Seasonal variation in environmental and behavioural drivers of annual‐cycle habitat selection in a nearshore seabird

Aim Conservation of highly mobile species often requires identifying locations or time periods of elevated vulnerability. Since both extrinsic habitat conditions and intrinsic behavioural and energetic requirements contribute to habitat use at the landscape scale, identifying spatial or temporal foci for conservation intervention requires understanding how habitat needs and distributions vary across the annual cycle. Nearshore marine birds inhabit highly dynamic systems and have widely varying habitat needs among breeding, moult and non‐breeding seasons, making them a useful case study for testing the relative contributions of individual resource requirements and environmental conditions in driving annual variation in distribution patterns. Location Northern Gulf of Mexico (USA). Methods We tracked Brown Pelicans using bird‐borne GPS transmitters and used a combination of Hidden Markov Models and multivariate selectivity analysis to compare the characteristics of preferred resident habitats used throughout the annual cycle. Results Habitat selection was driven by dynamic oceanographic variables during all stages of the annual cycle. Key habitat characteristics varied between seasons, with particularly strong selection on high productivity, low temperature and low salinity during the breeding and post‐breeding moult periods. The post‐breeding moult also corresponded to a time of limited availability of preferred habitats, resulting in extensive overlap between breeding populations from different administrative planning areas. Main conclusions By incorporating seasonal variation in individual behaviour and resource requirements into our habitat models, we were able to identify the post‐breeding moult as a period of high selectivity and restricted availability of preferred habitats for Brown Pelicans. Locations meeting preferred habitat criteria during the post‐breeding period, particularly estuarine habitats with high productivity and low salinity, would therefore be high‐value targets for management and restoration. Our analysis demonstrates the importance of accounting for both intrinsic and extrinsic temporal variation in evaluating habitat selection.