Multistressor ecotoxicological study on Culex mosquitoes: from toxicity mechanisms to biotic interactions

Natural populations are increasingly facing multiple stressors. Their ability to deal with interacting stressor combinations will be crucial for their local persistence. Populations of target species are typically exposed to pesticide mixtures and natural stressors, and are increasingly developing resistance to single pesticides. Nevertheless, we have poor knowledge whether natural stressors and the presence of pesticide resistance shape mixture toxicity. To assess the pesticide control efficiency and to reduce ecological damage to non-target species, it is important to quantify the effect of pesticide mixtures and compare them with the effect of their single pesticides on pest species, non-target species and their predator-prey interactions. Natural stressors that are widespread in freshwater systems are predation risk and warming; both may magnify the toxicity of single pesticides. The latter pattern has been captured in the 'Climate-induced toxicant sensitivity' concept (CITS). Nevertheless, deviations from this general pattern have been reported. To advance insights, there is a pressing need to identify the underlying molecular stress mechanisms of the CITS patterns. Moreover, while many studies on the toxicity of pesticides looked at the effects of a higher mean temperature, effects of the realistic scenario of daily temperature variation (DTV, which is a key component of global warming) are understudied. In this context, developmental stages and consequences for biotic interactions such as competition have also been largely ignored. In Chapters 1-2 (Part I), I tested the single and combined effects of the pesticide chlorpyrifos and the biopesticide Bacillus thuringiensis israelensis (Bti) on the survival of a chlorpyrifos-resistant and non-resistant strain of the Southern house mosquito Culex quinquefasciatus. In Chapter 1, I investigated whether these effects of chlorpyrifos and/or Bti were magnified by synthetic predator cues of Notonecta water bugs. Single exposure to Bti caused mortality in both strains and single exposure to chlorpyrifos caused only mortality in the non-resistant strain, while predator cues did not induce mortality. The chlorpyrifos-resistant strain was more sensitive to Bti, indicating a cost of resistance. The interaction types between chlorpyrifos and Bti (additive), between chlorpyrifos and predator cues (additive), and between Bti and predator cues (additive) were consistent in both strains. In Chapter 2, I tested the effects of chlorpyrifos and/or Bti on the mortality by predation by the pygmy backswimmer Plea minutissima. After exposure to the mixture, equal mortality by predation was caused in both mosquito strains. As expected, less mortality by predation in the chlorpyrifos-resistant mosquito strain compared to the non-resistant strain occurred after the chlorpyrifos exposure. Notably, the predator killed more mosquito larvae of the resistant strain compared to the non-resistant strain after Bti exposure indicating a predation cost of resistance in the presence of a biopesticide. Part I highlights the integrated use of Bti and natural control by predators as a promising strategy to counter the build-up of resistance and to keep vector mosquito populations under control. In chapters 3-5 (Part II), I tested whether daily temperature variation (DTV) magnifies the toxicity of chlorpyrifos in the Northern house mosquito Culex pipiens. In Chapter 3, I also studied whether DTV magnifies the single toxicity of Bti and/or its combined toxicity with chlorpyrifos. DTV was not lethal and did not change the toxicity of the individual pesticides. Yet, a key novel finding was that high DTV increased the mortality of the mixture by changing the interaction between both pesticides from additive to synergistic. In Chapter 4, I tested whether the effect of DTV and its interaction with chlorpyrifos was dependent on the developmental stage. DTV had again no direct negative effects and did not change the toxicity of the individual pesticides. Exposure to chlorpyrifos at a constant temperature-imposed mortality and reduced the heat tolerance in both larvae and adult males, but not in adult females. This pesticide-induced decrease of heat tolerance is in line with the TICS ("toxicant-induced climate change sensitivity") concept whereby the first evidence was provided that the TICS can be sex-specific. Notably, DTV interacted synergistically with chlorpyrifos for heat tolerance of the larvae and adult females. DTV increased the chlorpyrifos-induced decrease (CITS) of heat tolerance (TICS), providing support for the reciprocal effects between DTV and contaminants, hence the coupling of the TICS and CITS concepts (CITS --> TICS). This coupling of CITS and TICS was developmental stage specific as the interaction effect between DTV and chlorpyrifos was additive for adult males. In Chapter 5, I also studied the effect of competition with the water flea Daphnia magna on the combined effects of DTV and chlorpyrifos on the mosquito larvae. In this experiment, no pesticide-induced mortality was observed in the presence of DTV due to an accelerated pesticide breakdown under DTV, which contrasts with the general pattern of the CITS concept. There was neither a significant effect of competition on mortality, nor did competition influence the interaction between the single pesticide and DTV. Together these chapters underscore the importance to consider the effects of DTV on pesticide toxicity and on pesticide exposure in risk assessment of toxicants. In Chapters 6-7 (Part III), I tested three concentrations of the pesticide chlorpyrifos (absence, low-effect and high-effect) in the absence and presence of 4 °C warming on larvae of the Northern house mosquito Culex pipiens. In Chapter 6, I showed that both the low-effect and high-effect chlorpyrifos concentrations were lethal and generated mostly negative sublethal effects: activity of acetylcholinesterase (AChE) and total fat content decreased, and oxidative damage to lipids increased, yet growth rate increased. Warming was slightly lethal, yet had positive sublethal effects: growth rate, total fat content and metabolic rate increased, and oxidative damage decreased. The independent action model identified the expected synergistic interaction between chlorpyrifos and warming in four out of seven response variables. Notably, three variables (survival, AChE, and fat content) were strongly dependent on the chlorpyrifos concentration, and two of these (AChE and fat content) were not associated with a significant interaction in the general(ised) linear models. In Chapter 7, I tested the effects of chlorpyrifos under warming at the gene expression level. By applying the independent action model on RNA-seq data I confirmed my hypothesis that synergistic interactions between both stressors at the phenotypic level are underpinned by a higher frequency or strength of antagonistic upregulations (less upregulated than expected based on responses to the single stressors) and of synergistic downregulations (more downregulated than expected) of general stress defence response genes (protection of macromolecules, antioxidant processes, detoxification, and energy metabolism/allocation). These results are relevant to improve vector mosquito control strategies, resistance management and ecological risk assessment of pesticides. I identified a novel cost of resistance to a chemical pesticide in terms of increased vulnerability to a biopesticide, and to predation after exposure to the biopesticide. In addition, the finding of a higher toxicity of the mixture at high DTV compared to the typically used constant test temperatures in the laboratory urges caution when evaluating the environmental impact of pesticide mixtures. Not only the effects of DTV on pesticide toxicity, but also on pesticide degradation (hence exposure) should be considered in ecotox testing. I identified several factors that may affect the detection of the CITS concept: dependence on concentration, developmental stage, genetic strain and sex, and appropriate null model testing. The identified factors are important to take into account when studying the toxicity of contaminants in a warming world. Finally, I provide guidelines to formally test the stressor interaction type for gene expression data. Following this approach, my results highlight that a quantitative analysis of the frequency and strength of the interaction types of general stress response genes, specifically focusing on antagonistic upregulations and synergistic downregulations, may advance the mechanistic understanding of how other stressors modify the toxicity of contaminants.

Publication year:2020

Accessibility:Open