Use of a Bayesian Hierarchical Density Surface Model to Estimate Seasonal Distributions of Large Whales Off the East Coast of the United States With an Application to a Wind Energy Area

ABSTRACT Aim Offshore wind energy development off the East Coast of the United States is in the nascent but growing stage with consequences for wildlife. Marine mammals are particularly sensitive to pile driving activity, and therefore, it is critical to provide guidance about where and when impacts are most likely to be minimized. The aim of this study was to develop Bayesian hierarchical density surface models (BHDSMs) for five species of large whales and use model outputs specifically to characterize seasonal abundance within a wind energy area (WEA) off the East Coast of the United States. Location Western Atlantic Ocean, USA. Methods We use eight years of line transect survey data to develop BHDSMs for fin whales ( Balaenoptera physalus ), humpback whales ( Megaptera novaeangliae ), minke whales ( Balaenoptera acutorostrata ), sei whales ( Balaenoptera borealis ) and sperm whales ( Physeter macrocephalus ). We produced seasonal maps of densities at a basin‐wide scale with associated measures of uncertainty. We then apply our models to a WEA off the Massachusetts‐Rhode Island coast and estimate seasonal shifts in abundance and calculate the posterior probability of abundance exceeding user‐defined thresholds for each species. Results Seasonal distributions were largely consistent with previous studies of large whales for this region. Application to the WEA demonstrated an interaction between seasonal peaks in abundance and species. Posterior probabilities of exceeding the user‐defined threshold tracked seasonal changes in abundance, with differences amongst species in the magnitude of the probabilities. Main Conclusions This application of a BHDSM demonstrates the multipurpose use of habitat models to examine both large‐scale patterns in density with robust measures of uncertainty and localized variation in abundance in a probabilistic framework that can be used to assess risk. Model outputs can be readily applied to additional areas off the East Coast of the U.S. that are of interest.