Survival of the fattest: linking body condition to prey availability and survivorship of killer whales

Recovering small, endangered populations is challenging, especially if the drivers of declines are not well understood. While infrequent births and deaths may be important to the outlook of endangered populations, small sample sizes confound studies seeking the mechanisms underlying demographic fluctuations. Individual metrics of health, such as nutritive condition, can provide a rich data source on population status and may translate into population trends. We examined interannual changes in body condition metrics of endangered Southern Resident killer whales (SRKW) collected using helicopters and remotely operated drones. We imaged and measured the condition of the majority of all three social pods (J, K, and L) in each of seven years between 2008 and 2019. We used Bayesian multi‐state transition models to identify relationships between body condition changes and both tributary‐specific and area‐based indices of Chinook salmon abundance, and K‐fold cross‐validation to compare the predictive power of candidate salmon covariates. We found that Fraser River (tributary‐specific) and Salish Sea (area‐based) Chinook salmon abundances had the greatest predictive power for J Pod body condition changes, as well as the strongest relationships between any salmon covariates and SRKW condition across pods. Puget Sound (tributary‐specific) Chinook salmon abundance had the greatest predictive power for L Pod body condition changes, but a weaker relationship than Fraser River or Salish Sea abundance had with J Pod body condition. The best‐fit model for K Pod included no Chinook covariates. In addition, we found elevated mortality probabilities in SRKW exhibiting poor body condition (reflecting depleted fat reserves), 2–3 times higher than whales in more robust condition. Collectively, these findings demonstrate that (1) fluctuations in SRKW body condition can in some cases be linked to Chinook salmon abundance; (2) the three SRKW pods appear to have distinct patterns of body condition fluctuations, suggesting different foraging patterns; and (3) aerial photogrammetry is a useful early‐warning system that can identify SRKW at higher risk of mortality in the near future.