Modelling habitat preference and estimating the spatial distribution of Australian Sea Lions (Neophoca cinerea); "A first exploration "

Managing the Australian sea lion (Neophoca cinerea) population and mitigating its interactions with commercial fisheries, requires an understanding of their spatial distribution and habitat preference at sea. Numerous wildlife telemetry devices have been attached to individual seals from different colonies, providing a detailed insight into there movement and activities. However, as data are only available from some individuals from 16 of 40 colonies in South Australia, these data represents only a small proportion of the population. Moreover, some colonies are poorly represented. To estimate the spatial distribution of the entire South Australian population, one can first investigate why individuals visit certain places and use this information to predict the spatial distribution for other regions lacking data. In this study we fit Generalized Additive Mixed Models (GAMMs) to wildlife telemetry data collected from adult female Australian sea lions to investigate the species’ habitat preference for the variables distance to the colony, depth and slope. The results show that in general they have a higher preference for shallow areas, places close to the colony and a steep slope, but they also display large individual variability. Preference for these variables does not seem to differ between individuals of different sizes. In some of the most western colonies, some individuals seem to spend more time further away from the colony, but apart from this, there is no apparent effect of the longitudinal capture position on preference. The predicted spatial distribution is largely driven by the distance from the departure colony and the actual distribution of individuals among the different colonies. Although depth and slope does seem to effect their distribution to some extent, the most important variables that explain fine-scale foraging activities at-sea are probably missing. Future studies using GPS transmitters attached to animals, that yield high resolution locations and more detailed environmental data, in combination with the analytical technique presented here, should provide more insight there foraging decisions. This should eventually also improve the spatial prediction of the population as a whole