Travis J. Klee

Immune responses are crucial to not only control within‐host disease dynamics and potential negative outcomes to host health but also impose energetic trade‐offs with growth and other physiological processes. Although these trade‐offs play an important role in disease outcomes, we lack mathematical models that mechanistically describe how variations in energy investment and parental reproductive timing influence offspring survival under seasonal pathogen dynamics. Here, we determine the strategies (set of tactics—growth versus immune defence—over time) that maximize host fitness under persistent infections by implementing dynamic optimization state models. We modelled individual trajectories for 1 year and accounted for natural fluctuations in resource availability. We parameterized the model using field and experimental data from a well‐studied host–pathogen system: the coqui frog ( Eleutherodactylus coqui ) and the amphibian chytrid fungus ( Batrachochytrium dendrobatidis ). This data from field monitoring and controlled experiments enable robust parameter estimation and subsequent validation using empirical data. Our models identified critical windows that maximize individual growth while limiting mortality under increased pathogen burden. Individuals generally chose growth but shifted to defence when approaching near‐lethal infection levels. Seasonality in pathogen exposure and foraging success exacerbated growth–defence trade‐offs, leading to delayed maturity and lower survival rates when frogs hatched under unfavourable conditions. Our simulations provide a mechanistic view that explains empirical results showing that periods of highest reproductive activity align with high‐resource availability and low pathogen risk. Furthermore, our findings demonstrate that shifts in energy‐allocation modulate infections and constrain fitness traits. Our results highlight the utility of dynamic state‐variable models to examine how host fitness strategies impact early‐life growth rates and recruitment (number of offspring reaching maturity). We provide detailed instructions that allow extending the application of this model to other systems. By adjusting parameter values using empirical data, this model approach can be applied to predict fitness under current and future conservation challenges, such as extreme droughts and pathogen outbreaks, and to identify the optimal time for releasing individuals in species reintroduction programmes.