Stochastic within-host dynamics and climate-sensitive traits generate predictable patterns of variation in disease outcomes

Understanding how climatic variables impact host-pathogen relationships in temperature-sensitive ectothermic host organisms is crucial under global change. Few studies have explored how temperature gradients generate inter-individual variation in epidemiological traits like host susceptibility or pathogen replication. Here, we develop a mathematical model to explore a novel hypothesis: stochastic within-host dynamics and simulated thermal mismatches between host and pathogen traits generate predictable variation in infection outcomes among hosts and across thermal gradients. Our model demonstrates that varying trait thermal optima in host immunity and pathogen replication, and stochastic within-host processes produced variation in infection outcomes. Variability was low when temperatures strongly favored host or pathogen traits, but high and diverse when their performance was similar across a broad thermal range. In contrast, when trait performance was equal across all temperatures (no mismatch) variability remained low at all temperatures. Further, the magnitude of variation, quantified by entropy, exhibited predictable patterns depending on host-pathogen thermal mismatches. We conclude that interactions between trait thermal mismatches and within-host stochasticity provide a theoretical framework to improve ectotherm disease models under climate change, providing a valuable tool for exploring the impacts of environmental change on epizootic or epidemic dynamics, particularly in vulnerable marine ecosystems.