Abstract
Life-history performance of individuals in wildlife populations emerges from the interplay between the multiple processes that constitute an animal's health. Monitoring and modelling indicators of health can thus provide a way to assess and forecast the status of a population before its abundance changes. In this study, we develop a Bayesian state-space model that links multiple health indicators (representing energy, endocrine and morphometric status) and resulting female calving probability, using an 8-year dataset of repeated sightings, morphological measurements, faecal sampling and offspring observations of gray whales (Eschrichtius robustus) belonging to the Pacific Coast Feeding Group. Model results indicate that calving probability emerges from the combined effect of a female's structural body size and available energy reserves, while also showing a weak negative correlation with glucocorticoid levels prior to pregnancy. Assessment of population age structure suggests that the number of individuals in younger age classes is smaller than expected for a growing or a stable population, which, together with decreasing body size, could indicate an impending decline in this group. Model development was made possible by the collection of high-resolution, longitudinal data on individuals, although several mechanistic assumptions were imposed by the relatively short time series (8 years), influencing the results. Our modelling approach could inform similar efforts in other long-lived species where population dynamics cannot be easily monitored. Ultimately, models of wildlife health and vital rates can support assessments of the population-level consequences of multiple stressors, a key goal for management and conservation across systems and jurisdictions.
