Accounting for spatiotemporal sampling bias in a long-term dataset establishes a decline in abundance of endangered false killer whales Pseudorca crassidens in the main Hawaiian Islands Endangered Species Research

We estimated abundance of the endangered main Hawaiian Islands (MHI) insular population of false killer whales Pseudorca crassidens from 1999-2022 using a modeling technique that incorporates animal availability in a capture-recapture analysis. The population was sampled using different sampling methods, resulting in yearly encounter histories of 265 individuals and 53 satellite-tagged whales. Survey effort and animal location data were separately analyzed using kernel density estimators, and the degree of overlap between these 2 processes was used to model detection probability in a Bayesian Jolly-Seber population model. This approach better addresses spatiotemporally variable sampling effort than traditional capture-recapture methods, improving the estimation of reliable abundance trends. Using simulated data, the model was robust to many sampling and ecological complications, such as variable low detectability, unequal tag deployment lengths, and variable social group sizes. Fitting the model to the MHI false killer whale data set, we found that the insular population of false killer whales remains small, with an estimated 139 individuals (95% credible interval, CRI = [114, 162]) in 2022. The population appears to be in decline throughout the study period, with a mean annual percent change of -1.09 (95% CRI = [-2.11, -0.023]) over the entire time series and -3.51 (95% CRI = [-5.08, -1.88]) since 2013, when the population was listed as endangered. Given the magnitude of the decline, identifying which of the many factors affecting this population is most responsible is key in order to guide potential management responses.