Portal de Periódicos da CAPES

The interplay of past and current stress exposure on the water flea Daphnia

2011; Wiley; Volume: 25; Issue: 5 Linguagem: Inglês

10.1111/j.1365-2435.2011.01869.x

1365-2435

Mieke Jansen, Luc De Meester, Anke Cielen, Claudia C. Buser, Robby Stoks,

Fish Ecology and Management Studies

Functional EcologyVolume 25, Issue 5 p. 974-982 Free Access The interplay of past and current stress exposure on the water flea Daphnia Mieke Jansen, Corresponding Author Mieke Jansen Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, Belgium E-mail: [email protected]Search for more papers by this authorLuc De Meester, Luc De Meester Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this authorAnke Cielen, Anke Cielen Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this authorClaudia C. Buser, Claudia C. Buser EAWAG, Aquatic Ecology, Ueberlandstrasse 133, P.O. box 611, 8600 Dübendorf, Switzerland Institute of Integrative Biology, ETH Zürich, Universitaetsstrasse 16, 8092 Zürich, SwitzerlandSearch for more papers by this authorRobby Stoks, Robby Stoks Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this author Mieke Jansen, Corresponding Author Mieke Jansen Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, Belgium E-mail: [email protected]Search for more papers by this authorLuc De Meester, Luc De Meester Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this authorAnke Cielen, Anke Cielen Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this authorClaudia C. Buser, Claudia C. Buser EAWAG, Aquatic Ecology, Ueberlandstrasse 133, P.O. box 611, 8600 Dübendorf, Switzerland Institute of Integrative Biology, ETH Zürich, Universitaetsstrasse 16, 8092 Zürich, SwitzerlandSearch for more papers by this authorRobby Stoks, Robby Stoks Laboratory of Aquatic Ecology and Evolutionary Biology, Katholieke Universiteit Leuven, Ch. Deberiotstraat 32, 3000 Leuven, BelgiumSearch for more papers by this author First published: 20 May 2011 https://doi.org/10.1111/j.1365-2435.2011.01869.xCitations: 24AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Summary 1. Natural populations are exposed to multiple stressors, including both anthropogenic challenges such as xenobiotics and natural stressors associated with exposure to parasites and predators. While there is increasing concern and interest in the combined impact of current exposure to multiple stressors, little attention has been given to how past exposure to a stressor and its evolutionary response shapes the effects of current stressors. 2. Here, we performed a life-table experiment using the water flea Daphnia magna to study combined effects of current exposure to the pesticide carbaryl, parasite spores and fish predation risk and how these effects depend upon past exposure to carbaryl using clones obtained from a previous carbaryl selection experiment. 3. The current exposure to all three treatments affected life-history traits. Exposure to fish kairomones increased intrinsic population growth rate, while carbaryl and parasite exposure decreased this fitness measure. The three treatments interacted only in a few cases: carbaryl and fish kairomone exposure interacted in shaping intrinsic population growth rate and its component individual reproductive performance, yet the latter only in the animals not exposed to carbaryl stress in the past. 4. Our data revealed not only adaptive evolution of carbaryl resistance but also associated evolutionary costs in terms of reduced resistance to parasites, corroborating results of an earlier study. Importantly, both the evolutionary benefits and costs of past exposure to carbaryl stress were conditional on current environmental conditions, exposure to predation risk and parasites, respectively. 5. The emerging pattern showed that past stress interacted with current stress in shaping life history. Such evolution-driven carry-over effects across generations have been often ignored and may complicate the prediction of effects of current exposure to single and combined stressors even long after the past stress has disappeared. Introduction Natural populations are confronted with multiple stressors (van Straalen 2003). There is growing concern and scientific interest in the combined effects of multiple stressors because of accumulating evidence that they may interact, thereby complicating predictions of their joint effects (Sala et al. 2000; Darling & Côté 2008). These interactions may be beneficial when antagonistic interactions between stressors cancel out some of the negative effects of individual stressors, yet often interactions are synergistic and amplify the negative effects of individual stressors (Folt et al. 1999; Relyea & Mills 2001; Coors & De Meester 2008; Darling & Côté 2008). Because natural populations increasingly suffer human impact, there is special concern as to how anthropogenic stressors like xenobiotics interact with natural stressors like predators and parasites (Sih, Bell & Kerby 2004). While the focus of research on multiple stressors is on interactions between current stressors, organisms may have undergone selection by past exposure to a stressor, and current stressors may therefore also interact with such a past stressor. Studying such interactions is very important as they may shape the impact of single current stressors in unforeseen ways generating patterns that cannot be explained when ignoring the impact of the past stressor. This should be especially true for past stressors that are strong selective forces, which is the case for xenobiotics like pesticides that may induce high mortality rates on populations of non-target organisms. Pesticides often impose strong selection pressures leading to resistance (e.g. Palumbi 2001; Arnaud & Haubruge 2002; Gassmann, Onstad & Pittendrigh 2009). Adaptive genetic responses, with selection leading to a reduced susceptibility to the stressor, be it xenobiotics (Arnaud & Haubruge 2002) or natural stressors such as predators (Cousyn et al. 2001) and parasitism (Williams, Moore & Robbins 1990), are clear examples of how past exposure to a stressor can shape the response to current exposure to the same stressor. Yet, less-studied types of interactions between past and current stressors may occur where resistance to a past stressor negatively affects the ability of organisms to deal with other current stressors, so-called evolutionary costs of resistance (e.g. Berticat et al. 2002; Duron et al. 2006). For example, resistance against organophosphorous pesticides in mosquitoes resulted in increased vulnerability to current exposure to the biotic stressor Wolbachia (Duron et al. 2006). To what extent these interactions between a past and a certain current stressor in the form of evolutionary benefits (resistance) or costs depend themselves on interactions with other current stressors is largely unknown and asks for complicated study designs where animals selected with regard to a past stressor are exposed to combinations of current stressors. Such studies will not only explore the context dependence of evolutionary benefits and costs but also to what extent such interactions with a past stressor should be taken into account when evaluating the impact of exposure to single or combined current stressors. The zooplankter Daphnia magna (Straus, 1820) is an often used model organism in evolutionary ecology (Lampert 2006), general stress ecology (e.g. Pauwels, Stoks & De Meester 2005, 2010) and ecotoxicology (Walker, Hopkin & Peakall 2006) and therefore a good candidate to explore interactions between past and current stressors. Predatory fish and parasites are two of the most important and widely spread natural enemies of Daphnia (Kerfoot & Sih 1987; Ebert 2005). With regard to parasites, the effects of the widespread bacterial endoparasite Pasteuria ramosa (Metchnikoff) in particular have been studied extensively (e.g. Ebert et al. 1996; Ebert 2005; Jansen et al. 2010). As Daphnia inhabit ponds often surrounded by farm land, they are passively exposed to run-off from pesticide sprays (Declerck et al. 2006; Coors et al. 2009). Daphnia populations are therefore regularly exposed to mixtures of xenobiotics, predation risk and parasites, and because Daphnia are a keystone species in aquatic ecosystems (Lampert 1987), the study of such interactions has particular ecological relevance. Following the exciting work by Relyea & Mills (2001), interactions with the pesticide carbaryl have been studied in detail and revealed synergistic interactions between current exposure to carbaryl and to the parasite P. ramosa in D. magna (Coors & De Meester 2008; Coors et al. 2008; Jansen et al. 2011). Further, predation risk by phantom midges showed independent interactions with parasite challenge and current carbaryl exposure (Coors & De Meester 2008). In a previous study, we demonstrated evolved resistance after past carbaryl exposure as well as an evolutionary cost of this evolved resistance, with carbaryl-selected animals being more vulnerable when currently exposed to the parasite P. ramosa (Jansen et al. 2011). In this former study, however, we did not include fish predation risk, so we could not evaluate to what extent evolutionary benefits (increased resistance against carbaryl) and costs (decreased resistance against parasites) of carbaryl selection in the past are themselves context dependent, i.e. are modulated by current exposure to this other key biotic stressor. In the present study, we explore how current exposure to the pesticide carbaryl interacts with both key biotic stressors of Daphnia, exposure to parasite spores and to fish predation risk, and evaluate whether this interaction depends upon past exposure to carbaryl. This was carried out in a full-factorial experiment under all combinations of past exposure to carbaryl in a selection experiment and current exposure to pesticide, parasite and fish predation risk in a life-table experiment. Unlike our previous study (Jansen et al. 2011), this will enable us to test the context dependence of evolutionary benefits and costs of carbaryl resistance and to what extent such interactions with a past stressor should be taken into account when evaluating the impact of exposure not only to a single current stressor but also to mixtures of current stressors. Despite the increasing interest in effects of combined stressors, the dependency of their combined effects on a past stressor has not been explored before. Materials and methods Study population and pesticide For this study, we worked with a natural D. magna population from Oud-Heverlee (central Belgium, 50°50′22′′N, 4°39′18′′E). This population was chosen from a large set of 10 ponds that differ in land use in the immediate surroundings (see Coors et al. 2009) and within this set was one of the four populations that was used to set up a life-table experiment under all combinations of exposure to carbaryl and parasite spores (Jansen et al. 2011). The latter study revealed in all four populations a synergistic interaction between current exposure to carbaryl and to parasite spores, an increased resistance to carbaryl after selection, and an evolutionary cost of increased resistance in terms of increased susceptibility to parasites. We chose the Oud-Heverlee population for current extension of the Jansen et al. (2011) study because it showed high parasite infection rates. The carbarmate pesticide carbaryl is a model pesticide in ecotoxicology (Walker, Hopkin & Peakall 2006). It operates by inhibiting acetylcholinesterase, thereby preventing break down of acetylcholine, the chemical messenger responsible for transmitting the nervous signal through the synaptic cleft. The resulting build-up of acetylcholine will over-stimulate the postsynaptic receptors, and muscles will no longer be able to contract or relax in response to a nerve stimulus (Walker, Hopkin & Peakall 2006). Carbaryl selection experiment The preceding carbaryl selection experiment (summer of 2006, unpublished results) resulted in three types of experimental groups derived from each of the 10 natural populations: the start-control group, which was directly hatched from the natural dormant egg bank and represents the naturally available genetic variation within the pond; the container-control group, which was derived from the containers that were not exposed to carbaryl; and the carbaryl-selected group, which was derived from the containers that were exposed to carbaryl (Fig. 1). Both the start-control group and container-control group did not undergo carbaryl selection, yet the latter group may have undergone selection imposed by the container environment. Figure 1Open in figure viewerPowerPoint Schematic overview of the experimental design. In a first step, animals hatched from the egg bank of Oud-Heverlee population were exposed to carbaryl (past stressor) in a selection experiment in outdoor containers (Coors et al. in prep). In a second step, the isolates sampled from the containers were exposed to combinations of carbaryl, parasites and fish kairomones (current stressors) in a life-table experiment. The selection experiment was started with a standardized set of D. magna clones derived from the natural ponds. From each of the natural ponds, clonal lineages were hatched from the dormant egg bank, representing unique clones as they were produced by sexual reproduction. A total of 40 containers with a volume of 225 L were stocked with juvenile offspring from these clonal lineages by inoculating in a split-brood design 125 clones per population in a total volume of 180 L exposure medium. For each natural population, there were two replicate containers per selection treatment (presence or absence of carbaryl selection). In the containers with carbaryl selection, three pulses of carbaryl (32 μg L−1) were given every fourteen days, starting 24 days after inoculating the containers with juvenile D. magna. Twenty-six days after the last carbaryl pulse, D. magna females were isolated from the containers and individually cultured as clonal lineages without pesticide in the laboratory. Because the start-control and container-control groups did not differ in their responses to carbaryl exposure and parasite exposure (Jansen et al. 2011), we only included the start-control and the carbaryl-selected group associated with Oud-Heverlee in the current life-table experiment (Fig. 1). From both population types, we selected four females that were kept in the laboratory as clonal cultures for several generations (44 months) prior to the life-table experiment. Allozyme patterns for four different markers (PGI, MPI, MDH and AAT) of all isolates indicated that they differed in their multilocus genotypes, suggesting we worked with eight different clones in total. Clonal isolates were individually grown in 100 mL jars for two generations to minimize maternal effects prior to the life-table experiment. Aachener Daphnien Medium (ADaM, Klüttgen et al. 1994) with a conductivity of 300 μS was used to culture the clones in a temperature room at 20 °C and a 16 h light/8 h dark photoperiod. Before maturation, animals were fed 100 × 105 cells Scenedesmus obliquus daily, after which food levels were doubled to keep food levels ad libitum. Life-table experiment For each of the eight clonal isolates obtained from the selection experiment, we quantified key life-history traits in the life-table experiment under all eight combinations of the presence and absence of exposure to carbaryl, parasite spores and fish predation risk (Fig. 1). We ran four replicates per exposure combination for each clone resulting in a total of 8 clones × 8 exposure combinations × 4 replicates = 256 experimental units (vials). The experiment was started with second clutch juveniles <24 h old. Sets of sixty juveniles (1 Daphnia/1·6 mL water) from every clone were collectively exposed in a 500 mL glass vial to one of the eight treatment combinations. For exposure to carbaryl (1-napthyl methylcarbarmate, CAS 63-25-2, purity 99·8%, Sigma-Aldrich, Bornem, Belgium), we chose as nominal concentration 8 μg L−1 (measured concentration 7·75 μg L−1) that is sublethal and generates a synergistic effect with parasite exposure (Coors et al. 2008). In all treatments, the ethanol concentration, used as solvent control in the non-carbaryl treatments, was 0·05 μL L−1. For exposure to P. ramosa, we made a spore solution (375 × 102 mature P. ramosa spores per mL) as described in Jansen et al. (2010). Briefly, we collected P. ramosa infected D. magna, squashed and filtered them (nylon, mesh size 60 μm) before diluting them with distilled water to a final concentration of 50 × 105 spores per mL. Infected animals to generate spore solutions were obtained by infecting one clone (M10, Cousyn et al. 2001) with parasites collected as spores from two different ponds (Knokke In and Knokke Nat; see Coors et al. 2009). Grounded up healthy Daphnia were used as placebo in the treatments without parasites. For exposure to fish predation risk, fish-conditioned ADaM medium (300 μS) with fish kairomones was obtained by filtering the water (mesh size 0·450 μm) drawn from an aquarium where three Leuciscus idus (8 cm) were kept for 24 h. This fish species is not present in the Oud-Heverlee pond, but previous experiments showed the typical fish-induced responses using the kairomones from those fish in Daphnia (e.g. Cousyn et al. 2001; Pauwels, Stoks & De Meester 2005, 2010). To reach a final concentration of three fishes per 100 L water, this filtrate was five times diluted by daily inoculating 80 mL of freshly prepared fish-conditioned medium on a total amount of 400 mL in every jar of the fish-exposure treatment. Throughout the experiment, the fishes were fed Daphnia in a separate bucket to minimize the supplementation of nutrients or alarm substances to the experimental Daphnia. In the treatment combinations without predation risk, we did the same except for the fact we used 80-mL freshly diluted ADaM medium. Experimental cultures were daily fed 20 × 106 cells S. obliquus (33 × 104 per capita). On day 4, 1 h after renewing the medium, 15 individuals were transferred to 400 mL fresh ADaM and kept at the same exposure treatment combination to investigate life-history traits for another 17 days. The other 45 individuals in the jar were stored in RNA later® and flash-frozen in liquid nitrogen to perform genomic analyses (unpublished data). Daily food ratios were increased to 80 × 106 cells S. obliquus per vial from day 4 onwards. Every day juveniles were removed and 20% fresh fish-conditioned ADaM or 20% fresh ADaM medium was added. We completely renewed the medium every third day. During the 21-day life-table experiment, several key life-history traits were scored: the amount of juveniles born, the time at which they were born, and mortality. In addition, for the animals exposed to parasite spores, infection rates with immature and mature spores (proportion females infected with immature and mature spores, respectively) and infection intensity with mature spores (mean number of mature spores per infected female) were quantified on day 21. To avoid pseudoreplication, each of these traits was quantified as a mean or proportion per vial. Based on the first three traits, we calculated the intrinsic population growth rate 'r' using the Euler-Lotka equation (Stearns 1992). Additionally, we calculated individual reproductive performance (Van Doorslaer et al. 2009) using a modified version of the Euler-Lotka equation that does not take mortality into account, and where we used the total amount of offspring born on day x, divided by the total amount of living females (sterilized and non-sterilized) on day x for x going from day 4 to 21 (no juveniles were born before day 4). As this variable does not include mortality, it thus provides an estimate of individual performance related to reproduction only. This resulted in information on the intrinsic population growth rate 'r' and its two components: mortality and individual reproductive performance. Three infection-related variables were quantified at day 21 in the animals exposed to parasite spores. Daphnia infected with spores can be easily detected as they get sterilized (so show empty brood pouches), are darkish-red, non-transparent and get a larger body size (Ebert 2005). These infected females were individually squashed in 300 μL distilled water, and the developmental stage of the spores was scored at 400× magnification (phase-contrast microscope). The development of P. ramosa spores takes several days, and typically mature spores are present after 21 days. Immature spores are visible as cauliflowers which grow and develop further, turn into grape-like seeds and finally form spherical mature spores (Ebert 2005). We separately quantified the proportion of females per vial infected with immature and mature spores. After 21 days, nearly 100% of the Daphnia from all treatment combinations involving exposure to parasite spores were infected, making it meaningful to discriminate between animals with immature and mature spores. A higher infection rate with immature spores at day 21 in a given group would indicate that animals of this group were able to delay spore maturation, hence are more resistant against the parasite. Additionally, we quantified infection intensities, defined as the average number of mature spores per infected individual within one vial. Mature spores were counted from five infected females per vial using a Bürker counter (Bürker, Marienfield, Germany). Statistical analyses We tested for effects of the outdoor selection treatment (carb_sel: start-control group or carbaryl-selected group) and the laboratory exposure treatments, exposure to carbaryl (carb_lab), to parasites (par_lab) and to fish kairomones (fish_lab), on the different life-history traits using separate general linear models in STATISTICA v9. To take into account that we had several observations per clone, clone was nested in the carbaryl selection treatment, and all interactions with clone were included in the models. Yet, as our aims were not related to clonal variation and to keep the manuscript focused, we do not report results on this factor. The three infection-related response variables (infection rates of females with immature spores, and infection rates and infection intensities with mature spores) were analysed only for treatment combinations exposed to parasite spores. Mortality was log-transformed. Prospective power analyses were carried out using the power analysis module of STATISTICA v9. The power associated with medium and large effect sizes (according to Cohen 1988) is given in Table S1. Results Intrinsic population growth rate and its components Intrinsic population growth rate was affected by all three treatments (fish predation risk, parasite and carbaryl exposure; Table 1, Fig. 2a). While intrinsic population growth rate was slightly higher under fish predation risk in the absence of other stressors, it was considerably lower under exposure to parasites and to carbaryl. The opposite effects (with different strength) of fish predation risk and carbaryl exposure cancelled out when combined (carb_lab × fish_lab, Fig. 2a). Another way of looking at this interaction is that the effect of predation risk was weak in the absence of other stressors, while it was reasonably strong when the risk was added on top of other stressors (Fig. 2a). Table 1. Results of general linear models testing for effects of past exposure to carbaryl stress in the outdoor carbaryl selection (carb_sel), and the current laboratory exposures to carbaryl (carb_lab), to parasite spores (par_lab) and to fish kairomones (fish_lab) on the intrinsic population growth rate 'r', its two components, mortality (log transformed) and individual reproductive performance, and three infection-related variables: infection rates with immature spores, infection rates with mature spores and infection intensity with mature spores Intrinsic population growth rate Mortality (log transformed) Individual reproductive performance Infection rate with immature spores Infection rate with mature spores Infection intensity with mature spores d.f. F d.f. F d.f. F d.f. F d.f. F d.f. F carb_sel 1182 0·56 1182 14·20*** 1182 2·46 190 4·2* 190 5·87** 159 6·96* carb_lab 1182 11·63*** 1182 13·83*** 1182 10·45** 190 0 190 2·2 159 0·6 par_lab 1182 90·07*** 1182 0·39 1182 41·90*** fish_lab 1182 28·79*** 1182 20·08*** 1182 2·72 190 0·27 190 0·2 159 2·84 carb_sel × carb_lab 1182 0·01 1182 0·62 1182 2·63 190 0·42 190 1·82 159 0·25 carb_sel × par_lab 1182 0·53 1182 1·58 1182 3·09 carb_sel × fish_lab 1182 0·15 1182 0·51 1182 2·39 190 0·5 190 0·1 159 0·1 carb_lab × fish_lab 1182 4·40* 1182 0·37 1182 2·77 190 0·05 190 1·15 159 2·08 fish_lab × par_lab 1182 0·52 1182 0·37 1182 0·05 carb_lab × par_lab 1182 0·02 1182 0·30 1182 0·15 carb_lab × fish_lab × par_lab 1182 0·02 1182 0·17 1182 0·96 carb_sel × carb_lab × fish_lab 1182 0·91 1182 0·35 1182 5·90* 190 0 190 1·45 159 0·09 carb_sel × carb_lab × par_lab 1182 0·44 1182 0·31 1182 0·77 carb_sel × fish_lab × par_lab 1182 0·57 1182 0·02 1182 2·65 carb_sel × par_lab × fish_lab × carb_lab 1182 1·02 1182 2·28 1182 0·05 The significance of the F-values is indicated as follows: *P < 0·05, **P < 0·01 and ***P < 0·001. Figure 2Open in figure viewerPowerPoint Average (±1 SE) intrinsic population growth rate (a), and its two components, mortality (% dead individuals per vial) (b) and individual reproductive performance (c), in response to past carbaryl stress in the outdoor carbaryl selection treatment (start-control group vs. carbaryl-selected group) and current stress associated with the three laboratory exposure treatments: exposure to carbaryl, to parasite spores and to fish kairomones. While laboratory exposure to fish kairomones reduced mortality, exposure to parasite spores increased mortality (Table 1, Fig. 2b). Carbaryl exposure in the laboratory did not affect mortality. Animals that underwent carbaryl selection had an overall higher mortality than those of the start-control group. None of the interactions between stressors were significant (all P > 0·05). Individual reproductive performance was lower under exposure to parasite spores (Table 1, Fig. 2c). The effect of carbaryl exposure changed after carbaryl selection, and this was further modulated by the presence of fish kairomones (carb_sel × carb_lab × fish_lab, Table 1, Fig. 2c). In the absence of fish kairomones, carbaryl exposure reduced individual reproductive performance in the start-control group, but not in the carbaryl-selected animals. In the presence of fish kairomones, carbaryl exposure did not affect performance in any selection group. Another way of looking at this three-way interaction is that in the start-control group, exposure to carbaryl reduced performance and that this effect disappeared when also fish kairomones were added, while in the carbaryl-selected group, carbaryl did not reduce performance irrespective of the presence of fish kairomones. Infection-related variables None of the infection-related variables was affected by any of the three current stressors or their interactions (all P > 0·05, Table 1, Fig. 3). Carbaryl-selected animals had lower infection rates with immature spores and higher infection intensities and infection rates with mature spores compared to animals of the start-control group that were not exposed to carbaryl in the past. Figure 3Open in figure viewerPowerPoint Average (±1 SE) values of the three infection-related variables, infection rate with immature spores (a), infection rate with mature spores (b) and infection intensity with mature spores (c), in response to past carbaryl stress in the outdoor carbaryl selection treatment (start-control group vs. carbaryl-selected population) and current stress associated with the three laboratory exposure treatments: exposures to carbaryl, to parasite spores and to fish kairomones. Discussion All three treatments tested, exposure to fish kairomones, parasites and carbaryl, affected life history, but interacted only in a few cases: carbaryl and fish exposure interacted with each other in shaping intrinsic population growth rate and its component individual reproductive performance, yet the latter only in the animals not exposed to carbaryl stress in the past. The paucity of detected interactions may be partly because of the fact that our analyses only had considerable power to detect large effect sizes (Table S1), and the fact that all tests were run at non-stressful food levels (e.g. Marden et al. 2003). Despite this, our data revealed not only adaptive evolution of carbaryl resistance but also associated evolutionary costs. Importantly, both the evolutionary b