Environmental drivers of odontocete occurrence in a nearshore temperate habitat

Abstract

Nearshore coastal waters off Northumberland, UK, are important habitats for marine megafauna including marine mammals and seabirds. The area also features extensive anthropogenic activities including shipping, marine renewables development, fisheries, and tourism. Meanwhile there is lack of baseline data on odontocete occurrence to allow assessment of the potential impact from human disturbance. A recent increase in the number of sightings of common bottlenose dolphins (Tursiops truncatus) in the area has provided impetus for new data to inform conservation managers and policy makers. To provide this information, acoustic data were collected using broadband passive acoustic recorders at three sites (Druridge Bay, Newbiggin-by-the-Sea, and St Mary's Island) off the Northumberland coast and analyzed to identify species-specific click trains of common bottlenose dolphin, white-beaked dolphin (Lagenorhynchus albirostris) and harbor porpoise (Phocoena phocoena). Detection data were modelled for each species as detection-positive hours (DPH) using Generalized Additive Models with Generalized Estimating Equations (GAM-GEE) to investigate the effects of environmental covariates, including day of year, diel phase, lunar phase, salinity, year, month and sea surface temperature, and their interaction with location. A total of 20,845 h were recorded during the study resulting in a total of 1660, 71, and 1111 DPH identified for bottlenose dolphin, white-beaked dolphin and porpoise, respectively. The final model for bottlenose dolphins retained all covariates except tidal index and an interaction between location and day of year after model selection and all covariates in the final model were significant. The final model for white-beaked dolphin retained month, year, location, diel index, temperature, lunar index and interactions between location and diel index and all covariates except lunar index were significant. The final model for porpoise retained all covariates and interactions between location and diel phase, day of year, lunar index, and salinity, however, salinity and the interaction between location and lunar index were not significant. Bottlenose dolphin clicks were detected in the area year-round but with two peak periods in occurrence coinciding with May and late September; white-beaked dolphin clicks were detected predominantly in July and August; and porpoise clicks were present year-round but with much greater probability of detection at the Druridge Bay site, where there were peaks in winter and in September. The results of this study provide important information for conservation and management actions that may be needed to reduce anthropogenic pressures on odontocetes in UK waters.