Stoichiometric Responses to an Agricultural ... - ACS Publications

Dec 12, 2016 - cues. Therefore, we exposed damselfly larvae to chlorpyrifos ... The way the predator cues modulated the pesticide effects strongly dif...
0 downloads 0 Views 665KB Size
Article pubs.acs.org/est

Stoichiometric Responses to an Agricultural Pesticide Are Modified by Predator Cues Lizanne Janssens,* Lin Op de Beeck, and Robby Stoks Laboratory of Aquatic Ecology, Evolution and Conservation, University of Leuven, Charles Deberiotstraat 32, B-3000 Leuven, Belgium S Supporting Information *

ABSTRACT: Current ecological risk assessment of pesticides fails to protect aquatic ecosystem health. To get better insight in how pesticides may affect aquatic ecosystems, we tested how sublethal pesticide concentrations modify body stoichiometry. Moreover, as interactions with natural stressors may cause underestimates of the impact of pesticides, we also tested whether this pathway depended on the presence of predator cues. Therefore, we exposed damselfly larvae to chlorpyrifos and cues from predatory dragonflies and focused on body stoichiometry and associated explanatory variables (growth rate, RNA:DNA, and energy storage molecules). The way the predator cues modulated the pesticide effects strongly differed between endpoints. Exposure to chlorpyrifos affected the key body stoichiometric ratios: chlorpyrifos consistently increased N:P, while its effects on C:N (decrease with predator cues) and C:P (increase without predator cues) strongly depended upon the presence of the natural stressor. These stoichiometric responses could be explained by associated changes in growth, RNA:DNA, and in C-rich fat and sugars and N-rich proteins. The observed changes in body stoichiometry may affect the damselflies’ food quality and have the potential to cascade through the food web and shape nutrient cycling.



INTRODUCTION Current ecological risk assessment of pesticides seems ineffective to protect freshwater ecosystems.1−3 This could be explained by the presence of unknown pathways of how sublethal pesticide effects may cascade through food chains and eventually generate effects at the ecosystem level. Furthermore, synergisms between stressors,4−6 whereby stressors enhance each other’s effects may further contribute to the underestimation of the impact of pesticide concentrations that current legislation considers environmentally protective. In this context, the stressor combination of pesticides and predator cues, which is widespread in aquatic ecosystems,7−9 is getting increased attention as this combination may drastically magnify the impact of sublethal pesticide concentrations.10−12 Stoichiometric responses to stressors are an understudied sublethal effect that may cascade up the food chain by changing the nutritional value of the prey for higher trophic levels. For example, copepods suffer from a reduced growth when feeding on algae with modified C:P and N:P ratios.13 Changes in body stoichiometry can also affect key ecosystem functions related to nutrient cycling and energy fluxes such as litter breakdown, primary production, and community respiration.14,15 For example, in a terrestrial ecosystem a 4% higher body C:N ratio of grasshopper carcasses due to predator exposure resulted in a 3-fold decrease in the mineralization of plant litter.16 Body stoichiometry may therefore provide an unexplored organismal trait explaining how pesticides shape ecosystem functions, a key challenge in ecological risk assessment.2,17 Yet, whether and © XXXX American Chemical Society

how pesticides affect body stoichiometry and whether this is mediated by predator cues is largely unknown. A powerful predictive framework to understand how stressors shape body stoichiometry is the general stress paradigm (GSP).14 The GSP predicts that in the presence of stress, animals will elevate their metabolism and allocate resources to defense mechanisms and maintenance including an increased production of C-rich biomolecules (partially from the breakdown of proteins). Therefore, the quantity of C that can be allocated to production of new tissues will be lower, which will also reduce the quantity of N and P that can be used for production. Since animals are limited in the ability to store excess inorganic elements and to maintain homeostasis, the excess of N and P will be excreted,15,18 resulting in increased body ratios of C:N and C:P.14 In line with the assumed generality of these stress-related responses, the changes in body stoichiometry predicted by the GSP have been shown in response to both predator cues15,19 and warming.20 The aim of this study was to examine the effects of a sublethal concentration of a widespread agricultural pesticide (chlorpyrifos) on the body stoichiometry of an aquatic insect and to investigate how these effects were modulated by exposure to predator cues. To obtain mechanistic insights Received: Revised: Accepted: Published: A

July 6, 2016 December 7, 2016 December 12, 2016 December 12, 2016 DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

avoid any bias due to larvae being associated with a specific container (set of conspecific larvae and predator), we randomly redistributed vials among containers of the same predation risk treatment on a daily basis.26 Throughout the exposure period, larvae were daily fed ad libitum with Artemia nauplii (mean daily dose ±1 SE: 1224 ± 108 nauplii, n = 10 daily doses). The number of larvae tested at each treatment combination was 65 (total of 260 larvae). In the treatment with chlorpyrifos present, larvae were exposed to 1 μg/L chlorpyrifos. The chosen concentration has a negative effect on larval growth rate (see results range finding in SI1, Supporting Information) and is within the range of chlorpyrifos concentrations reported in nature.27 Based on a pooled sample that combined the medium of 10 experimental vials, the initial chlorpyrifos concentration was 0.84 μg/L, and after 24 h (before renewal of the medium) the chlorpyrifos concentration decreased to 0.54 μg/L. Details on the chlorpyrifos treatment are given in SI1, Supporting Information. Predation risk was manipulated using visual and chemical predator cues, reflecting the cocktail of predator cues that damselfly larvae encounter in nature. To provide visual predator cues, a large Anax dragonfly larva (mean size: 3.59 ± 0.25 cm), important predators of damselfly larvae,28 was placed in the outer container of the treatment with predator cues. Additionally, larvae could see the conspecific larvae in the other vials in the container (damselfly larvae are cannibalistic).29 To avoid visual predator cues in the treatment without predation risk, the walls of these vials were made nontransparent using tape (this did not affect light levels in the vials). To provide chemical predator cues, we followed the method described in McPeek et al.30 and homogenized one E. cyathigerum larva in 20 mL of water from an aquarium filled with 300 mL of aged tap water in which a large Anax dragonfly larva had eaten one E. cyathigerum larva. As a result, this predator medium contained both predator kairomones and prey alarm cues, which are known to induce strong antipredator stress responses in aquatic organisms.31 We daily added 1 mL of this predator medium to each vial of the treatments with predator cues, and to the other vials we daily added 1 mL of aged tap water. Concentrations of NH3 and NO3 did not differ between treatments (SI4, Supporting Information). The NO2 concentration was higher in the vials to which predator cues were added. However, a separate test showed no decrease in growth rate in damselfly larvae exposed to this NO 2 concentration (t1, 18 = 1.70; p = 0.11), indicating that the observed effects of predation risk (see the Results section) cannot be explained by differences in NO2 concentration. Response Variables. To quantify growth rate, we weighed each larva to the nearest 0.01 mg at the start and at the end of the 7-day exposure period. Growth rate was calculated as [ln(final mass) − ln(initial mass)]/7 days.30 After determining final mass, the larvae were flash frozen using liquid nitrogen and stored at −80 °C. A detailed description of the assays used for the quantification of the physiological variables can be found in SI2, Supporting Information. Given that not all physiological variables could be measured on the same larva, we worked with two randomly chosen subsets of larvae. Whenever duplicate or triplicate readings were done, larval means were used for the statistical analyses. In a first set of larvae (20 per treatment combination, total of 80 larvae) we quantified RNA:DNA ratios fluorometrically based on the protocol by Vrede et al.32 We first measured the

underlying the single and interactive effects of both widespread stressors in aquatic ecosystems, we also studied key physiological variables related to the GSP. Building on the GSP,14 we predicted exposure to the pesticide to result in increased C and decreased N and P body contents, resulting in increased C:N and C:P, linked with a reduced growth rate and associated reduction in P-rich RNA (quantified as the RNA:DNA ratio, which is also a proxy for protein synthesis). Given their importance for shaping body stoichiometry,21 we also quantified the main energy storage molecules: N-rich proteins and C-rich fat and sugars. As the fundamental stress responses are general,14 we expected the pesticide and predator cues to evoke similar responses on growth, physiology, and stoichiometry. We further predicted that the effects of the pesticide would be stronger in animals exposed to predator cues given the widespread occurrence of synergistic interactions between both stressors.10,22,23 We studied this in larvae of the damselfly Enallagma cyathigerum for which we have shown synergistic interactions between pesticide exposure and predator cues for variables linked with oxidative stress.23 Damselflies are important intermediate predators in aquatic food webs, being predators of small invertebrates and prey for larger organisms, indicating that stressors that negatively affect the damselflies have the potential to disturb the whole ecosystem.24 Moreover, damselflies have a complex life cycle with the potential of carrying over stress effects experienced in the larval stage to the adult, terrestrial stage. The effects of the predator cues are partly discussed in an accompanying study on the effects of climate change in this species.25 The focus here will be on the pesticide effects and how these are modulated by the presence of predator cues, to get insight in this overlooked pathway how low sublethal pesticide levels may impair ecosystem health in natural water bodies.



METHODS Collecting and Housing. Eggs were obtained from 20 females collected in “Het Stappersven” (Kalmthout, Belgium) a shallow fishless lake with larvae of the dragonfly Anax imperator as top predator. This lake is situated in a large nature reserve (>100 ha) with no agriculture in a radius of >1.5 km. After hatching, larvae were placed individually in 200 mL cups at 20 °C and a photoperiod 14:10 L:D (light intensity ca. 3000 lx). Damselfly larvae were fed ad libitum with Artemia nauplii 5 days a week (mean daily dose ±1 SE: 673 ± 53 nauplii, n = 10 daily doses). Experimental Setup. To test for the single and combined effects of pesticide exposure and predator cues on growth rate, physiology, and body stoichiometry, we set up a full factorial experiment with two pesticide treatments (chlorpyrifos absent and present) × two predator cues treatments (predator cues absent and present). We exposed final instar larvae for 7 days to the stressors with daily renewal of the medium. In a separate experiment we showed that the renewal procedure did not affect the growth rate across 7 days (t test, t1, 18 = 1.16, p = 0.31). An exposure period of 7 days is long enough to evaluate the effects on growth rate (an important fitness-related variable), while still being a realistic exposure scenario in agricultural ponds where the study species occurs (for more details see SI1, Supporting Information). During the exposure period, larvae were placed individually in glass vials (100 mL) filled with 50 mL of synthetic pond water (for detailed information see SI1, Supporting Information). Sets of four vials were placed together in a larger outer container (750 mL). To B

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology total amount of RNA plus DNA and after degrading the RNA with RNase measured the total amount of DNA. Subtracting the amount of DNA from the total amount of RNA and DNA resulted in the RNA concentration in the samples. RNA and DNA concentrations were measured in triplicate. In a second set of larvae (15 per treatment combination, total of 60 larvae) we quantified energy storage levels and body elemental composition. Fat content was measured in triplicate following Janssens and Stoks.33 For total sugar content (glucose + glycogen), we used the protocol described in Stoks et al.34 based on the glucose kit of Sigma-Aldrich USA. Measurements were done in duplicate. The results for glucose and glycogen concentrations were very similar; therefore, we only report the total sugar content in the Results and Discussion sections. Protein content was measured in triplicate using the Bradford method.35 Fat content, total sugar content, and protein content were expressed as μg per mg dry mass (dry mass being derived from the total mass30). For the quantification of body elemental composition, we divided the homogenate in two parts: 1/4 of the sample was used for C and N analyses and 3/4 of the sample for P analyses. C and N content were quantified using an elemental analyzer (Carlo Erba 1108). For P analysis the samples were analyzed using inductively coupled plasma mass spectrometry (Agilent 7700x ICP-MS). C, N, and P concentrations were expressed as μg per mg dry mass, whereas C:N, C:P, and N:P were expressed as molar ratios. Statistical Analyses. We used two-way ANOVAs to test for the effects of pesticide exposure and predator cues for each response variable separately. Additionally, we performed two MANOVAs, the first one testing for the effects of the pesticide and predator cues on the energy reserves (fat, sugars, and proteins) and the second one testing for effects on the elemental body composition (C, N, and P). To be able to compare the responses to both stressors, we will also report the main effects of the predator cues (Figures 1−3; Table S3, Supporting Information), although these have been integrated in another paper on global warming.25 All tests were done in STATISTICA v12 (StatSoft, Tulsa, OK, U.S.A.). Values of p < 0.05 were considered significant. For all variables the assumptions of ANOVA (normally distributed errors and homogeneity of variances) were met without the need for transformations. To identify the type of interaction, whenever we detected an interaction between the stressors, we calculated the interaction effect size (estimated as Hedges’d) with its 95% confidence interval following Jackson et al.36 An interaction effect size larger than zero indicates a synergistic interaction, while an interaction effect size smaller than zero indicates an antagonistic or reversal interaction, with the latter occurring when the observed effect is opposite to the predicted combined additive effect. The results for these calculations are presented in SI5, Supporting Information.

Figure 1. Mean (A) growth rate and (B) RNA:DNA ratios of Enallagma cyathigerum larvae as a function of pesticide and predation risk exposure. Given are least-squares means +1 SE.

F1, 76 = 8.25; p = 0.0053) (Figure 1B; Table S3 Supporting Information). This indicates an antagonistic interaction between the two stressors as supported by the negative interaction effect size (Hedges’d = −1.57; 95% CI: [−1.97; −1.18]. The change in RNA:DNA ratio could be fully attributed to changes in the RNA concentration in animals exposed to the stressors, while the DNA concentration was unaffected (see SI6, Supporting Information). Overall, exposure to the pesticide and exposure to predator cues reduced the energy reserves and this in an interactive way (MANOVA: Pesticide: F3, 54 = 26.99; p < 0.001; Predator cues: F3, 54 = 19.65; p < 0.001; Pesticide × Predator cues: F3, 54 = 4.76; p = 0.0051; Figure 2; Table S3 Supporting Information). The pesticide-induced reduction in fat content was only present in the absence of predator cues (without predator cues: −28.7%, with predator cues: +3.4%) (Pesticide exposure × Predator cues, F1, 56 = 9.68; p = 0.0029, Figure 2A), and the pesticide-induced reduction in sugar content was stronger in the absence of predator cues (without predator cues: −28.0%, with predator cues: −16.8%) (F1, 56 = 8.12; p = 0.0061, Figure 2B). This indicates antagonistic interactions between the two stressors for both fat and sugar contents as supported by the negative interaction effect sizes (fat: Hedges’d = −18.97; 95% CI: [−22.27; −15.67]; sugars: Hedges’d = −6.59; 95% CI: [−7.43; −5.75]. The chlorpyrifos effect on the protein content (−17.5%) was not influenced by the predator cues (F1, 56 = 2.56; p = 0.11 Figure 2C).



RESULTS No animals died during the pesticide exposure period. Growth rate was lower in larvae exposed to the pesticide (−44.9%) (F1, 256 = 30.52; p < 0.001) and to predator cues (−32.7%) (F1, 256 = 13.93; p < 0.001). Chlorpyrifos and predator cues did not interact for growth rate (F1, 256 = 0.017; p = 0.88) (Figure 1A; Table S3 Supporting Information). Larvae exposed to chlorpyrifos had a lower RNA:DNA ratio, especially in the absence of predator cues (without predator cues: −35.7%, with predator cues: −18.0%) (Pesticide exposure × Predator cues, C

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

CI: [5.59; 12.65]). The negative pesticide effect on P was especially strong in the absence of predator cues (without predator cues: −36.8%, with predator cues: −22.5%) (antagonistic interaction: Hedges’d = −1.62; 95% CI: [−2.14; −1.10]). The pesticide-induced reduction in N (−13.2%) did not depend on exposure to the predator cues (Pesticide exposure × Predator cues: F1, 56 = 0.35; p = 0.56). The pesticide effect on the stoichiometric ratios was strongly dependent on predator cues for C:N (F1, 56 = 6.37; p = 0.014; reversal interaction: Hedges’d = −1.71; 95% CI: [−2.35; −1.06]). Larvae exposed to chlorpyrifos only had a lower C:N ratio (−30.5%) in the presence of predator cues (without predator cues: −2.4%) (Figure 3D; Table S3 Supporting Information). Also for C:P the pesticide and predator cues interacted (F1, 56 = 9.06; p = 0.0039). More specifically, we could show an antagonistic interaction (Hedges’d = −2.69; 95% CI: [−15.70; −9.67]) with the pesticide only resulting in a higher C:P ratio (+34.0%) in the absence of predator cues (with predator cues: +18.0%) (Figure 3E; Table S3 Supporting Information). Both exposure to chlorpyrifos (+28.2%, F1, 56 = 12.39; p < 0.001) and predator cues (+13.8%, F1, 56 = 4.61; p = 0.036) resulted in a higher N:P ratio, with no interaction between the two stressors (F1, 56 = 1.59; p = 0.21) (Figure 3F; Table S3 Supporting Information).



DISCUSSION

The applied sublethal dose of this widely used agricultural pesticide reduced growth rate, RNA:DNA ratios, and energy reserves and altered the elemental stoichiometry, mostly in a similar way as exposure to predator cues did. This extends the view that contaminant-induced mortality can often be similar to the lethal effects of predators,37 toward sublethal pesticide effects that may have as strong impact on populations than lethal effects.38 Moreover, for most of these variables the pesticide interacted with the predator cues, leading to both increased (synergistic interaction for C content, reversal interaction for C:N) and decreased (antagonistic interactions for P content and C:P ratio) stoichiometric responses to the pesticide when it was combined with predator cues. Interactive effects of pollutants and predator cues are increasingly reported;10 here we report the first evidence for such effects on body stoichiometry which has the potential to cascade up to the ecosystem level by changing nutrient fluxes.39 Growth, RNA:DNA, and Energy Reserves. Growth rate was lower in animals exposed to the pesticide, and this was unaffected by exposure to predator cues (Figure 1A). Chlorpyrifos-induced growth reductions are common39−41 and can be explained by a reduced food intake39 associated with the inhibition of acetylcholinesterase39,40,42 and a higher investment in detoxification and repair.40,43,44 These mechanisms may as well have caused the observed pesticide-induced reduction in energy storage in terms of total fat, total sugars, and protein contents (Figure 2). The pesticide-induced reduction in the RNA:DNA ratio (Figure 1B) may reflect the reduced growth rate and associated reduced production of new tissues (proteins) in the presence of the pesticide.45,46 The decreases in RNA:DNA ratio and energy reserves due to pesticide exposure were less strong (RNA:DNA and total sugars; Figure 1B and 2B) or even absent (total fat; Figure 2A) when animals were also exposed to predator cues. This smaller response in animals exposed to predator cues can be explained by the already lower baseline levels of these variables in the

Figure 2. Mean levels of energy storage molecules: (A) fat content, (B) sugar content, and (C) protein content of Enallagma cyathigerum larvae as a function of pesticide and predation risk exposure. Given are least-squares means +1 SE.

Overall, both exposure to the pesticide and to predator cues resulted in lower body contents of C, N, and P; moreover both stressors interacted (MANOVA: Pesticide: F3, 54 = 25.28; p < 0.001; Predator cues: F3, 54 = 17.87; p < 0.001; Pesticide × Predator cues: F3, 54 = 4.25; p = 0.0090; Figure 3A-C; Table S3 Supporting Information). The pesticide effect on elemental body content depended on predator cues for C (trend, Pesticide exposure × Predator cues, F1, 56 = 3.08; p = 0.084) and for P (F1, 56 = 9.94; p = 0.0041). The negative pesticide effect on C tended to be stronger in animals exposed to predator cues (without predator cues: −19.5%, with predator cues: −36.1%) (synergistic interaction: Hedges’d = 9.12; 95% D

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 3. Mean (A) C content, (B) N content, (C) P content, (D) C:N ratios, (E) C:P ratios, and (F) N:P ratios of Enallagma cyathigerum larvae as a function of pesticide and predation risk exposure. Given are least-squares means +1 SE.

to a stronger reduction in P-rich RNA), while the C content was much less affected. The pattern in P indeed closely followed the RNA:DNA ratio (Figure 1B) and to a large extent matched the negative effects of chlorpyrifos on growth rate and the associated synthesis of proteins.21 Finally, exposure to chlorpyrifos resulted in an increased N:P ratio (Figure 3F) which was not dependent on the presence of predator cues. This could be explained by a stronger reduction in P (Figure 3C; linked to the reduction in P-rich RNA) than in N (Figure 3B; linked to the reductions in N-rich proteins). General Stress Response or Stressor-Specific? When comparing the impact of the anthropogenic and the natural stressor in isolation, most of the observed responses were qualitatively similar: both exposure to the pesticide alone and exposure to the predator cues alone caused a decreased growth rate, RNA:DNA ratio, energy reserves, and to a large extent also body stoichiometry. This suggests that these stressor-induced responses might be general responses to stressors, although they did not fit the GSP framework.14 One exception was the C:N ratio: our results showed a decrease in C:N under chlorpyrifos exposure (only in the presence of predator cues) and an increase in C:N under predation risk, while both stressors reduced C and N, indicating that the relative reductions differed. The pesticide-induced reduction in C was stronger compared to the effect of the predator cues, while the decrease in N was less strong. We have no direct data to explain this less strong effect of the predator cues on the C content compared to the effect of the pesticide. One testable hypothesis is that this may be explained by an increase in molecules with a high C:N ratio in association with protective anticipatory responses evoked by predator cues. In response to predator cues, arthropods have been shown to increase the thickness of their exoskeleton (e.g., mayfly larvae,48 water fleas49) of which the major compound chitin is a polysaccharide with a high C:N (5:1).21 This is a specific morphological defense response to

presence of predator cues, thereby probably reaching lower threshold limits. Body Stoichiometry. Exposure to chlorpyrifos had considerable effects on all three body stoichiometric ratios which for C:N and C:P were strongly dependent upon the presence of predator cues, thereby providing a novel way how natural stressors may modify the impact of pesticides in natural systems. These stoichiometric responses to the pesticide did not follow the GSP which predicts increased C:N and C:P ratios (as a result of an increased metabolic rate and the allocation of resources to emergency functions away from the production of new tissues, see the Introduction).14 In contrast, the changes in body stoichiometry could for a large extent be explained by changes in the energy storage molecules and RNA:DNA. Chlorpyrifos only reduced the C:N ratio in the presence of predator cues (Figure 3D) as a result of stronger declines in C than in N. Indeed, the chlorpyrifos-induced decline of C (linked to the reduced C-rich sugar content) was much stronger in the presence (−36.1%) than in the absence (−19.5%) of predator cues (Figure 3A). In contrast, the chlorpyrifos-induced decline of N (linked to the decreased Nrich protein content) was as moderate in the presence (−13.2%) as in the absence (−13.7%) of predator cues (Figure 3B). A decrease in C:N has also been observed in Daphnia magna exposed to lindane and explained by less incorporation of carbon in structural biomass and higher use of carbon for the synthesis of compounds involved in detoxification, increased respiration, and oxidized products.47 Increased detoxification (e.g., increased activity of glutathione-S-transferase, increased expression levels of the stress protein Hsp70) and increased respiration have indeed been observed in E. cyathigerum when exposed to chlorpyrifos.40 Chlorpyrifos exposure increased the C:P ratio (Figure 3E) only in the absence of predator cues. This occurred because in the absence of predator cues the pesticide-induced reduction in P was especially strong (linked E

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology avoid being wounded during predator attacks.48,49 Further, as an anticipatory response to heal the cuticular wounds caused by nonlethal strikes of a dragonfly predator, odonate larvae have been shown to increase their levels of melanin, another molecule with a high C:N (9:1).50 These anticipatory defensive responses may potentially have partly counteracted the reduction in C content due to the decrease in fat and sugar content. Yet, this conjecture awaits experimental testing for our study species. Differences in the applied concentrations of the two stressors may also have shaped the differential strength of the reductions in C and N contents but likely not the patterns in their ratio. Relevance for Ecological Risk Assessment. The here observed pesticide-induced changes in the body stoichiometric ratios of the damselfly larvae may have two important and overlooked effects in natural communities that are of high relevance for ecological risk assessment. First, changed body stoichiometry alters the nutritional value (match between the elemental compositions of resources and consumers21) of the damselfly larvae for higher trophic levels such as fish and birds.51 The disturbed stoichiometric balance may result in lower growth rates of the predators as has, for example, been shown in an algae-copepod resource-consumer couple.13 It thereby contributes to an important, yet overlooked way how sublethal pesticide concentrations may generate trait-mediated indirect effects in natural communities.11 Second, the stoichiometric responses to pesticide exposure provide a novel way how pesticides may shape ecosystem functions. Pesticides are thought to affect ecosystem functions (such as litter breakdown, primary production, and community respiration)2 both directly by influencing the ecosystem properties (for example, pH) and indirectly by contaminant-induced declines in biodiversity.52 The mechanisms behind this are complex but are often linked with species declines at different trophic levels. 53,54 Our experiment identified a novel mechanism how pollutants can affect ecosystem functions related to nutrient transfer: through changing the body stoichiometry. Indeed, the disturbance of body stoichiometry can strongly affect nutrient cycling. For example, soil samples that received carcasses of grasshoppers with a 4% higher C:N body content due to exposure to predators showed a 3-fold decrease in the mineralization of plant litter.15 Pollutantinduced changes in body stoichiometry may therefore provide an important mechanism how sublethal pollutant concentrations may negatively affect food web structure and nutrient transfer. Because in natural communities compensatory responses may exist to counteract the effects of stoichiometric disturbances, research under field conditions will be, however, necessary. The observation under natural field conditions of predator-induced changes in body stoichiometry that translated into strong changes in plant mineralization rates15 highlights the potential for this mechanism to influence nutrient cycling in natural communities. Given that damselflies have a complex life cycle, these changes in body stoichiometric ratios in the aquatic larval stage will likely be transferred to the terrestrial adult stage,24 indicating that pesticides present in water bodies can influence the terrestrial ecosystem by disturbing the transfer of nutrients. This would add a novel mechanism to the research topic of how pollutants can influence the coupling of aquatic and terrestrial ecosystems by affecting the fluxes of organisms and nutrients between these ecosystems.55,56

To conclude, we here presented proof-of-principle that pollutants may change body stoichiometry and provided insights in the underlying mechanisms and highlighted the importance of integrating predation risk when evaluating the net impact of sublethal pesticide concentrations on body composition. Studies that include pesticide-induced changes in body stoichiometry may provide a better understanding on how sublethal pesticide concentrations may disturb ecosystem functioning and thereby may help to get a better understanding why current risk assessment of pesticides fails in protecting freshwater biodiversity and ecosystem health.1,3



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.6b03381. SI1, detailed information about the pesticide treatment; SI2, detailed protocols for the quantification of the physiological variables; SI3, effects of the pesticide exposure and predator cues on the RNA and DNA concentrations; Table S1, results of the different (M)ANOVAs testing for effects of pesticide exposure and predator cues on growth, physiology, and elemental stoichiometry; Table S2, concentration of NH3, NO2, and NO3 in the control treatment versus the predation risk treatment; Table S3, effect sizes for the pesticide exposure, predator cues, and their interactions (PDF)



AUTHOR INFORMATION

Corresponding Author

*Phone: +3216323857. Fax: +3216324575. E-mail: lizanne. [email protected]. ORCID

Lizanne Janssens: 0000-0001-9126-4682 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Marie Van Dievel and Ria Van Houdt who assisted during the experiment and three anonymous reviewers for their constructive comments. L.J. is a postdoctoral fellow of FWOFlanders; L.O.D.B. is a Ph.D. Fellow of IWT-Flanders. Financial support came from the Belspo project SPEEDY, KULeuven Excellence Center Financing PF/2010/07, and FWO research grant G.0943.15.



REFERENCES

(1) Beketov, M. A.; Kefford, B. J.; Schafer, R. B.; Liess, M. Pesticides reduce regional biodiversity of stream invertebrates. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 11039−11043. (2) Peters, K.; Bundschuh, M.; Schäfer, R. B. Review on the effects of toxicants on freshwater ecosystem functions. Environ. Pollut. 2013, 180, 324−329. (3) Malaj, E.; von der Ohe, P. C.; Grote, M.; Kuhne, R.; Mondy, C. P.; Usseglio-Polatera, P.; Brack, W.; Schafer, R. B. Organic chemicals jeopardize the health of freshwater ecosystems on the continental scale. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 9549−9554. (4) Darling, E. S.; Côté, I. M. Quantifying the evidence for ecological synergies. Ecol. Lett. 2008, 11, 1278−1286. (5) Bancroft, B. A.; Baker, N. J.; Blaustein, A. R. A meta-analysis of the effects of ultraviolet B radiation and its synergistic interactions with pH, contaminants and disease on amphibian survival. Conserv. Biol. 2008, 22, 987−996.

F

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology (6) Lindenmayer, D. B.; Likens, G. E.; Krebs, C. J.; Hobbs, R. J. Improved probability of detection of ecological “surprises. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 21957−21962. (7) Sih, A.; Bell, A. M.; Kerby, J. L. Two stressors are far deadlier than one. Trends Ecol. Evol. 2004, 19, 274−276. (8) Schwarzenbach, R. P.; Escher, B. I.; Fenner, K.; Hofstetter, T. B.; Johnson, C. A.; Von Gunten, U.; Wehrli, B. The challenge of micropollutants in aquatic systems. Science 2006, 313, 1072−1077. (9) Liess, M.; Schäfer, R. B.; Schriever, C. A. The footprint of pesticide stress in communities − species traits reveal community effects of toxicants. Sci. Total Environ. 2008, 406, 484−490. (10) Relyea, R. A.; Mills, N. Predator-induced stress makes the pesticide carbaryl more deadly to gray treefrog tadpoles (Hyla versicolor). Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 2491−2496. (11) Relyea, R.; Hoverman, J. Assessing the ecology in ecotoxicology: a review and synthesis in freshwater systems. Ecol. Lett. 2006, 9, 1157− 1171. (12) Teplitsky, C.; Piha, H.; Laurila, A.; Merilä, J. Common pesticide increases costs of antipredator defenses in Rana temporaria tadpoles. Environ. Sci. Technol. 2005, 39, 6079−6085. (13) Plum, C.; Hüsener, M.; Hillebrand, H. Multiple vs. single phytoplankton species alter stoichiometry of trophic interaction with zooplankton. Ecology 2015, 96, 3075−3089. (14) Hawlena, D.; Schmitz, O. J. Physiological stress as a fundamental mechanism linking predation to ecosystem functioning. Am. Nat. 2010, 176, 537−556. (15) Hawlena, D.; Schmitz, O. J. Herbivore physiological response to predation risk and implications for ecosystem nutrient dynamics. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 15503−15507. (16) Hawlena, D.; Strickland, M. S.; Bradford, M. A.; Schmitz, O. J. Fear of predation slows plant-litter decomposition. Science 2012, 336, 1434−1438. (17) Power, A. G. Ecosystem services and agriculture: tradeoffs and synergies. Philos. Trans. R. Soc., B 2010, 365, 2959−2971. (18) Polednik, L.; Rehulka, J.; Kranz, A.; Polednikova, K.; Hlavac, V.; Kazihnitkova, H. Physiological responses of over-wintering common carp (Cyprinus carpio) to disturbance by Eurasian otter (Lutra lutra). Fish Physiol. Biochem. 2008, 34, 223−234. (19) Van Dievel, M.; Janssens, L.; Stoks, R. Short-term and long-term behavioural, physiological and stoichiometric responses to predation risk indicate chronic stress and compensatory mechanisms. Oecologia 2016, 181, 347−357. (20) Schmitz, O. J. Global climate change and the evolutionary ecology of ecosystem functioning. Ann. N. Y. Acad. Sci. 2013, 1297, 61−72. (21) Sterner, R. W.; Elser, J. J. Ecological stoichiometry: the biology of elements from molecules to the biosphere; Princeton University Press: Princeton, 2002. (22) Campero, M.; Slos, S.; Ollevier, F.; Stoks, R. Sublethal pesticide concentrations and predation jointly shape life history: behavioral and physiological mechanisms. Ecol. Appl. 2007, 17, 2111−2122. (23) Janssens, L.; Stoks, R. Synergistic effects between pesticide stress and predator cues: conflicting results from life history and physiology in the damselfly Enallagma cyathigerum. Aquat. Toxicol. 2013, 132− 133, 92−99. (24) Stoks, R.; Cordoba-Aguilar, A. Evolutionary ecology of Odonata: a complex life cycle perspective. Annu. Rev. Entomol. 2012, 57, 249−265. (25) Janssens, L.; Van Dievel, M.; Stoks, R. Warming reinforces nonconsumptive predator effects on prey growth, physiology and body stoichiometry. Ecology 2015, 96, 3270−3280. (26) McPeek, M. A. The growth/predation risk trade-off: so what is the mechanism? Am. Nat. 2004, 163, E88−E111. (27) Schulz, R. Field studies on exposure, effects and risk mitigation on aquatic nonpoint-source insecticide pollution. J. Environ. Qual. 2004, 33, 419−448. (28) Stoks, R.; De Block, M.; McPeek, M. A. Alternative growth and energy storage responses to mortality threats in damselflies. Ecol. Lett. 2005, 8, 1307−1316.

(29) De Block, M.; Stoks, R. Cannibalism-mediated life history plasticity to combined time and food stress. Oikos 2004, 106, 587− 597. (30) McPeek, M. A.; Grace, M.; Richardson, J. M. L. Physiological and behavioral responses to predators shape the growth/predation risk trade-off in damselflies. Ecology 2001, 82, 1535−1545. (31) Ferrari, M. C. O.; Wisenden, B. D.; Chivers, D. P. Chemical ecology of predator-prey interactions in aquatic ecosystems: a review and prospectus. Can. J. Zool. 2010, 88, 698−724. (32) Vrede, T.; Persson, J.; Aronsen, G. The influence of food quality (P:C ratio) on RNA:DNA ratio and somatic growth of Daphnia. Limnol. Oceanogr. 2002, 47, 487−494. (33) Janssens, L.; Stoks, R. Reinforcing effects of non-pathogenic bacteria and predation risk: from physiology to life history. Oecologia 2014, 176, 323−332. (34) Stoks, R.; De Block, M.; McPeek, M. A. Physiological costs of compensatory growth in a damselfly. Ecology 2006, 87, 1566−1574. (35) Bradford, M. M. A rapid and sensitive method for the quantification of microgram quantities of protein, utilizing the principle of protein-dye landing. Anal. Biochem. 1976, 72, 248−254. (36) Jackson, M. C.; Loewen, C. J. G.; Vinebrooke, R. D.; Chimimba, C. T. Net effects of multiple stressors in freshwater ecosystems: a meta-analyses. Glob. Change Biol. 2016, 22, 180−189. (37) Rohr, J. R.; Kerby, J. L.; Sih, A. Community ecology as a framework for predicting contaminant effects. Trends Ecol. Evol. 2006, 21, 606−613. (38) Kö hler, H. R.; Triebskorn, R. Wildlife ecotoxicology of pesticides: can we track effects on the population level and beyond? Science 2013, 341, 759−765. (39) Dinh Van, K.; Janssens, L.; Debecker, S.; Stoks, R. Temperatureand latitude-specific individual growth rates shape the vulnerability of damselfly larvae to a widespread pesticide. J. Appl. Ecol. 2014, 51, 919− 928. (40) Janssens, L.; Stoks, R. Exposure to a widespread non-pathogenic bacterium magnifies sublethal pesticide effects in the damselfly Enallagma cyathigerum: from the suborganismal level to fitness-related traits. Environ. Pollut. 2013, 177, 143−149. (41) Widder, P. D.; Bidwell, J. R. Cholinesterase activity and behavior in chlorpyrifos exposed Rana sphenocephala tadpoles. Environ. Toxicol. Chem. 2006, 25, 2446−2454. (42) Xuereb, B.; Lefevre, E.; Garric, J.; Geffard, O. Acetylcholinesterase activity in Gammarus fossarum (Crustacea Amphipoda): Linking AChE inhibition and behavioural alteration. Aquat. Toxicol. 2009, 94, 114−122. (43) Cinzia, A.; Baldracchini, F.; Piazzoli, A.; Frosini, R.; Talesa, V.; Elvio, G. Activity changes of glyoxalase system enzymes and glutathione-S-transferase in the bivalve mollusk Scapharca inaequivalvis exposed to the organophosphate chlorpyrifos. Pestic. Biochem. Physiol. 2006, 86, 72−77. (44) Scheil, V.; Zürn, A.; Köhler, H. R.; Triebskorn, R. Embryo development, stress protein (Hsp70) responses and histopathology in zebrafish (Danio rerio) following exposure to nickel chloride, chlorpyrifos and binary mixtures of them. Environ. Toxicol. 2010, 25, 83−93. (45) Barron, M. G.; Adelman, I. R. Nucleic acid, protein content and growth of larval fish sublethally exposed to various toxicants. Can. J. Fish. Aquat. Sci. 1984, 41, 141−150. (46) Revankar, P. R.; Shyama, S. K. Genotoxic effects of monocrotophos, an organophosphorous pesticide, on an estuarine bivalve. Food Chem. Toxicol. 2009, 47, 1618−1623. (47) Ek, C.; Karlson, A. M. L.; Hansson, S.; Garbaras, A.; Gorokhova, E. Stable isotope composition in Daphnia is modulated by growth, temperature and toxic exposure: implications for trophic magnification factor assessment. Environ. Sci. Technol. 2015, 49, 6934−6942. (48) Dahl, J.; Peckarsky, B. Induced morphological defenses in the wild: predator effects on a mayfly, Drunella coloradensis. Ecology 2002, 83, 1620−1634. (49) Rabus, M.; Sollradl, T.; Clausen-Schaumann, H.; Laforsch, C. Uncovering ultrastructural defences in Daphnia magna - an G

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology interdisciplinary approach to assess the predator-induced fortification of the carapace. PLoS One 2013, 8, e67856. (50) Duong, T. M.; McCauley, S. J. Ecology 2016, 97, 1605−1610. (51) Corbet, P. S. Dragonflies: behavior in ecology of Odonata; Cornell University Press: Ithaca, NY, USA, 1999. (52) McMahon, T. A.; Halstead, N. T.; Johnson, S.; Raffel, T. R.; Romansic, J. M.; Crumrine, P. W.; Rohr, J. R. Fungicide-induced declines of freshwater biodiversity modify ecosystem functions and services. Ecol. Lett. 2012, 15, 714−722. (53) Duffy, J. E.; Cardinale, B. J.; France, K. E.; McIntyre, P. B.; Thebault, E.; Loreau, M. The functional role of biodiversity in ecosystems: incorporating trophic complexity. Ecol. Lett. 2007, 10, 522−538. (54) Reiss, J.; Bridle, J. R.; Montoya, J. M.; Woodward, G. Emerging horizons in biodiversity and ecosystem functioning research. Trends Ecol. Evol. 2009, 24, 505−514. (55) Schulz, R.; Bundschuh, M.; Gergs, R.; Bruhl, C. A.; Diehl, D.; Entling, M. H.; Fahse, L.; Frohr, O.; Jungkunst, H. F.; Lorke, A.; Schafer, R. B.; Schaumann, G. E.; Schwenk, K. Review on environmental alterations propagating from aquatic to terrestrial ecosystems. Sci. Total Environ. 2015, 538, 246−261. (56) Soininen, J.; Bartels, P.; Heino, J.; Luoto, M.; Hillebrand, H. Toward more integrated ecosystem research in aquatic and terrestrial ecosystems. BioScience 2015, 65, 174−182.

H

DOI: 10.1021/acs.est.6b03381 Environ. Sci. Technol. XXXX, XXX, XXX−XXX