Community regulation models as a framework for direct and indirect effects of climate change on species distributions

Existing projections of climate change impacts focus primarily on direct abiotic impacts on individuals and populations. However, these models often ignore species interactions, which are vital for determining community composition and structure. To evaluate both direct and indirect effects of climate change on species distributions, we applied the Menge–Sutherland model, which describes the relative role of predation and environmental stress in regulating community structure. Using a space‐for‐time approach, we tested the predictions that (1) predators are more strongly impacted by increasing environmental stress than prey (as described in the Menge–Sutherland model) and (2) incorporating indirect (predator) effects increases our ability to predict impacts of increased temperature on prey distributions. We surveyed vertical distributions of predators (sea stars) and a foundational prey species (mussels) at 20 intertidal sites spanning a thermal gradient along the West Coast of the United States. Using generalized linear models and structural equation models, we found that as temperature increased, the upper limits of foundational prey species decreased (a direct effect), while prey lower limits also shifted downward, due to an indirect effect of temperature on predator distributions. Under future climate change, mussel ranges may undergo vertical shifts toward subtidal habitats, allowing for localized persistence of mussels and associated species. Our model comparisons indicate that this framework—incorporating both direct and indirect environmental stress effects within a classic community regulation model—can improve prediction of responses to warming. Community regulation models could be expanded to inform management and conservation efforts during unprecedented climate and ecological change. Climate change is one of the greatest threats to global biodiversity and ecosystem functioning (Sala et al. 2000), and it is becoming increasingly apparent that predicting and mitigating the effects of ongoing and accelerating climate change require going beyond studies examining single species or spatial scales to understand the role of community‐level interactions. Environmental stress affects organisms both directly, through altered physiology, as well as indirectly, through changes to species interactions (Fleeger et al. 2003, Adams 2005). While understanding species’ tolerance limits and modeling these as a function of predicted environmental changes can provide an initial framework, it is also clear that models may be incomplete if biological interactions