The data for shell regeneration are presented in Figure 4.The two-way ANOVA showed a significant effect of the two main factors (time, initial and final, F 1,68 = 14.02, p = 0.0004; different treatment groups, F 3,68 = 4.71, p = 0.004) but not the interaction effect (F 3,68 = 1.73, p = 0.167).These data showed that snails from Whitesand Lake in Spirit Creek pond water had the greatest repair success.We observed five individuals repair their fractures just 5 days after the damage occurred.Whitesand Lake snails in Spirit Creek pond water was the only group to show a statistically significant change (t = 3.46, p = 0.003; Figure 4A) from the initial fracture to the final measurement.For Whitesand Lake snails in Spirit Creek pond water, we observed one mortality occurring 5 days after the initial fracture.Whitesand Lake snails had the most variable results and the least success in shell regeneration.In multiple cases for Whitesand Lake snails, the fractures had incurred further damage between the measuring days and were larger in size than the initial fracture 8 days later.There were minimal successful signs of damage repair for Whitesand Lake snails 8 days after the damage was inflicted (t = 1.21, p = 0.645; Figure 4A).Spirit Creek snails had three individuals successfully repair their damage within the time; however, the repaired areas were brittle, transparent, and not hardened.We assume that the condition of these repairs would improve with time (t = 2.53, p = 0.053; Figure 4B).Spirit Creek snails in Whitesand Lake pond water were observed to slowly repair their shell damage; however, no individuals from the trials had completely repaired their damage after 8 days (t = 0.39, p = 0.991; Figure 4B).In Spirit Creek snails in Whitesand Lake pond water there were three mortalities, occurring 5 days after the fractures were made.

Configural learning
We hypothesized that the Spirit Creek snails if maintained and trained as in Spirit Creek pond water would be competent to show configural learning memory as with a previous study looking at configural learning and memory in freshly collected snails (Kagan & Lukowiak, 2019).In that study, the freshly collected strains were maintained in the Lukowiak-laboratory pond water (see Methods) for at least 5 days before being trained and tested in that pond water.We studied three different Spirit Creek snail cohorts (Figure 5).The first cohort (n = 9; Figure 5A) was maintained and then trained in Spirit Creek pond water.An ANOVA followed by a Tukey's post hoc test revealed the following.There were significant differences in the feeding responses in this experiment (ANOVA, F (1.106,7.743)= 30.69,p = 0.0005).There was a significant increase in the rasping rate between pond water and their initial response to the carrot slurry (pond water vs. carrot odor pre; p = 0.0010).Three hours following the configural learning training procedure the response to the carrot slurry (3 h compared to carrot odor pre) following the configural learning training procedure was not significantly different (p = 0.0884).Importantly, in this cohort maintained in Spirit Creek water, configural learning memory was not seen.That is, the feeding response to the carrot slurry following the configural learning training procedure was not significantly reduced compared with the response elicited by the carrot slurry before the training procedure.
In the second cohort (n = 11; Figure 5B) Spirit Creek snails were first maintained for 5 days in Lukowiak-laboratory pond water before being trained in the same water.An ANOVA followed by a Tukey's post hoc test revealed the following.There were significant differences in the feeding responses in this experiment (ANOVA, F (1.813,18.13)= 57.41,p < 0.0001).There was a significant increase in the rasping rate between pond water and their initial response to the carrot slurry (pond water vs. carrot odor pre; p < 0.0001).Three hours following the configural learning training procedure the response to the carrot slurry (3 h compared to carrot odor pre) was significantly less (p = 0.0003).Thus, in this cohort configural learning memory formed.That is, the snails remembered that the carrot slurry now signaled predator presence.
In the final cohort (n = 10; Figure 5C), first maintained in Lukowiak-laboratory pond water and then maintained and trained in Spirit Creek pond water, the configural learning training procedure was again employed.An ANOVA followed by a Tukey's post hoc test revealed the following.There were significant differences in the feeding responses in this experiment (ANOVA, F (1.799,16.19)= 43.39,p < 0.0001).There was a significant increase in the rasping rate between pond water and their initial response to the carrot slurry (pond water vs. carrot odor pre; p < 0.0001).Three hours following the configural learning training procedure the response to the carrot slurry (3 h compared to carrot odor pre) was not significantly different (p = 0.1461).Thus, in this cohort configural learning did not occur.The response to carrot following the configural learning training procedure was not significantly reduced compared with the response elicited by the carrot slurry before the training procedure.Thus, training in Spirit Creek pond water resulted in snails not forming a configural learning memory.

DISCUSSION
Insecticides and fertilizers can leach into freshwater systems via surface runoff and thus could pose a significant threat to these systems in areas of high agricultural activity.The present study serves as a baseline for outlining the potential impacts that neonicotinoids and diamides have on some important homeostatic behaviors and processes (shell repair and cognition) in an aquatic organism, L. stagnalis.By comparing individuals collected from an insecticide-contaminated environment (i.e., Spirit Creek snails) to those from an insecticide-free environment (Whitesand Lake snails) we can analyze the impacts of "naturally occurring" concentrations and combinations of insecticides.We aim to determine differences in behaviors and physiologies of L. stagnalis from the two environments and if these alterations can be induced with exposure to insecticide-containing pond water over a relatively short time.

Locomotion
The locomotion of L. stagnalis has been shown to be altered by drops in temperature, low-calcium environments, and predator presence (Dalesman & Lukowiak, 2010;Orr et al., 2007).To our knowledge, no studies have examined the relationship between a cocktail of "naturally occurring" insecticides and movement in L. stagnalis prior to the present study.Previous studies have shown that insecticides do not significantly alter locomotion, despite having significant neurological and reproductive impacts (Ihara & Matsuda, 2018;Lavtiža et al., 2016;Liu et al., 2018;Smagghe et al., 2019).A study conducted by Rodrigues et al. (2016) found that chlorantraniliprole, a neonicotinoid also found in Spirit Creek, had an impact on the locomotor activities of the freshwater planarian Dugesia; however, this type of movement and organism differs substantially from those of L. stagnalis.
The purpose of this experiment was to outline if the neonicotinoid cocktail in Spirit Creek pond water impacted the locomotion of L. stagnalis.The results of our study demonstrated no significant difference in velocity between Spirit Creek and Whitesand Lake snails, indicating that this "cocktail" had no discernible impact on the snail's movement.Thus, we conclude that while this insecticide cocktail may alter other behaviors in Lymnaea, locomotory behavior was not significantly impacted.

Rasping
Because feeding is essential for survival, we needed to determine if the "cocktail" altered this essential homeostatic behavior.We hypothesized that Spirit Creek Lymnaea would have a higher rasping rate than Whitesand Lake Lymnaea because of an increased energy requirement due to the presence of insecticides.As expected, in the presence of the carrot slurry (i.e., carrot odor) there was a significant increase in the rasping rate between the spontaneous level (i.e., that occurring in the absence of a food stimulus) and the rate elicited by the carrot slurry in both Whitesand Lake and Spirit Creek snails in both Whitesand Lake and Spirit Creek pond water.What was apparent is that the Whitesand Lake snails in Spirit Creek pond water had many more high responders than Spirit Creek snails in Spirit Creek pond water.In the carrot slurry, however, this is not as apparent.If we focus on the Whitesand Lake snails, we found that their response to the carrot slurry was significantly greater in Spirit Creek pond water than it was in Whitesand Lake pond water.This was not the case with the Spirit Creek snails; their response to the carrot slurry was similar in both Whitesand Lake and Spirit Creek pond water.We speculate that the response difference seen in the Whitesand Lake snails may relate to insecticide-driven stress that served to increase the feeding rate.Because the Spirit Creek snails had adapted to this stressor, we did not see a difference in feeding between the two pond waters.
It was beyond the scope of our initial study to determine if a combination of stressors such as heat (i.e., as a consequence of climate change) and an insecticide cocktail would have even greater effects on feeding.These experiments will be initiated in the next field season.

Shell regeneration
For L. stagnalis and other mollusks, the shell is essential for survival, and thus repairing damage to reestablish its integrity should be of high priority (Blundon & Vermeij, 1983).Injury to the shell promotes a physiological response to repair the damaged area, typically at the expense of normal shell growth (Kunigelis & Saleuddin, 1982).We hypothesized that in the Spirit Creek cocktail, snails would have a slower shell repair rate.However, contrary to our hypothesis, we observed rapid shell repair rates and reductions in the overall size of the shell fractures in Spirit Creek and Whitesand Lake snails in Spirit Creek pond water.In addition, the only significant difference was observed in the Whitesand Lake snails in Spirit Creek pond water.However, we only measured an 8day period, and a longer observational time may have yielded different results.
Surprisingly, our data show a depressed shell repair rate with no significant changes between the initial and final measurements in the snails housed in Whitesand Lake pond water compared with those housed in Spirit Creek pond water.Our first thought was that there might be a difference in ambient calcium levels between Whitesand Lake and Spirit Creek pond waters.Low environmental calcium levels (i.e., <50 mg L −1 ) impact important homeostatic behaviors of L. stagnalis (Dalesman & Lukowiak, 2010;Swinton, Swinton, & Lukowiak, 2019) and cognition (Knezevic et al., 2011(Knezevic et al., , 2016)).Differences in calcium content of the pond waters might alter shell repair.However, as shown in Table 1, both sites had high calcium levels during the time of the present study (April-August 2019).What we did observe, however, was that the Whitesand Lake snails incurred spontaneous additional damage.That is, between measurements their fracture size grew beyond what was initially measured.We are uncertain as to why or how this happened.

Configural learning and memory formation
In the configural learning training procedure snails are exposed simultaneously to the carrot slurry made by using crayfish effluent for 45 min.This is the configural learning training procedure (Swinton, Swinton, Shymansky et al., 2019).We compare the rate of rasping to the carrot slurry before the configural learning training procedure and then 3 h after the training procedure.It was suggested (Batabyal et al., 2021;Kagan & Lukowiak, 2019;Swinton, Swinton, Shymansky et al., 2019) that the configural learning training procedure creates a "landscape of fear" in the snails, such that the snails now no longer increase their rasping rate to the carrot slurry.
Previously it has been shown that freshly collected snails from Whitesand Lake were capable of configural learning and memory formation (Kagan & Lukowiak, 2019).In that study, however, the freshly collected snails were housed in pond water made in the Lukowiak laboratory for at least 5 days before undergoing configural learning training.In the present study, we show that Spirit Creek snails, maintained and trained in Spirit Creek pond water, were not capable of forming such a memory.However, the Spirit Creek snails could successfully learn and form a configural learning memory if they were maintained and trained in Lukowiak-laboratory pond water.Thus, while in Spirit Creek pond water they were incapable of showing such cognitive ability, this deficit could be overcome by maintaining and training them in pesticide-free pond water.That is, the insecticide-containing Spirit Creek pond water, and possibly other agricultural runoff contaminants, was not conducive for this higher form of associative learning.This cognitive deficit occurred even after Spirit Creek snails were maintained in Lukowiak-laboratory pond water for an initial 5 days but were then placed back into Spirit Creek pond water.We interpret these data to suggest that the presence of the insecticide cocktail in the Spirit Creek pond water prevents this form of cognition.Thus, the Spirit Creek pond water insecticide cocktail has negative consequences on the cognitive ability of the snails.It remains to be investigated what neuronal mechanisms are altered in the CNS of L. stagnalis that are exposed to Spirit Creek pond water.We speculate that essential second messenger systems such as cyclic adenosine monophosphate response element-binding protein might be affected by the cocktail of insecticides occurring in Spirit Creek pond water.

CONCLUSION
Neonicotinoids and diamides which are known as "systemic insecticides," are widely used in agriculture across the globe.These insecticides are persistent in the environment and have high leaching potential, thus posing toxicity risks to various aquatic organisms.They can disrupt the food chain by blocking nutrient supply, affecting aquatic species directly or indirectly.
Our study shows how a mix of such insecticides that is present in a natural pond can have sublethal effects occurring even over a short timescale.Lymnaea stagnalis is an ecologically important freshwater organism that has been used as a model system in many different types of studies (Rivi et al., 2021) that is at high risk for exposure to various contaminants within Saskatchewan, Canada.Understanding the threshold levels of neonicotinoids in water is crucial to mitigate their persistent effects on such aquatic communities.However, research on the adverse effects of neonicotinoids is more limited for aquatic vertebrates compared with aquatic invertebrates, hindering our comprehension of their mechanisms of action.Because aquatic vertebrates depend on aquatic invertebrates for food, it is essential to study the effects of neonicotinoids on invertebrate models to establish protective guidelines for the ecosystem.The scarcity of research on the indirect effects caused by these insecticides is more significant than their direct toxicity on model organisms.The impacts of neonicotinoids and diamides on aquatic species, including L. stagnalis, are thus extremely important and serve as a starting point for further research into this area.Addressing information gaps will aid in understanding the regulatory mechanisms of such insecticides, leading to well-informed guidelines for safeguarding the aquatic ecosystem.

FIGURE 1 :
FIGURE 1: A detailed schematic is provided for the four broad questions addressed in the present study.Details of the experimental groups and sample sizes are listed for each experiment.Darker-shelled animals represent Spirit Creek snails, and lighter-shade animals represent Whitesand Lake snails.

FIGURE 2 :
FIGURE 2: Velocity of individual Lymnaea stagnalis within the locomotion trials for both Spirit Creek (n = 44) and Whitesand Lake (n = 59) snails.No significant difference in velocity was observed between the two populations.The graphical representation shows the mean, and the error bars are the SEM.ns = not significant as p > 0.05; WS = Whitesand Lake; SC = Spirit Creek.

FIGURE 3 :
FIGURE 3: Feeding experiment comparing basal rasping in pond water and carrot effluent in each of the four experimental groups: (A) Whitesand Lake (WSL) snails in WSL pond water control (WSL in WSL water, n = 20) and WSL snails in Spirit Creek (SC) pond water (WSL in SC water, n = 10).Both groups show a significant increase in the rasping rate in carrot slurry compared with pond water.Also, WSL snails in SC pond water show a significantly high rasping rate in carrot compared to WSL snails in WSL pond water.(B) Spirit Creek control (SC in SC water, n = 21) and SC snails in WSL pond water (SC in WSL water, n = 10).Both groups show a significant increase in rasping rate in carrot slurry compared with pond water.The graphical representation shows the mean, and the error bars are the SEM.***p < 0.001; ****p < 0.0001.ns = not significant as p > 0.05.

FIGURE 4 :
FIGURE 4: Shell damage and healing progression of Lymnaea stagnalis after creating intentional fractures to the shells.The initial and final fracture sizes are illustrated for 10 individuals in each trial group: (A) Whitesand Lake (WSL) snails control (WSL in WSL water, n = 10) and WSL snails in Spirit Creek (SC) pond water (WSL in SC water, n = 10).Healing is significant in the WSL snails in SC pond water.(B) Spirit Creek control (SC in SC water, n = 10) and SC snails in WSL pond water (SC in WSL water, n = 10).The graphical representation shows the mean, and the error bars are the SEM.**p < 0.01.ns = not significant as p > 0.05.

FIGURE 5 :
FIGURE 5: Configural learning in Sprit Creek pond water versus artificial pond water.Configural learning in three groups of Spirit Creek snails.Three separate cohorts of Spirit Creek snails were used.The timeline of each experiment is presented above the data.(A) Spirit Creek snails (n = 8) were housed in Spirit Creek pond water for 5 days as well as throughout the configural learning experiment.These snails did not show configural learning memory.(B) Spirit Creek snails (n = 11) were housed in artificial pond water for 5 days as well as throughout the configural learning experiment.This group showed configural learning memory.(C) Spirit Creek snails (n = 10) were housed in artificial pond water for 5 days, followed by Spirit Creek pond water for 5 days and throughout the configural learning experiment.These snails did not show configural learning memory after the training procedure.The graphical representation shows the mean, and the error bars are SEM.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.SCW = Spirit Creek water; PW = pond water; C = carrot odor; CL = configural learning; C pre = initial response to carrot slurry; ns = not significant as p > 0.05.

TABLE 1 :
Mean water chemistry characteristics of Whitesand Lake and Spirit Creek during snail collection, April-August 2019

TABLE 2 :
Mean annual concentrations and standard deviations of neonicotinoids within Spirit Creek