A General Framework for Predicting the Temperature-Dependence of Microbial Interactions

A general mechanistic understanding of the impacts of temperature on resource-mediated interactions between microbial species is currently lacking. Here, we develop a trait-based mathematical framework to derive the temperature-dependence of effective pairwise species interactions within a microbial community that accounts for higher-order interactions. Parameterising this framework using bacterial metabolic traits leads to three key predictions. First, species' effective interactions respond unimodally to environmental temperature and are typically more thermally sensitive than their underlying metabolic processes (e.g., resource uptake and respiration). Second, the temperature response of interactions is mainly governed by resource competition. Third, the mean and variance of intraspecific interaction strengths increase with temperature faster than interspecific interactions. These findings have key implications for our understanding of microbial community dynamics in natural environments, paving the way for predicting the impacts of temperature fluctuations and change on their coexistence, diversity, and stability.