Habitat-based Density Models for a Marine Predator - the Harbour Porpoise in the North Sea

Effective species conservation and management requires information on species distribution patterns, which is particularly challenging for highly mobile and cryptic species that may be subject to multiple anthropogenic stressors across international boundaries. Understanding species-habitat relationships can improve the assessment of population size, trends and distribution by explicitly allowing high resolution data on habitats to inform abundance estimation and the identification of spatial management units. In this study, we aggregated an unprecedented set of survey data of a marine top predator, the harbour porpoise Phocoena phocoena, collected in the UK (Dogger Bank, SCANS II), Belgium, the Netherlands, Germany and Denmark, to develop seasonal habitat-based density models for the North Sea. Visual survey data were collected year-round over nine years (2005-2013) by means of dedicated line-transect surveys, taking into account the proportion of missed sightings on the transect. Generalized additive models of porpoise density were fitted to 156,630 km of on-effort survey data with 14,356 sightings of porpoise groups. Selected predictors included static and dynamic variables, such as depth, distance to shore and to sandeel (Ammodytes spp.) grounds, satellite-derived sea surface temperature (SST), spatial and temporal variation in SST (as proxies for fronts) and day length. Day length and a 3-dimensional tensor smooth representing the spatial distribution of daily SST proved to be good proxies for “season”, allowing predictions in both space and time. The density models captured previously described seasonal shifts in distribution of harbour porpoise across international boundaries. By combining the large scale international SCANS II survey with the more frequent, small scale national surveys it has been possible to provide seasonal maps that can assist the EU Habitats and Marine Directives in more effectively assessing the conservation status of the harbour porpoise. Moreover, our results can facilitate the identification of management units where human activities and disturbances are likely to impact the population.