Quantifying the abundance, distribution and behaviour of a generalist marine consumer, the loggerhead turtle, in coastal foraging habitat

Conservation of large-bodied marine taxa, many of which have undergone population declines, is often hindered by difficulty in quantifying spatiotemporal patterns of abundance. This is especially true for diving species that spend little time at the surface, and is largely responsible for gaps in our understanding of the habits of highly migratory, long-lived marine turtles in coastal foraging areas. Here, I use a variety of field and analytical techniques to study these elusive ectotherms on a foraging ground, focusing on loggerhead turtles (Caretta caretta) in Shark Bay, Western Australia. First, I combine stable isotope analysis, animal-borne video and mark-recapture data to describe loggerhead foraging ecology. Low stable isotope (δ13C and δ15N) variance in slow-turnover tissue, particularly among large individuals, combined with evidence of a general diet and strong fidelity to a foraging location suggest that adult loggerheads in Shark Bay are foraging generalists and site specialists. While isotopic evidence of polymodal foraging has been found among females on nesting grounds, this may reflect isotopic characteristics of preferred migratory routes or foraging habitat as opposed to prey specialization within foraging areas. Subsequently, I focus on an important methodological problem in population ecology: detection probability during abundance surveys. During transects for long-diving taxa (e.g., marine turtles, beaked whales) a large proportion of the population will be missed because they are submerged, leading to ‘availability bias’ in count data. I collected dive data for loggerhead and green turtles (Chelonia mydas) and demonstrated that dive-surfacing patterns were highly heterogeneous and correlated with water temperature and habitat depth. Current approaches do not incorporate diving variability, without which abundance indices can be biased and spatiotemporal comparisons unreliable. For example, in seasonal environments marine turtle counts often decrease in colder months. Usually interpreted as emigration, this may also reflect reduced availability (extended dive times), leading to misinterpretation of local population dynamics. Finally, Bayesian methods of incorporating diving variability are applied to survey data, revealing that seasonality, habitat depth and regional characteristics drive variation in loggerhead distribution. Loggerhead density was stable over ten years, and the foraging population in Shark Bay may be substantially larger than believed.