Predicting the Combined Effects of Multiple Stressors and Stress Adaptation in Gammarus pulex

Abstract

Global change confronts organisms with multiple stressors causing nonadditive effects. Persistent stress, however, leads to adaptation and related trade-offs. The question arises: How can the resulting effects of these contradictory processes be predicted? Here we show that Gammarus pulex from agricultural streams were more tolerant to clothianidin (mean EC₅₀ 148 μg/L) than populations from reference streams (mean EC₅₀ 67 μg/L). We assume that this increased tolerance results from a combination of physiological acclimation, epigenetic effects, and genetic evolution, termed as adaptation. Further, joint exposure to pesticide mixture and temperature stress led to synergistic interactions of all three stressors. However, these combined effects were significantly stronger in adapted populations as shown by the model deviation ratio (MDR) of 4, compared to reference populations (MDR = 2.7). The pesticide adaptation reduced the General-Stress capacity of adapted individuals, and the related trade-off process increased vulnerability to combined stress. Overall, synergistic interactions were stronger with increasing total stress and could be well predicted by the stress addition model (SAM). In contrast, traditional models such as concentration addition (CA) and effect addition (EA) substantially underestimated the combined effects. We conclude that several, even very disparate stress factors, including population adaptations to stress, can act synergistically. The strong synergistic potential underscores the critical importance of correctly predicting multiple stresses for risk assessment.

Keywords: combined effects; fitness costs; genetic adaptation; mixture toxicity; synergism.

Figure 1

Overview of the experimental design. Gammarus pulex was collected from agricultural and reference streams. Both populations were acclimatized to different temperatures (green open rectangle, 16; yellow open rectangle, 19; red open rectangle, 22 °C) for 10 days and exposed to a range of clothianidin (0 to 1000 μg/L) for 48 h under nine different conditions: three prochloraz treatments (gray filled rectangle, 0; blue filled rectangle, 1; magenta filled rectangle, 10 μg/L) × three temperatures (green open rectangle, 16; yellow open rectangle, 19; red open rectangle, 22 °C). Subsequently, the interaction between both pesticides was predicted at different temperatures. For multiple stress, the EC₅₀ values of clothianidin under elevated temperate were compared with the control at 16 °C.

Figure 2

Effects of pesticide contamination on macroinvertebrate community structure quantified with the biological indicator SPEAR: Local pesticide contamination changes the macroinvertebrate community structure (linear regression, adjusted R² = 0.79, F = 41.86, residual df = 10, p < 0.001). Shaded areas represent 95% confidence intervals.

Figure 3

Pesticide tolerance of populations from eight contaminated and four reference streams quantified with their clothianidin tolerance (EC₅₀) under different warming conditions: EC₅₀ of G. pulex collected from control (green) and agricultural streams (red) after exposure (48 h) to clothianidin under different temperature regimes (16, 19, and 22 °C). The lower and upper boundaries of the box represent the 25th and 75th percentile, the horizontal line denotes the median, and the whiskers correspond to the lowest and highest values. Dashed lines represent fitted regressions with confidence intervals displayed by shaded areas. The significance level is displayed as * for p < 0.05, ** for p < 0.01, and *** for p < 0.001.

Figure 4

Relationship between total General-Stress and strength of synergistic effects: combined stress of multiple stressors including prochloraz, suboptimal temperature, and fitness cost of pesticide adaptation significantly increased the clothianidin sensitivity expressed by synergism with effect addition as null model (linear regression, adjusted R² = 0.80, F = 93.28, residual df = 22, p < 0.001). Shaded areas represent 95% confidence intervals.