Box 5: How behavior can shape stressor effects on communities
Behavioural responses to stressors can have important indirect implications for community structure and function (Figure 5). Adaptive responses to stressors often involve shifts in foraging or antipredator behaviour that change the functional role of organisms within the food-web. For instance, in response to stress, herbivores may shift their diet preference to select plants with higher rich, digestible carbohydrates to achieve nutritional and homeostatic balance. This selective herbivory not only changes the plant community composition, but has consequences for nutrient cycling and energy flow, because it alters the elemental composition and non-processed plant litter reaching the detrital pool (Hawlena & Schmitz 2010). Importantly, however, such behavioural responses to environmental stressors, and the corresponding ramifications for communities, are largely mediated by how these stressors alter species interactions, such as predator-prey dynamics, competition and mutualisms (Miller et al. 2014; Thompson et al. 2018). Understanding how stressors impact behaviorally mediated species interactions is essential for predicting when synergistic interactions will arise, but this fundamental aspect of ecology is often absent from multiple stressor studies (Tylianakis et al. 2008; Thompson et al. 2018).
If stressors hit predators harder than prey, stressors can benefit prey and moderate both the individual and combined effects of stressors by allowing prey to feed with reduced risk, and divert energy into other life-history and physiological processes (Francis et al. 2009; White et al. 2018). If stressors specifically inhibit keystone predators, however, this can lead to the restructuring or collapse of entire ecosystems (Breitburg et al. 1998; Rumschlag et al.2019). For example, the combination of organophosphate insecticides and triazine herbicides dramatically increased trematode pathogens in pond mesocosms, because these conditions favoured populations of the intermediate host (snails) by reducing the top-down behaviorally- and density-mediated effects of their predators and increasing periphyton food abundance (Rumschlag et al. 2019). In contrast, if stressors disproportionately change the behavior and feeding ecology of prey, for instance, by increasing their foraging activity or reducing antipredator responses, predation rates are likely to increase, leading to stronger top-down effects on intermediate consumers (Shears & Ross 2010; Milleret al. 2014).
Single and multiple stressors can also directly and indirectly disrupt mutualisms (Hegland et al. 2009; Schweiger et al. 2010). Changes in the timing and spatial synchronicity of species behavioural patterns in response to stress may lower species co-occurrence rates leading to the deterioration of mutualistic interactions (Hegland et al. 2009, Schweiger et al. 2010, Figure III). Furthermore, negative impacts of stressors on a single member of a mutualism can have large costs for other members, including indirect beneficiaries of the mutualism (Hegland et al. 2009; Schweiger et al. 2010; Barton & Ives 2014). For example, the negative impact of agricultural pesticides on honeybee foraging behaviour and colony health can reduce plant numbers (via reduced pollination) and, in turn, have negative effects on herbivores (Goulson et al. 2015; Tosi et al. 2017). Alternatively, mutualisms can help buffer ecosystems against negative effects of environmental stress (Thompson et al. 2018). For instance, an ant-aphid mutualism can protect plants from indirect effects of increasing temperatures by limiting positive-effects of rising temperatures on the abundance of pest predator species (Marquiset al. 2014).
Multiple stressor effects on communities are difficult to predict but are likely contingent upon the degree of tolerance and co-tolerance of species and functional groups to those stressors (Vinebrooke et al. 2004), as well as the trophic level(s) upon which these stressors have the greatest impact. In situations when key functional groups or species are particularly sensitive to stressors and are subsequently reduced or eliminated from the community pool, pronounced behavioural and numerical effects can propagate across trophic levels if the ecosystem functionality of those groups is not replaced (Galic et al. 2018; Dib et al. 2020). On the other hand, when remaining species can compensate functionally for this loss, stressor effects tend to be weakened across trophic levels, leading to more resistant communities (Jackson et al. 2016).