Abstract
Understanding the ecological processes that underpin the dynamics of community turnover in response to environmental change is critical to predicting how warming will influence ecosystem functioning. Here, we quantify the effect of changing temperature on community composition and ecosystem functioning via the action of ecological selection on population-level thermal traits. To achieve this, we use microbes isolated from a network of geothermal streams in Iceland where in situ temperatures span 8-38°C within a single catchment. We first quantified variability in thermal tolerance between taxa, and then assembled synthetic communities along a broad thermal gradient to explore how temperature-driven selection on thermal tolerance traits shaped the emergent community structures and functions. We found marked changes in community structure and composition with temperature, such that communities exposed to extreme temperatures (10, 35°C) had highly asymmetric biomass distributions and low taxonomic richness. Thermal optima were a good predictor of the presence and relative abundance of taxa in the high-temperature treatments. We also found that the evenness of the abundance distribution was related to ecosystem production, such that communities with more equitable abundance distribution were also the most productive. Our results highlight the utility of using a multi-level approach that links population-level traits with community structure and ecosystem functioning to better understand how ecological communities will respond to global warming.