Integrating multiple data sources for assessing blue whale abundance and distribution in Chilean Northern Patagonia

Abstract Aim Species distribution models are useful tools for depicting important habitat, assessing abundance and orienting conservation efforts. For small populations in poorly studied ecosystems, available data are often scarce and patchy. To overcome this limitation, we aim to evaluate the use of different data types within a hierarchical Bayesian framework with the goal of modelling the abundance and distribution of a small and highly migratory population of blue whale ( BW , Balaenoptera musculus ) summering in Chilean Northern Patagonian ( CNP ). Location CNP, Eastern South Pacific (ESP). Methods We constructed a Bayesian hierarchical species distribution Model ( HSDM ), combining a binomial N‐mixture model used to model BW groups counts in line‐transect data (2009, 2012 and 2014) with a logistic regression for modelling presence‐availability data (2009–2016), allowing both models to share covariate parameters for borrowing strength in estimations. Results Distance to areas of high chlorophyll‐ a concentration during spring before summering season ( AHCC ‐s) was the most important and consistent explanatory variable for assessing BW abundance and distribution in CNP . Incorporating accessorial presence‐only data reduced uncertainty in parameters estimation when comparing with a model using only line‐transect data, although other covariates of secondary importance failed to be retained in this model. Main conclusions Our results remark the capability of HSDM for integrating different data types providing a potential powerful tool when data are limited and heterogeneous. Results indicate that AHCC ‐s, and possibly thermal fronts, could modulate BW abundance and distribution patterns in CNP . Preliminary model‐based delimitations of possible priority conservation areas for BW in CNP overlap with highly used vessel navigation routes and areas destined to aquaculture.