Exploring habitat selection in sea snakes using passive acoustic monitoring and Bayesian hierarchical models
Udyawer, V., Simpfendorfer, C.A., Read, M., Hamann, M., and Heupel, M.R. (2016) Exploring habitat selection in sea snakes using passive acoustic monitoring and Bayesian hierarchical models. Marine Ecology Progress Series, 546. pp. 249-262.
PDF (Published Version)
- Published Version
Restricted to Repository staff only |
Abstract
Resource selection studies often use analytical techniques that provide information at either a population or an individual level. We applied a Bayesian hierarchical model that simultaneously estimates population- and individual-level habitat selection to explore how varying levels of dietary specialisation affect resource requirements of 2 species of sea snakes that occupy the same coastal environment. We used passive acoustic telemetry to monitor the movements of the 2 species—a dietary generalist, Hydrophis (Lapemis) curtus, and a dietary specialist, H. elegans—and investigated how individuals select habitats based on habitat type, depth and proximity to sources of freshwater within a nearshore environment. Composition of diets in both species was also assessed using regurgitated material from captured individuals. Selection of habitats by the 2 species differed, with H. elegans displaying an affinity for mudflat and seagrass habitats <4 km from sources of freshwater and depths <3 m. H. curtus selected for slightly deeper seagrass habitats (1-4 m) further from freshwater sources (2-5 km). Data from regurgitated material showed that the diet of H. curtus comprised at least 4 families of fish and displayed some level of intraspecific predation, whereas H. elegans preyed solely on eels. Both species predominantly selected seagrass areas, indicating that these habitats provide key resources for sea snakes within nearshore environments. The results illustrated the utility of Bayesian hierarchical models when analysing passive acoustic monitoring data to provide population-level habitat selection metrics and incorporate individual-level variability in selection, both of which are necessary to inform targeted management and conservation practices.