Hotter, faster, sicker? Warming shifts the cost of infection from individuals to populations

The thermal environment can profoundly modify the damage that pathogens cause to hosts. Yet while the interacting impacts of infection and temperature on individual life-history traits are well documented, how these effects scale to demographic consequences across thermal gradients remains poorly understood. Using the Daphnia magna-Pasteuria ramosa system across a 10-30°C gradient, we tested whether virulence in individual-level life history metrics reliably reflects the consequences for predicted population growth rates. Our results reveal a stark mismatch between scales: the cost of infection on individual life-history traits (fecundity, body size, and lifespan) was maximised at low temperatures, whereas the impact on population growth rate was maximised at high temperatures and at the thermal optima of healthy hosts. These findings demonstrate that standard virulence metrics at the scale of individual life-history traits may underestimate demographic risk in warming environments. To reliably predict population persistence in the face of disease under a changing climate, it is important to account for the temperature-dependent scaling of pathogen virulence and host life-history to population growth rates.