Sean T. Giery

Ameliorating the impacts of climate change on communities requires understanding the mechanisms of change and applying them to predict future responses. One way to prioritize efforts is to identify biotic multipliers, which are species that are sensitive to climate change and disproportionately alter communities. We first evaluate the mechanisms underlying the occupancy dynamics of marbled salamanders, a key predator in temporary ponds in the eastern United States We use long‐term data to evaluate four mechanistic hypotheses proposed to explain occupancy patterns, including autumn flooding, overwintering predation, freezing, and winterkill from oxygen depletion. Results suggest that winterkill and fall flooding best explain marbled salamander occupancy patterns. A field introduction experiment supports the importance of winterkill via hypoxia rather than freezing in determining overwinter survival and rejects dispersal limitation as a mechanism preventing establishment. We build climate‐based correlative models that describe salamander occupancy across ponds and years at two latitudinally divergent sites, a southern and middle site, with and without field‐collected habitat characteristics. Correlative models with climate and habitat variation described occupancy patterns better than climate‐only models for each site, but poorly predicted occupancy patterns at the site not used for model development. We next built hybrid mechanistic metapopulation occupancy models that incorporated flooding and winterkill mechanisms. Although hybrid models did not describe observed site‐specific occupancy dynamics better than correlative models, they better predicted the other site's dynamics, revealing a performance trade‐off between model types. Under future climate scenarios, models predict an increased occupancy of marbled salamanders, especially at the middle site, and expansion at a northern site beyond the northern range boundary. Evidence for the climate sensitivity of marbled salamanders combined with their disproportionate ecological impacts suggests that they might act as biotic multipliers of climate change in temporary ponds. More generally, we predict that top aquatic vertebrate predators will expand into temperate‐boreal lakes as climate change reduces winterkill worldwide. Predaceous species with life histories sensitive to winter temperatures provide good candidates for identifying additional biotic multipliers. Building models that include biological mechanisms for key species such as biotic multipliers could better predict broad changes in communities and design effective conservation actions.