Non-conventional endpoints show higher sulfoxaflor toxicity to Chironomus riparius than conventional endpoints in a multistress environment

Aquatic Toxicology Non-conventional endpoints show higher sulfoxaflor toxicity to Chironomus riparius than conventional endpoints in a multistress environment --Manuscript Draft-- Manuscript Number: Article Type: Research Paper Keywords: Ecotoxicity; Neurotoxin; Biomechanical endpoints; Behaviour; Multi-stress; Benthic species; Chronic Corresponding Author: Martina Vijver Leiden University Institute of Environmental Sciences Leiden, Netherlands First Author: Sofie B Rasmussen Order of Authors: Sofie B Rasmussen Thijs Bosker S. Henrik Barmentlo Olof Berglund Martina G. Vijver Abstract: Evidence grows that standard toxicity testing might underestimate the environmental risk of neurotoxic insecticides. Behavioural endpoints such as locomotion and mobility have been suggested as sensitive and ecologically relevant additions to the standard tested endpoints. Possible interactive effects of chemicals and additional stressors are typically overlooked in standardised testing. Therefore, we aimed to investigate how concurrent exposure to environmental stressors (increased temperature and predation cues) and a nicotinic acetylcholine receptor (nAChR)-modulating insecticide (‘sulfoxaflor’) impact Chironomus riparius across a range of conventional and nonconventional endpoints. We used a multifactorial experimental design encompassing three stressors, sulfoxaflor (2.0-110 µg/L), predation risk (presence/absence of predatory cues), and elevated temperature (20 °C and 23 °C), yielding a total of 24 distinct treatment conditions. To assess potential additive effects, we applied an Independent Action (IA) model to predict the impact on eight endpoints, including conventional endpoints (growth, survival, total emergence, and emergence time) and less conventional endpoints (the size of the adults, swimming abilities and exploration behaviour). For the conventional endpoints, observed effects were either lower than expected or well-predicted by the IA model. In contrast, we found greater than predicted effects of predation cues and temperature in combination with sulfoxaflor on adult size, larval exploration, and swimming behaviour. However, in contrast to the non-conventional endpoints, no conventional endpoints detected interactive effects of the neurotoxic insecticide and the environmental stressors. Acknowledging these interactions, increasing ecological context of ecotoxicological test systems may, therefore, advance environmental risk analysis and interpretation as the safe environmental concentrations of neurotoxic insecticides depend on the context of both the test organism and its environment. Suggested Reviewers: Henriette Selck [email protected] Dave Spurgeon [email protected] Susana Loureiro [email protected] Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Leiden University Institute of Environmental Sciences (CML) Einsteinweg 2 2333 CC Leiden The Netherlands Leiden, July 24st 2024 Manuscript for publication in Aquatic Toxicology. Authors: Sofie B. Rasmussena, Thijs Boskera, S. Henrik Barmentloa, Olof Berglundb, Martina G. Vijvera,* a Institute of Environmental Sciences, University of Leiden, Leiden, The Netherlands b Department of Biology, Lund University, Lund, Sweden *Corresponding author: Martina G. Vijver, [email protected]nuniv.nl Dear Editor-in-Chief, We hereby submit our manuscript entitled “Non-conventional endpoints show higher sulfoxaflor toxicity to Chironomus riparius than conventional endpoints in a multistress environment”. Highlights of the paper are:  Emergence and swimming of C. riparius is altered when exposed to sulfoxaflor  Additional stressors did not change the sensitivity of C. riparius to sulfoxaflor  Conventional endpoints predict no to positive effects of multiple stressors  Behavioural endpoints showed greater effects than predicted by IA modelling  Environmental stressors were more influential at higher sulfoxaflor concentrations As such, we are convinced that this work fits the scope of Aquatic Toxicology. This paper has never been published (not in whole nor substantial part). All persons entitled to authorship have been included. Furthermore, we have not submitted the manuscript to a preprint server before submitting it to Aquatic Toxicology. We thank you for considering this manuscript for publication and look forward to discussing any queries. Yours faithfully, On behalf of all authors, Sofie B. Rasmussen, Thijs Bosker, S. Henrik Barmentlo, Olof Berglund, and Martina G. Vijver Cover letter Non-conventional endpoints show higher sulfoxaflor toxicity to Chironomus riparius than conventional endpoints in a multistress environment Sofie B. Rasmussen1, Thijs Bosker 1, S. Henrik Barmentlo1, Olof Berglund2, Martina G. Vijver1,* 1 Institute of Environmental Sciences, Leiden University, P.O. Box 9518, 2300 RA Leiden, the Netherlands 2 Department of Biology, Lund University, Lund, Sweden *Corresponding author, Institute of Environmental Sciences, Leiden University, P.O. Box 9518, 2300 RA Leiden, the Netherlands. Email: [email protected]. Tel.: +31 71 527 1487 Highlights Highlights:  Emergence and swimming of C. riparius is altered when exposed to sulfoxaflor  Additional stressors did not change the sensitivity of C. riparius to sulfoxaflor  Conventional endpoints predict no to positive effects of multiple stressors  Behavioural endpoints showed greater effects than predicted by IA modelling  Environmental stressors were more influential at higher sulfoxaflor concentrations Page 1 of 31 Non-conventional endpoints show higher sulfoxaflor toxicity 1 to Chironomus riparius than conventional endpoints in a 2 multistress environment 3 Sofie B. Rasmussen1, Thijs Bosker 1, S. Henrik Barmentlo1, Olof Berglund2, Martina G. Vijver1,* 4 5 1 Institute of Environmental Sciences, Leiden University, P.O. Box 9518, 2300 RA Leiden, the 6 Netherlands 7 2 Department of Biology, Lund University, Lund, Sweden 8 *Corresponding author, Institute of Environmental Sciences, Leiden University, P.O. Box 9518, 2300 RA 9 Leiden, the Netherlands. Email: [email protected]. Tel.: +31 71 527 1487 10 11 12 Manuscript File Click here to view linked References Page 2 of 31 Highlights: 13  Emergence and swimming of C. riparius is altered when exposed to sulfoxaflor 14  Additional stressors did not change the sensitivity of C. riparius to sulfoxaflor 15  Conventional endpoints predict no to positive effects of multiple stressors 16  Behavioural endpoints showed greater effects than predicted by IA modelling 17  Environmental stressors were more influential at higher sulfoxaflor concentrations 18 19 20 21 Page 3 of 31 Abstract 22 Evidence grows that standard toxicity testing might underestimate the environmental risk of neurotoxic 23 insecticides. Behavioural endpoints such as locomotion and mobility have been suggested as sensitive and 24 ecologically relevant additions to the standard tested endpoints. Possible interactive effects of chemicals 25 and additional stressors are typically overlooked in standardised testing. Therefore, we aimed to 26 investigate how concurrent exposure to environmental stressors (increased temperature and predation 27 cues) and a nicotinic acetylcholine receptor (nAChR)-modulating insecticide (‘sulfoxaflor’) impact 28 Chironomus riparius across a range of conventional and non-conventional endpoints. We used a 29 multifactorial experimental design encompassing three stressors, sulfoxaflor (2.0-110 µg/L), predation 30 risk (presence/absence of predatory cues), and elevated temperature (20 °C and 23 °C), yielding a total of 31 24 distinct treatment conditions. To assess potential additive effects, we applied an Independent Action 32 (IA) model to predict the impact on eight endpoints, including conventional endpoints (growth, survival, 33 total emergence, and emergence time) and less conventional endpoints (the size of the adults, swimming 34 abilities and exploration behaviour). For the conventional endpoints, observed effects were either lower 35 than expected or well-predicted by the IA model. In contrast, we found greater than predicted effects of 36 predation cues and temperature in combination with sulfoxaflor on adult size, larval exploration, and 37 swimming behaviour. However, in contrast to the non-conventional endpoints, no conventional endpoints 38 detected interactive effects of the neurotoxic insecticide and the environmental stressors. Acknowledging 39 these interactions, increasing ecological context of ecotoxicological test systems may, therefore, advance 40 environmental risk analysis and interpretation as the safe environmental concentrations of neurotoxic 41 insecticides depend on the context of both the test organism and its environment. 42 43 Keywords: 44 Ecotoxicity, Neurotoxin, Biomechanical endpoints, Behaviour, Multi-stress, Benthic species, Chronic 45 Page 4 of 31 1 Introduction: 46 Neurotoxic insecticides are ubiquitous in agricultural practices (Costa et al., 2008; Berens et al., 2021; 47 Casillas et al., 2022), and as a result, their residues are commonly detected in surface waters globally 48 (Borsuah et al., 2020; Thompson et al., 2021; Wang et al., 2023). This contamination has raised concerns 49 for aquatic ecosystems, particularly for the health of aquatic organisms (Casillas et al., 2022). Sulfoxaflor 50 is a recently introduced insecticide in the sulfoximine class, whose primary site of action is nerve action 51 (Gauthier & Mabury, 2021). Although classified as a unique class (Sparks et al., 2013), it shares many 52 characteristics with the extensively used neonicotinoids (Cutler et al., 2012). These characteristics have 53 previously been observed to cause effects on aquatic communities, including declines in macro54 invertebrate (Van Dijk et al., 2013; Sánchez-Bayo et al., 2016) and insect abundance (Goulson, 2014; 55 Barmentlo et al., 2021). 56 Regulations on pesticides rely on a tiered approach of risk assessments, starting with short-term tests 57 involving single species under standardised conditions (EFSA, 2013; Diepens et al., 2016; Schuijt et al., 58 2021). However, these conventional toxicity metrics may underestimate the ecological impact of 59 neurotoxic agents in aquatic environments (Vijver et al., 2017; Legradi et al., 2018; de Campos et al., 60 2022). Sensitivities to neonicotinoid and lufenuron at LOEC levels up to 2500 times lower have been 61 found for multiple organisms, including Daphnia magna, Chironomus riparius and Hyalella Azteca when 62 exposed in situ compared to the standard laboratory OECD test (Barmentlo et al., 2018; Brock et al., 63 2016) 64 Traditional ecotoxicity assessments tend to focus on endpoints such as mortality and reproduction, often 65 not reflecting the species responses that occur before these life-history endpoints (Sarma & Nandini, 2006; 66 Forbes et al., 2017; Legradi et al., 2018). Giving the primary mode of action is nerve action for neurotoxic 67 insecticides, behavioural endpoints (e.g., locomotion and mobility) are expected to show higher sensitivity 68 compared to conventional endpoints (Augusiak & Van den Brink, 2016; Weichert et al., 2017). To 69 illustrate, Raby et al. (2018) report a difference in sensitivity of up to three orders of magnitude between a 70 Page 5 of 31 standard endpoint (mortality) and less conventional endpoints (limited swimming behaviour or muscle 71 spasms) of C. dipterum larvae when exposed to acute doses of neonicotinoids. Impaired ability to burrow 72 in C. riparius larvae has been linked to them being significantly more prone to predation by zebrafish 73 (Danio rerio), highlighting the relevance of mobility for survival and, hence, suggesting potential impacts 74 on population dynamics (Langer-Jaesrich et al., 2010). Incorporating behavioural endpoints into 75 ecological risk assessments of neurotoxins holds promise due to their sensitivity and ecological relevance 76 (Ågerstrand et al., 2020; Bownik & Wlodkowic, 2021). 77 There is a low level of environmental realism in lower-tier risk screening, as species are exposed under 78 optimal conditions and often to a single stressor (Holmstrup et al., 2010; Schuijt et al., 2021). Adding 79 stressors in the experimental design, such as elevated surface water temperatures, can alter the sensitivity 80 of aquatic species to chemical exposure (Scherer et al., 2013; Macaulay et al., 2020). This finding pleads 81 for enhancing environmental relevance within the test settings (Holmstrup et al., 2010; Jackson et al., 82 2016), allowing for a broader impact assessment. Abiotic conditions fluctuate naturally and are in the 83 future likely more subject to change given climate change; lake surface temperatures have increased 84 worldwide by 0.34 °C per decade (Woolway et al., 2020) and are further expected to increase by 2.7 to 7 85 °C in summer periods (Hardenbicker et al., 2017; Rajesh & Rehana, 2022). Higher water and air 86 temperatures can influence vital responses related to organisms' fitness, such as increased metabolism 87 (Shah et al., 2020) and emergence of aquatic insect species (Heye et al., 2019; Dellar et al., 2022). 88 Furthermore, behavioural changes as a result of increased water temperature have been shown for 89 Diamesa zernyi (Chironomidae) larvae. However, in contrast to exposure to neurotoxic insecticides, 90 distance and speed were increased compared to controls after just 24 hours of exposure (Lencioni et al., 91 2021). 92 Moreover, one of the most important biotic forms of stress is predation. Predatory cues can significantly 93 influence the behaviour of prey organisms, which can directly influence exposure as well as response 94 parameters to the toxicant (Langer-Jaesrich et al., 2010). Indirectly, the predatory cues potentially 95 Page 12 of 31 the identified positions of larvae were adjusted where needed. Smoothing over 35 frames was done to 227 lower inconsistencies in the precise position due to larvae sizes. Analysed endpoints included swimming 228 ratio (%), average swimming speed (cm/s), and exploration of the arena (cm2). For the swimming ratio, a 229 threshold of 0.5 cm/s was chosen to be sufficient to successfully locate swimming phases and to avoid 230 including non-moving periods. After localising all swimming events, the swimming ratio was calculated 231 as time spent swimming (v<0.5 cm/s) over total recording time. Average velocities were taken per larvae, 232 only including frames that were moving at 0.5 cm/s. Lastly, for continuity, an area of 0.21 cm2 per larvae 233 independent of larva size was used for explorations. 234 Living larvae (visually checked on health) were returned to corresponding treatment beaker for 235 measurements of emergence. Emergence was quantified by counting the number of adults in each 236 replicate daily, after the first emerging adult until day 23 (Figure 2). Adults were trapped in their 237 respective beakers using parafilm. After removal with a tweezer, all adults were sexed to account for 238 sexual dimorphisms in size, and images were taken for length measurement. The length of the adults was 239 determined using ImageJ, from the tip of the head until the tip of the tail. All data for total emergence, 240 time of emergence, and adult size were pooled for both sexes to optimise replicate numbers, as no 241 significant impact of gender was found. 242 243 2.5 Statistical analyses 244 Potential differences between controls (20°C, no predation cue), further referred to as culture controls of 245 the different exposure scenarios, were tested using a one-way ANOVA, followed by a Tukey's posthoc 246 test. Normality of error distribution was tested with a Shapiro-Wilk test and Levene’s test to ensure 247 homogeneity of variances. Furthermore, dose-response curves for the effects of sulfoxaflor on C. riparius 248 for each of the four scenarios: predation (presence /absence), and temperature (20 and 23 °C) were 249 performed in R, using the drc package and a 3 par log-logistic function, the corresponding 95 % 250 Page 13 of 31 confidence intervals were calculated by non-linear regression. All dose-response curves were based on 251 data normalised to the treatment with no added sulfoxaflor. 252 Based on each dose-response curve, the 50% effect concentration (EC50) and corresponding 95 % 253 confidence interval were determined. The effect concentrations and slope of each curve was compared 254 with a f-test using the command compParm from the drc-package (Ritz et al., 2015). The significance 255 level for both EC50 values and slope was 0.05. 256 An Independent Action (IA) model was applied to the eight chosen endpoints to calculate the chemical, 257 predator, and/or increased temperature impact on the organisms’ vitality. We expected the stressors to act 258 in a response additive way. Deviations in observed values from the modelled value were interpreted as 259 positive (when lower) and negative (when higher). For the treatment combining all three stressors, 260 responses of the individual stressors were used for the model. The model was as followed (Loewe et al., 261 1926): 262 𝐸𝑚𝑖𝑥 =∏(1 − 𝑖𝑒𝑖−𝑒𝑐𝑜𝑛𝑡𝑟𝑜𝑙 𝑒𝑚𝑎𝑥−𝑒𝑐𝑜𝑛𝑡𝑟𝑜𝑙) 263 Emix is the predicted effect of the joint stress, eI is the observed effect of the single stressor i, econtrol being 264 the effect observed under control conditions, and emax is the maximum effect that is possible. For survival 265 data, emax was set at 0. For other endpoints where a maximum effect of zero was not meaningful, this was 266 set as the largest observed effect. 267 As the actual measured concentrations of sulfoxaflor varied between scenarios, the individual stress of the 268 chemical on each endpoint was estimated based on the dose-response relationships under standard 269 conditions (20 °C and absence of predation cues). For the endpoints, mean emergence day and adult size, 270 the expected values were hence extrapolated from effects in 80 µg/L, as no adults were observed in the 271 highest nominal concentration of 160 µg/L. 272 273 Page 14 of 31 3 Results: 274 3.1 Water chemistry and verification of test setup 275 Water chemistry parameters in the exposure solutions were not affected by the addition of sulfoxaflor, rise 276 in temperature, or the addition of predation cues. All measured pH values were within the range of 7.5 to 277 8.2, and salinities were measured to be below 2 ‰ for all treatments. Dissolved oxygen was above 90 % 278 saturation, independent of concentration or scenario treatment. Measurements of sulfoxaflor 279 concentrations found an average of 60 % recovery in the treatments. No differences in sulfoxaflor 280 recoveries were found over time or between scenarios. All measures related to water chemistry can be 281 found in the supplementary information (supplementary information, B). For all scenarios, culture controls 282 of C. riparius were included to ensure stable conditions for the vitality of the tested organisms at all 283 timeslots. Survival on day 8 and total emergence after 23 days were not impacted over time, and all 284 culture controls showed over 70 % total emergence, validating the test setup according to the OECD 285 guideline (OECD, 2004). However, some minor differences were found when assessing the behavioural 286 endpoints (see supplementary information, Figure C.1), and therefore, all results were normalised to each 287 exposure scenario's respective culture control. 288 289 3.2 Effect values 290 Dose-response relationships and corresponding EC50 values were estimated separately per scenario and 291 per endpoint (Table 1). We found few significant differences in EC50 values per endpoint across different 292 scenarios (see supplementary information, Table D.1). The exception is total emergence, where the 293 standard exposure scenario (20 °C and absence of predation cues) showed significantly lower EC50 values 294 compared to the other scenarios. However, it should be noted that there were relatively large confidence 295 intervals in most of the measured endpoints. 296 Sensitivities among endpoints differed slightly across scenarios. Overall, total emergence was the most 297 sensitive of the conventional endpoints, with EC50 values ranging from 15 µg/L in the standard exposure 298 Page 15 of 31 scenario to 79 µg/L sulfoxaflor in the scenario at 20 °C and presence of predation cues. Noteworthy, all 299 scenarios were found to have steep dose-response curves for this endpoint, with no emerging adults found 300 in any scenario for the highest concentration (all modelled dose-response curves can be found in 301 supplementary information, Figure D.1). The standard exposure scenario showed no or low dose302 dependent effect on mean emergence date and larval growth (Table 1). 303 For the standard exposure scenario, all behavioural endpoints (swimming, speed, and exploration) showed 304 EC50 values between 20 to 50 µg/L, making them less sensitive than the most sensitive standard endpoint: 305 total emergence (Table 1). However, for the three additional stress scenarios, estimated EC50 values were 306 either lower or in a similar range as the total emergence, except for exploration at 20 °C in presence of 307 predation cues and swimming time at 23 °C in presence of predation cues (Table 1). In addition, effects of 308 sulfoxaflor on adult size were within a similar sensitivity range as the behavioural endpoints for the three 309 stress scenarios. 310 Page 16 of 31 Table 1. Effect values of 50% effect (EC50) on C. riparius exposed to sulfoxaflor under four different stress scenarios and for selected endpoints. 311 Scenario 20 °C No Predation 23 °C No Predation 20 °C Predation 23 °C Predation Endpoint EC50 µg/L (CI) P value EC50 µg/L (CI) P value EC50 µg/L (CI) P value EC50 µg/L (CI) P value Conventional endpoints Survival 3,8 (0 – 43,000 ) 0.84 110 (85 – 140) < 0.001 13 (91 – 160) < 0.001 120 (91 – 160) < 0.001 Total emergence 14 (0 – 45) 0.34 72 (66 – 77) < 0.001 79 (72 – 86) < 0.001 74 (65. – 83) < 0.001 Mean emergence day 15 (NaN) NaN 35 (0 - 110) 0.34 150 (NaN) NaN 31 (NaN) NaN Growth 320 (0 – 3,600) 0.84 200 (0 – 86) 0.54 820 (7.0 – 160) 0.03 86 (75 – 97) < 0.001 Non-conventional endpoints Size of adult 0.24 (NaN) NaN 160 (0 – 2,400) 0.88 82.(0 - 330) 0.49 76 (0 - 220) 0.29 Swimming 36 (0 – 160) 0.57 100 (0 – 450) 0.56 104 (0 – 1,30) 0.86 180 (NaN) NaN Speed 47 (0 – 100) 0.08 64 (56 – 72) < 0.001 82.0 (0 – 990) 0.85 63(1.0 – 130) 0.047 Exploration 22 (0 – 230,000) 0.99 734 (0 – 160) 0.08 5340 (0 – 10,000) 0.91 60 (40 – 79) < 0.001 Legend: 20 °C No Predation = standard conditions of 20 °C without the addition of predation cues, 23 °C No Predation = exposed to 23 °C without 312 the addition of predation cues, 20 °C Predation = exposed at 20 °C in presence of predation cues, 23 °C Predation = exposed to 23 °C in presence 313 of predation cues. CI = confidence intervals, NaN= values could not be established. Non-conventional endpoints chosen were all related to 314 biomechanistic endpoints. 315 316 Page 17 of 31 3.3 Independent action 317 When comparing differences between observed and modelled responses using the IA model, we found 318 that conventional endpoints, overall, showed either limited differences with the standard exposure 319 scenario or a larger observed than expected value (Figure 3; Table 2). Here, low values for the measured 320 endpoints are regarded as a potential effect of the organisms. Generally, the higher the values, the better 321 we regard the animal's fitness. For example, high total emergence would mean that the organisms could 322 survive and thrive to adulthood, increasing the potential to reproduce and benefitting the population. 323 Besides mean emergence days, significant correlations between expected and observed values were found 324 for the conventional endpoints (Table 2). However, when the model predicted low values, observed values 325 still showed varying but limited effects, making the slope of the correlation relatively flat (see 326 supplementary information, Figure E.1). Some exemptions were found in larval length, mean emergence 327 day and total emergence, especially for higher concentrations of sulfoxaflor, where the observed values 328 were lower than expected (Figure 3). Furthermore, the conventional endpoints showed limited differences 329 between the exposure scenarios (Table 2). 330 In contrast to the conventional endpoints, all four non-conventional endpoints showed significant 331 deviations from the 1:1 observed expected values (Table 2). For biomechanistic endpoints, swimming 332 time, swimming velocity and exploration, negative correlations between observed and expected values 333 were found for all concentrations of sulfoxaflor and across all three stress scenarios (Figure 3; 334 supplementary information, Figure E.1). For adult size, strong correlations were observed across all three 335 exposure scenarios (Table 2), however, only minor deviations between observed and expected values were 336 found, with, in general, minor effects on adult size (see supplementary information, Figure E.1). The 337 difference between observed and IA-predicted effects seemed to depend on scenarios for several of the 338 endpoints including swimming time and velocity, where the combination of high temperature and 339 predation cues showed the highest negative deviations between observed and expected values (Figure 3), 340 and the lowest correlations; Spearman ρ is 0.44 (p=0.015) and 0.55 (p=0.002), respectively (Table 2). This 341 Page 18 of 31 tendency did not seem to depend on sulfoxaflor concentration (Figure 3). Instead, high swimming values 342 were expected based on increased temperature and predation cues alone. In contrast, these effects were not 343 observed in combined treatments (see supplementary information, Figure E.1). Interestingly, for 344 exploration and adult size, the difference between observed and expected values became more evident in 345 higher sulfoxaflor concentrations, indicating that potential stress on stress responses depends on 346 compound concentrations (Figure 3). 347 In brief, negative deviations from additive responses were not observed for conventional endpoints but 348 were observed when focusing on biomechanistic endpoints. This illustrates the added value and 349 importance of explaining sensitivity, making use of non-conventional endpoints for neurotoxin exposure 350 at realistic environmental concentrations. This added value was most dominant when larvae were exposed 351 to the combination of all three stressors: sulfoxaflor, increased temperature, and predation stress. Lastly, 352 the higher the concentration of sulfoxaflor, the stronger the influence of environmental factors was. 353 Page 19 of 31 354 Figure 3, Visual representation of additive effects based on the Independent Action model. The differences between observed and expected values 355 explain deviations from additivity and, hence, represent other types of joint effects. Data was normalised and shown as differences to the standard 356 exposure scenario (20 °C and absence of predation cues). Blue ▴ = 23 °C and no predation cues, yellow ▪ = 20 °C and added predation cues, 357 orange • = 23 °C and added predation cues. Error bars represent standard error (SE). 358 359 Page 20 of 31 Table 2. Spearman rank correlations between observed and expected values for the three stress scenarios. 360 Scenario 23 °C No Predation 20 °C Predation 23 °C Predation Endpoint df ρ P value df ρ P value df ρ P value Conventional endpoints Survival 28 0.523 0.003 28 0.453 0.012 28 0.387 0.035 Total emergence 28 0.789 <0.001 28 0.819 <0.001 28 0.771 <0.001 Mean emergence day 20 -0.275 0.216 23 0.391 0.053 23 0.065 0.756 Growth 28 0.725 <0.001 28 0.806 <0.001 28 0.797 <0.001 Non-conventional endpoints Size of adult 19 0.747 <0.001 23 0.694 <0.001 22 0.737 <0.001 Swimming 28 0.770 <0.001 28 0.644 <0.001 28 0.439 0.015 Speed 28 0.863 <0.001 28 0.614 <0.001 28 0.554 0.002 Exploration 28 -0.819 <0.001 28 -0.696 <0.001 28 -0.785 <0.001 Legend: 23 °C No Predation = exposed to 23 °C and without the addition of predatory cue, 20 °C Predation = exposed at 20 °C and the addition of 361 predation cues, 23 °C Predation = exposed to 23 °C and addition of predation cues. Statistic abbreviations: df = degrees of freedom, ρ = 362 Spearman's rank correlation coefficient. 363 364 Page 21 of 31 365 4 Discussion: 366 4.1 Effects of single stressors 367 In the present study, the toxicity of sulfoxaflor on C. riparius at standard test conditions was determined 368 by dose-response curves for eight different endpoints for which non-conventional endpoints were most 369 sensitive. Standardised endpoints of emergence (OECD, 2004) had an EC50 value of 14.6 µg/L, in line 370 with previous studies reporting an EC50 of 18.3 µg/L for total emergence (Rasmussen et al., 2024). 371 Robustness and replicability were further confirmed for another chironomid species, Chironomus 372 kiinensis larvae, with EC50 on emergence at 20 µg/L (Liu et al., 2021). 373 When exposing the larvae to increased temperature without the influence of other stressors, we did not 374 observe any deviations from already expected effects on development. Increasing the temperature to 23 °C 375 compared to standard conditions at 20 °C significantly increased the growth of the larvae and resulted in 376 faster emergence. Length at day 10 was increased to 1.04 cm (±0.03) at 23 °C compared to 0.91 cm 377 (±0.03) at 20 °C and mean emergence day was at day 14 (±0.7) and 17 (±0.4) for 23 °C and 20 °C, 378 respectively (see supplementary information, Figure C.2). Based on previous literature, lower mean 379 emergence times in higher temperatures were expected as Heye et al. (2019) showed faster female 380 emergences already at 22 °C. They also found that faster emergence at higher temperatures could come 381 with a trade-off, resulting in smaller adults. However, we did not observe any significant changes in adult 382 size, nor larval survival and total emergence (see supplementary information, Figure C.2). Contrary to 383 known literature on one similar species, we did not find significant changes for any of the behavioural 384 endpoints when elevating the temperature. Increased temperature has been shown to cause hyperactivity in 385 Chironomidae larvae (Diamesa zernyi), with significant increases in both travelled distance and velocity 386 resulting from 72 h of exposure (Lencioni et al., 2021). Interspecies differences, or longer exposure 387 periods giving the larvae time to acclimate, could explain the discrepancies (Shaw, 2020). Based on our 388 observations, increasing the temperature to 23 °C did not seem to impact the C. riparius adversely. 389 Page 28 of 31 Augusiak, J., & Van den Brink, P. J. (2016). The influence of insecticide exposure and environmental 520 stimuli on the movement behaviour and dispersal of a freshwater isopod. Ecotoxicology, 25, 1338521 1352. 522 Barmentlo, S. H., Parmentier, E. M., de Snoo, G. R., & Vijver, M. G. (2018). Thiacloprid‐induced toxicity 523 influenced by nutrients: Evidence from in situ bioassays in experimental ditches. Environmental 524 toxicology and chemistry, 37(7), 1907-1915. 525 Barmentlo, S. H., Schrama, M., De Snoo, G. R., Van Bodegom, P. M., van Nieuwenhuijzen, A., & Vijver, 526 M. G. (2021). Experimental evidence for neonicotinoid driven decline in aquatic emerging 527 insects. Proceedings of the National Academy of Sciences, 118(44), e2105692118. 528 Berens, M.J., Capel, P.D. and Arnold, W.A. (2021), Neonicotinoid Insecticides in Surface Water, 529 Groundwater, and Wastewater Across Land-Use Gradients and Potential Effects. Environ Toxicol 530 Chem, 40: 1017-1033. https://doi.org/10.1002/etc.4959 531 Borsuah, J. F., Messer, T. L., Snow, D. D., Comfort, S. D., & Mittelstet, A. R. (2020). Literature review: 532 Global neonicotinoid insecticide occurrence in aquatic environments. Water, 12(12), 3388. 533 Bownik, A., & Wlodkowic, D. (2021). Applications of advanced neuro-behavioral analysis strategies in 534 aquatic ecotoxicology. Science of the Total Environment, 772, 145577. 535 Brock, T. C. M., Bas, D. A., Belgers, J. D. M., Bibbe, L., Boerwinkel, M. C., Crum, S. J. H., Diepens, N. 536 J., Kraak, M. H. S., Vonk, J. A. & Roessink, I. (2016). Effects of sediment-spiked lufenuron on 537 benthic macroinvertebrates in outdoor microcosms and single-species toxicity tests. Aquatic 538 Toxicology, 177, 464-475. 539 Bundschuh, M., Zubrod, J. P., Klöttschen, S., Englert, D., & Schulz, R. (2020). Infochemicals influence 540 neonicotinoid toxicity—impact in leaf consumption, Growth, and predation of the amphipod 541 Gammarus fossarum. Environmental Toxicology and Chemistry, 39(9), 1755-1764. 542 de Campos, B. G., Figueiredo, J., Perina, F., Abessa, D. M. D. S., Loureiro, S., & Martins, R. (2022). 543 Occurrence, effects and environmental risk of antifouling biocides (EU PT21): Are marine 544 ecosystems threatened?. Critical Reviews in Environmental Science and Technology, 52(18), 545 3179-3210. 546 Casillas, A., de la Torre, A., Navarro, I., Sanz, P., & de los Ángeles Martínez, M. (2022). Environmental 547 risk assessment of neonicotinoids in surface water. Science of The Total Environment, 809, 548 151161. 549 Chiara, V., & Kim, S. Y. (2023). AnimalTA: A highly flexible and easy‐to‐use program for tracking and 550 analysing animal movement in different environments. Methods in Ecology and Evolution. 551 14:1699–1707. DOI: 10.1111/2041-210X.14115 552 Costa, L. G., Giordano, G., Guizzetti, M., & Vitalone, A. (2008). Neurotoxicity of pesticides: a brief 553 review. Frontiers in Bioscience-Landmark, 13(4), 1240-1249. 554 Cutler, P., Slater, R., Edmunds, A.J., Maienfisch, P., Hall, R.G., Earley, F.G., Pitterna, T., Pal, S., Paul, 555 V.L., Goodchild, J., Blacker, M., Hagmann, L., Crossthwaite, A.J. (2012). Investigating the mode 556 of action of sulfoxaflor: a fourth-generation neonicotinoid. Pest Manag Sci. 2013 May;69(5):607557 19. doi: 10.1002/ps.3413. Epub 2012 Oct 30. PMID: 23112103. 558 Dellar, M., Boerlijst, S. P., & Holmes, D. (2022). Improving estimations of life history parameters of 559 small animals in mesocosm experiments: A case study on mosquitoes. Methods in Ecology and 560 Evolution, 13(5), 1148-1160. 561 EFSA (European Food Safety Authority), 2013. Guidance on tiered risk assessment for plant protection 562 products for aquatic organisms in edge-of-field surface waters. EFSA J. 11 (7), 3290, 268 pp. 563 Forbes, V. E., Salice, C. J., Birnir, B., Bruins, R. J., Calow, P., Ducrot, V., Galic, N., Garber, K., Harvey, 564 B. C., Jager, H., Kanarek, A., Pastorok, R., Railsback, S. F., Rebarber, R. & Thorbek, P. (2017). A 565 framework for predicting impacts on ecosystem services from (sub) organismal responses to 566 chemicals. Environmental toxicology and chemistry, 36(4), 845-859. 567 Page 29 of 31 Gauthier, J. R., & Mabury, S. A. (2021). The Sulfoximine Insecticide Sulfoxaflor and Its Photodegradate 568 Demonstrate Acute Toxicity to the Nontarget Invertebrate Species Daphnia magna. 569 Environmental Toxicology and Chemistry, 40(8), 2156-2164. 570 Goulson, D. (2014). Pesticides linked to bird declines. Nature 511, 295–296. https://doi571 org.ezproxy.leidenuniv.nl/10.1038/nature13642 572 Hardenbicker, P., Viergutz, C., Becker, A., Kirchesch, V., Nilson, E., & Fischer, H. (2017). Water 573 temperature increases in the river Rhine in response to climate change. Regional Environmental 574 Change, 17, 299-308. 575 Heye, K., Lotz, T., Wick, A., & Oehlmann, J. (2019). Interactive effects of biotic and abiotic 576 environmental stressors on carbamazepine toxicity in the non-biting midge Chironomus riparius. 577 Water research, 156, 92-101. 578 Hölker, F., & Stief, P. (2005). Adaptive behaviour of chironomid larvae (Chironomus riparius) in 579 response to chemical stimuli from predators and resource density. Behavioral Ecology and 580 Sociobiology, 58, 256-263. 581 Holmstrup, M., Bindesbøl, A. M., Oostingh, G. J., Duschl, A., Scheil, V., Köhler, H. R., Loureiro, S., 582 Soares, A. M. V. M., Ferreira, A. L. G., Kienle, C., Gerhardt, A., Laskowski, R., Kramarz, P. E., 583 Bayley, M., Svendsen, C. & Spurgeon, D. J. (2010). Interactions between effects of environmental 584 chemicals and natural stressors: a review. Science of the Total Environment, 408(18), 3746-3762. 585 Jackson, M. C., Loewen, C. J., Vinebrooke, R. D., & Chimimba, C. T. (2016). Net effects of multiple 586 stressors in freshwater ecosystems: a meta‐analysis. Global change biology, 22(1), 180-189. 587 Karima, Z., (2021). Chironomidae: Biology, Ecology and Systematics. The Wonders of Diptera588 Characteristics, Diversity, and Significance for the World’s Ecosystems. DOI: 589 10.5772/intechopen.95577 590 Langer-Jaesrich, M., Kienle, C., Köhler, H. R., & Gerhardt, A. (2010). Impairment of trophic interactions 591 between zebrafish (Danio rerio) and midge larvae (Chironomus riparius) by chlorpyrifos. 592 Ecotoxicology, 19, 1294-1301. 593 Legradi, J. B., Di Paolo, C., Kraak, M. H. S., Van der Geest, H. G., Schymanski, E. L., Williams, A. J., 594 Dingemans, M. M. L., Massei, R., Brack, W., Cousin, X., Begout, M.-L., van der Oost, R., 595 Carion, A., Suarez-Ulloa, V., Silvestre, F., Escher, B. I., Engwall, M., Nilén, G., Keiter, S. H., 596 Pollet, D., Waldmann, P., Kienle, C., Werner, I., Haigis, A.-C., Knapen, D., Vergauwen, L., 597 Spehr, M., Schulz, W., Busch, W., Leuthold, D., Scholz, S., vom Berg, C. M., Basu, N., Murphy, 598 C. A., Lampert, A., Kuckelkorn, J., Grummt, T., & Hollert, H. (2018). An ecotoxicological view 599 on neurotoxicity assessment. Environmental Sciences Europe, 30(1), 1-34 600 Lencioni, V., Di Nica, V., & Villa, S. (2021). Investigation of the combined effects of rising temperature 601 and pesticide contamination on the swimming behaviour of alpine chironomids. Water, 13(24), 602 3618. 603 Liu, P., Wu, F., Li, H., & You, J. (2021). The neonicotinoid alternative sulfoxaflor causes chronic toxicity 604 and impairs mitochondrial energy production in Chironomus kiinensis. Aquatic Toxicology, 235, 605 105822. 606 Loewe, S. T. (1926). Effect of combinations: mathematical basis of problem. Arch. Exp. Pathol. 607 Pharmakol., 114, 313-326. 608 Macaulay, S. J., Buchwalter, D. B., & Matthaei, C. D. (2020). Water temperature interacts with the 609 insecticide imidacloprid to alter acute lethal and sublethal toxicity to mayfly larvae. New Zealand 610 Journal of Marine and Freshwater Research, 54(1), 115-130. 611 Marziali, L., Rosignoli, F., Valsecchi, S., Polesello, S., & Stefani, F. (2019). Effects of perfluoralkyl 612 substances on a multigenerational scale: a case study with Chironomus riparius (Diptera, 613 Chironomidae). Environmental toxicology and chemistry, 38(5), 988-999. 614 Page 30 of 31 Moore, E. A., Babbitt, C. W., Connelly, S. J., Tyler, A. C., & Rogalskyj, G. (2019). Cascading ecological 615 impacts of fullerenes in freshwater ecosystems. Environmental toxicology and chemistry, 38(8), 616 1714-1723. 617 Nowak, C., Jost, D., Vogt, C., Oetken, M., Schwenk, K., & Oehlmann, J. (2007). Consequences of 618 inbreeding and reduced genetic variation on tolerance to cadmium stress in the midge Chironomus 619 riparius. Aquatic Toxicology, 85(4), 278-284. 620 OECD (2004), Test No. 219: Sediment-Water Chironomid Toxicity Using Spiked Water, OECD 621 Guidelines for the Testing of Chemicals, Section 2, OECD Publishing, Paris, 622 https://doi.org/10.1787/9789264070288-en. 623 Pestana, J. L., Loureiro, S., Baird, D. J., & Soares, A. M. (2009). Fear and loathing in the benthos: 624 responses of aquatic insect larvae to the pesticide imidacloprid in the presence of chemical signals 625 of predation risk. Aquatic Toxicology, 93(2-3), 138-149. 626 Pestana, J. L., Loureiro, S., Baird, D. J., & Soares, A. M. (2010). Pesticide exposure and inducible 627 antipredator responses in the zooplankton grazer, Daphnia magna Straus. Chemosphere, 78(3), 628 241-248. 629 Raby, M., Nowierski, M., Perlov, D., Zhao, X., Hao, C., Poirier, D. G., & Sibley, P. K. (2018). Acute 630 toxicity of 6 neonicotinoid insecticides to freshwater invertebrates. Environmental Toxicology and 631 Chemistry, 37(5), 1430-1445. 632 Rajesh, M., Rehana, S. (2022). Impact of climate change on river water temperature and dissolved 633 oxygen: Indian riverine thermal regimes. Sci Rep 12, 9222. https://doi.org/10.1038/s41598-022634 12996-7 635 Rasmussen, S. B., Bosker, T., Ramanand, G. G., & Vijver, M. G. (2024). Participatory hackathon to 636 determine ecological relevant endpoints for a neurotoxin to aquatic and benthic invertebrates. 637 Environmental Science and Pollution Research, 31:22885–22899. https://doi.org/10.1007/s11356638 024-32566-w 639 Ritz C., Baty F., Streibig J., C., Gerhard D. (2015) Dose-Response Analysis Using R. PLoS ONE 10(12): 640 e0146021. https://doi.org/10.1371/journal.pone.0146021 641 Sánchez-Bayo F, Goka K and Hayasaka D (2016) Contamination of the Aquatic Environment with 642 Neonicotinoids and its Implication for Ecosystems. Front. Environ. Sci. 4:71. doi: 643 10.3389/fenvs.2016.00071 644 Sarma, S. S. S., & Nandini, S. (2006). Review of recent ecotoxicological studies on cladocerans. Journal 645 of Environmental Science and Health, Part B, 41(8), 1417-1430. 646 Scherer, C., Seeland, A., Oehlmann, J., & Müller, R. (2013). Interactive effects of xenobiotic, abiotic and 647 biotic stressors on Daphnia pulex—Results from a multiple stressor experiment with a fractional 648 multifactorial design. Aquatic toxicology, 138, 105-115. 649 Schuijt, L. M., Peng, F. J., van den Berg, S. J., Dingemans, M. M., & Van den Brink, P. J. (2021). (Eco) 650 toxicological tests for assessing impacts of chemical stress to aquatic ecosystems: Facts, 651 challenges, and future. Science of the Total Environment, 795, 148776. 652 Sibley, P.K., Ankley, G.T. & Benoit, D.A. (2001), Factors affecting reproduction and the importance of 653 adult size on reproductive output of the midge Chironomus tentans. Environmental Toxicology 654 and Chemistry, 20: 1296-1303. 655 Silva, C. J., Machado, A. L., Campos, D., Soares, A. M., & Pestana, J. L. (2022). Combined effects of 656 polyethylene microplastics and natural stressors on Chironomus riparius life-history traits. 657 Environmental Research, 213, 113641. 658 Shah, A. A., Woods, H. A., Havird, J. C., Encalada, A. C., Flecker, A. S., Funk, W. C., Guayasamin, J. 659 M., Kondratieff, C. B., Poff, N. L., Thomas, S. A., Zamudio, K. R., & Ghalambor, C. K. (2021). 660 Temperature dependence of metabolic rate in tropical and temperate aquatic insects: Support for 661 Page 31 of 31 the Climate Variability Hypothesis in mayflies but not stoneflies. Global Change Biology, 27(2), 662 297-311. 663 Shaw, A. K. (2020). Causes and consequences of individual variation in animal movement. Movement 664 ecology, 8(1), 12. 665 Sparks, T. C., Watson, G. B., Loso, M. R., Geng, C., Babcock, J. M., & Thomas, J. D. (2013). Sulfoxaflor 666 and the sulfoximine insecticides: Chemistry, mode of action and basis for efficacy on resistant 667 insects. Pesticide Biochemistry and Physiology, 107(1), 1-7. 668 Tiselius, P., Hansen, B., Jonsson, P., Kiørboe, T., Nielsen, T. G., Piontkovski, S., & Saiz, E. (1995). Can 669 we use laboratory-reared copepods for experiments? A comparison of feeding behaviour and 670 reproduction between a field and a laboratory population of Acartia tonsa. ICES Journal of 671 Marine Science, 52(3-4), 369-376. 672 Thompson, D. A., Hruby, C. E., Vargo, J. D., & Field, R. W. (2021). Occurrence of neonicotinoids and 673 sulfoxaflor in major aquifer groups in Iowa. Chemosphere, 281, 130856. 674 US EPA (US Environmental Protection Agency), 2013. Environmental fate and ecological risk assessment 675 for sulfoxaflor registration. Washington, DC. [cited 2023, May 14]. Available from: 676 https://www.regulations.gov/document/EPA-HQ-OPP-2010-0889-0022 677 Van Dievel, M., Janssens, L., & Stoks, R. (2019). Additive bioenergetic responses to a pesticide and 678 predation risk in an aquatic insect. Aquatic Toxicology, 212, 205-213. 679 Van Dijk, T. C., Van Staalduinen, M. A., & Van der Sluijs, J. P. (2013). Macro-invertebrate decline in 680 surface water polluted with imidacloprid. PLoS ONE, 8, e62374. 681 doi.org/10.1371/journal.pone.0062374 682 Van Praet, N., De Jonge, M., Stoks, R., & Bervoets, L. (2014). Additive effects of predator cues and 683 dimethoate on different levels of biological organisation in the non-biting midge Chironomus 684 riparius. Aquatic toxicology, 155, 236-243. 685 Vijver, M. G., Hunting, E. R., Nederstigt, T. A., Tamis, W. L., van den Brink, P. J., & van Bodegom, P. 686 M. (2017). Postregistration monitoring of pesticides is urgently required to protect ecosystems. 687 Environmental Toxicology and Chemistry, 36(4), 860-865. 688 Wang, J., Yin, R., Liu, Y., Wang, B., Wang, N., Xiao, P., Xiao, T. & Hirai, H. (2023). Meta-analysis of 689 neonicotinoid insecticides in global surface waters. Environmental Science and Pollution 690 Research, 30(1), 1039-1047. 691 Weichert, F. G., Floeter, C., Artmann, A. S. M., & Kammann, U. (2017). Assessing the ecotoxicity of 692 potentially neurotoxic substances–Evaluation of a behavioural parameter in the embryogenesis of 693 Danio rerio. Chemosphere, 186, 43-50. 694 Weis, J. S., Smith, G., Zhou, T., Santiago-Bass, C., & Weis, P. (2001). Effects of contaminants on 695 behavior: biochemical mechanisms and ecological consequences. Bioscience, 51(3), 209-217. 696 Woolway, R. I., Kraemer, B. M., Lenters, J. D., Merchant, C. J., O’Reilly, C. M., & Sharma, Sapna (2020) 697 Global lake responses to climate change. Nature Reviews Earth & Environment, 1. pp. 388-403. 698 ISSN 2662-138X doi: https://doi.org/10.1038/s43017-020-0067-5 699 Xia, J. G., Ma, Y. J., Guo, W. M., Huang, L., & Fu, S. J. (2015). Temperature-dependent effects of PFOS 700 on risk recognition and fast-start performance in juvenile Spinibarbus sinensis. Aquatic Biology, 701 24(2), 101-108. 702 Ågerstrand, M., Arnold, K., Balshine, S., Brodin, T., Brooks, B. W., Maack, G., McCallum, E. S., Pyle, 703 G., Saaristo, M. & Ford, A. T. (2020). Emerging investigator series: use of behavioural endpoints 704 in the regulation of chemicals. Environmental Science: Processes & Impacts, 22(1), 49-65. 705 706 707 708 Declaration of interests ☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☒ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Martina G. Vijver reports financial support was provided by European Union. Martina G. Vijver reports financial support was provided by Leiden Bio Science Park. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Conflict of Interest Figure 1 Click here to access/download;Figure;Figure 1.PNG Figure 2 Click here to access/download;Figure;Figure 2.PNG Figure 3 Click here to access/download;Figure;Figure 3.PNG Supplementary material A Click here to access/download Supplementary Material Supplementary information A.docx Supplementary material B Click here to access/download Supplementary Material Supplementary information B.xlsx