Drivers for Spatial Modelling of a Critically Endangered Seabird on a Dynamic Ocean Area: Balearic Shearwaters Are Non‐vegetarian

Spatial modelling is an important research tool to improve our knowledge about the distribution of wildlife in the ocean. Using different modelling techniques (MaxEnt and a generalized linear mixed model), a predictive habitat suitability model was developed for one of the most threatened seabirds in the world: the Balearic shearwater, Puffinus mauretanicus. Models were developed using a 10-year dataset from the Gulf of Cadiz (on the south-western Iberian Peninsula), a key foraging area for Balearic shearwaters during migration and the non-breeding season. Predictive habitat maps strongly matched the observed distribution patterns, pointing to bathymetric features as the main modelling drivers. The species was concentrated on shallow areas (up to approximately 100 m in depth) of the continental shelf, very close to the mouth of the Guadalquivir River. In contrast with previous studies, Balearic shearwater distribution in the highly dynamic Gulf of Cadiz was not correlated with areas of high chlorophyll a concentration. This lack of spatial correlation probably arises from the delay between the phytoplankton bloom and the response of the zooplankton and small fish that are preyed upon by Balearic shearwaters, which may result in important displacements of this trophic chain across the Gulf of Cadiz. The analysis presented contributes to a better understanding of the spatial distribution and ecology of the critically endangered top predator in the Gulf of Cadiz and offers important information to improve management plans.