Evolutionary-ecotoxicological study of the sensitivity to pesticides in aquatic systems under global warming: effects of temperature variation and temperature extremes

In natural ecosystems animals encounter several environmental stressors that furthermore can interact with pollutants and potentially increase their toxicity. This is considered one of the underlying reasons why current ecological risk assessment is possibly failing to protect natural ecosystems as it may be lacking realism by being based on pesticide toxicity testing under standard laboratory conditions. While it is widely known that an increase in mean temperature increases the toxicity of pollutants, studies largely ignored the effects of daily temperature fluctuations (DTFs) on pesticide toxicity and never tested pollutant toxicity under both global warming stressors (the more realistic scenario in which both the mean temperature and the DTF increase). Furthermore, how DTFs shape pesticide toxicity is mostly unknown, which asks for an integrated approach including life history and physiology. In this thesis, I tested for the single and combined effects of exposure to the pesticide chlorpyrifos (CPF) and warming (+ 4 °C increase in mean temperature and a + 5 °C increase in DTF) in Ischnura elegans damselfly larvae by executing a series of common-garden experiments. I studied effects on life history traits, heat tolerance and candidate underlying physiological mechanisms. Thereby, I used damselfly larvae across a latitudinal gradient (European low- and high-latitude populations) to study the role of thermal adaptation in shaping the sensitivity to pesticides under global warming. The mean summer water temperatures and maximum summer DTFs in shallow freshwater ponds located in southern France (low latitude) are 24 °C with 10 °C DTF, and in southern Sweden (high latitude) are 20 °C with 5 °C DTF. This 4 °C difference in mean temperature and 5 °C difference in maximum DTF also matches the predicted increase in both factors by 2100 under the IPCC RCP 8.5 scenario, which allows for applying a space-for-time substitution to test if gradual thermal evolution in high-latitude populations may buffer for the increased pesticide toxicity under global warming (increase in both mean temperature and DTF). I reviewed the strengths and weaknesses of this approach in chapter I.

            In the second chapter, I studied the effects of global warming in the absence of the pesticide and I found that DTFs only had negative effects on growth rate under the 4 °C warming treatment. While 4 °C warming was beneficial for larvae of both latitudes, this changed in a negative effect in the presence of high DTF. These negative effects of DTF were stronger in high-latitude larvae and already occurred at low DTF, indicating local thermal adaptation. This also suggests that if high-latitude populations are able to gradually thermally evolve into ‘low-latitude’ populations, they would no longer suffer a growth reduction in the presence of DTF under 4 °C warming.

            When studying the effects of DTFs on pesticide toxicity in chapter III, I found the striking result that while the used chlorpyrifos concentration was not affecting the damselfly larvae’s life history at a constant temperature of 20 °C, it did strongly decrease survival and growth in the presence of DTFs around 20 °C. These results suggest that in standard pesticide toxicity tests, which are carried out by current risk assessment, this concentration would have been regarded as safe. Thereby, this highlights it is crucial to integrate DTFs in current risk assessment to reach more realistic predictions about pesticide toxicity in natural systems.

            In the last three chapters, I studied the effects of pesticides under global warming, thereby including the predicted increases in both mean temperature and DTF. The toxicity of chlorpyrifos was magnified by the increase in both mean temperature and DTF, but especially at their combination. Furthermore, I described in chapter V a novel, likely general mechanism that contributes to the higher chlorpyrifos toxicity under global warming by coupling two general principles: the widespread temperature-size rule and the size-pesticide sensitivity pattern. Larvae got smaller under DTFs, and these smaller larvae were more vulnerable to the pesticide, hence the higher chlorpyrifos toxicity under DTFs was partly mediated through DTF-induced reductions in body size. In terms of physiological variables, I found evidence of chlorpyrifos-induced effects being stronger under DTFs in terms of oxidative damage to lipids, which may contribute to the mortality patterns. Further, in my last chapter, I observed that the chlorpyrifos-induced reductions in bioenergetic response variables (energy availability and net energy budget) were stronger when the high mean temperature was combined with the high DTF. Moreover, I also showed that the bioenergetic responses contributed to the higher chlorpyrifos toxicity under global warming as treatment combinations with lower net energy budgets showed higher mortality and lower growth rates. Although I did not always find evidence that possible gradual thermal evolution in high-latitude damselflies would buffer for the increased chlorpyrifos toxicity under global warming, I did find a strong signal in chapter IV. Latitude-specific thermal adaptation to both mean temperature and DTF buffered for the chlorpyrifos-induced reduction in heat tolerance, meaning that possible gradual thermal evolution in high-latitude populations may buffer for the negative effects of chlorpyrifos on heat tolerance under warming, unless the DTF increase is taken into account.

My results indicate that it is crucial to not only consider DTFs, but also their interaction with increasing mean temperatures, and to integrate gradual thermal evolution to make more realistic predictions of pesticide toxicity in the current climate and in a warming world.