Heatwaves and turbidity influxes have interacting and lasting effects on the foraging behaviour of three-spined stickleback Gasterosteus aculeatus

Abstract

Freshwater ecosystems support a disproportionately high share of fish species and biodiversity yet are among the most threatened ecosystems globally and are increasingly being exposed to multiple, co-occurring environmental stressors. In addition, predicting the combined impacts of environmental stressors remains a major challenge, particularly when stressors differ in how quickly they arise and how long their effects persist. Understanding how these stressors interact across time is critical for predicting their ecological consequences. Here, we examine how warming and turbidity, two widespread and intensifying stressors, interact to influence predator-prey dynamics in the three-spined stickleback (\textit{Gasterosteus aculeatus}). Using a fully factorial experiment, fish were exposed to ambient, warmed, turbid, or combined conditions for short (one week) or prolonged (six weeks) periods, before feeding rates were quantified across prey densities using a novel stochastic Bayesian model. We found that foraging behaviour was shaped primarily by prior exposure rather than the immediate experimental conditions. Short-term exposure to both stressors increased feeding rates, whereas prolonged warming reduced them. In contrast, turbidity continued to increase feeding rates over both time periods and, when combined with warming, offset the negative effects of extended heat exposure. Changes in feeding performance at the individual level may propagate through food webs, potentially triggering trophic cascades that reduce biodiversity. Given the exceptional biodiversity of freshwater systems, such dynamics are of particular concern. We also found that behavioural changes persisted even after stressors were removed, indicating that exposure history can have lasting effects. These findings suggest that short-term experiments, such as the functional response, may actually be measuring subject experiential history rather than responses to specific experimental conditions.