Redistribution of Cetaceans Based on Ocean Temperature Using Spatiotemporal Regression Models

ABSTRACT Aim A warming climate is predicted to drive shifts in marine species' distribution. Determining the environmental variables influencing the distribution and habitat use of large predators, such as cetaceans, is critical for conservation management, yet remains poorly understood. New Zealand is a global cetacean hotspot, with ecosystem productivity supporting multi‐species cohorts throughout coastal waters. Cetaceans are early indicators of change as shifts in productivity influence prey availability, making them potential sentinel species. Location Hauraki Gulf, Aotearoa New Zealand. Methods We used cetacean sighting data collected by a whale‐watch operator from 2000 to 2019, alongside monthly averaged sea surface temperature (SST) measurements. We modelled the locations and times of sightings to understand spatiotemporal patterns in the occurrence of frequently sighted cetaceans within the Gulf. Using latent Gaussian Markov random fields to model spatially varying effects of SST on cetacean occurrence and to account for spatiotemporal correlations, we investigated distributional shifts over time and how these were related to SST. Results Hotspots and spatial clustering varied depending on prey trophic level. For Bryde's whales and common dolphins, hotspots in density shifted gradually with a spatially varying effect of SST depending on location. These effects were not observed for bottlenose dolphins or killer whales, for which hotspots in density often shifted rapidly and could not be attributed to SST. Probability of occurrence of Bryde's whale and common dolphin peaked along the northern coastline in cooler months, with movement further out under warmer months. Similar seasonal shifts were not evident for the other species. Main Conclusions Finding relatively simple and low‐cost approaches that inform adaptive management decisions is important, particularly for endangered species. We highlight the value of a long‐term dataset, a simple environmental measure (SST), and robust model design to understand the dynamic distribution patterns of cetaceans and associated species in a rapidly changing ocean.