Loggerhead and green turtles of the south west Pacific: investigating foraging ecology and migration using integrated analytical methods

Abstract

This study sought to refine the feeding behaviours and foraging regions of resident loggerhead and green turtle populations in Queensland, Australia, using a combination of diet investigatory techniques, satellite telemetry and stable isotope analysis. Discoveries consequent to the development and application of new analytical techniques, highlights the need to re-evaluate longstanding assumptions about marine turtle foraging ecology. The study of marine turtles has historically been limited by the challenges of investigating animals in the marine environment. Due to the complex developmental history of marine turtles, which follow distinct life stages, generally research has focused on sampling at reproductive events where large numbers of female turtles are readily accessible. Consequently, using capture-mark-recapture methods, the approximate migratory range of a sampled nesting population could be established. However, the proportion of the nesting population originating from distinct foraging regions has not yet been resolved. The incorporation of satellite telemetry and stable isotope analysis (SIA) have the potential to document the ecology of marine turtle populations more precisely than was previously possible. The use of these methods in tandem with more traditional foraging ecology tools (e.g. gut content analysis, oesophageal lavage or faecal sampling), provides important information on foraging locations and the food sources that can be used to develop conservation strategies at a local through to a national scale for marine turtles.nStudies conducted in the north-west Atlantic have successfully defined the foraging regions of nesting loggerhead turtles based on differences in the stable isotope ratios d13C and d15N from sampled blood and skin tissues. However, despite long-term turtle monitoring programs identifying nesting and foraging populations of green and loggerhead turtles throughout Queensland, the question of how distinct foraging regions contribute to nesting populations has yet to be resolved. In addition, foraging ecology studies of marine turtles using SIA have discovered that within a species, individuals can have very different foraging strategies. This finding contrasts with traditional views that most marine turtle species, once they grow to the size of juveniles, recruit to the near-shore environment and adopt the same benthic feeding strategy. Hence, this study aimed to use SIA methodology on the east Australian populations of loggerhead and green turtles to discover more information on their feeding regions and foraging ecology.nLoggerhead and green turtles are found throughout coastal waters, bays and reef structures along the tropical and sub-tropical eastern coast of Australia. Investigating the variation in the foraging ecology of these species will provide information for the development of local management strategies targeting the long-term conservation of marine turtles in Australian waters.The aim of this thesis was to incorporate SIA and telemetry data with the analysis of faecal and oesophageal samples to:1) determine habitat use and dietary composition of loggerhead and green turtle foraging populations of south east Queensland;2) ascertain if d13C and d15N stable isotope values from foraging turtles could be used to differentiate local or regional foraging habitats; and3) identify the foraging regions of green and loggerhead turtles nesting at major nesting sites in Queensland waters.nFollowing the success of studies that used SIA to define foraging regions on nesting loggerheads in the north-west Atlantic, SIA was used to investigate if the foraging regions of Australian nesting loggerhead and green turtles could be identified from epidermis or blood samples. The isotopic values of nesting loggerhead turtles from Mon Repos, central Queensland were sampled and compared to a cohort of nesting individuals with known foraging locations and a subset of foraging loggerheads from a known foraging region. Unlike comparable studies from the north-west Atlantic, the isotopic values of sampled nesting turtles from the east coast of Australia did not group into potential foraging regions and no correlation was observed between latitude and the isotopic values of sampled turtle tissues. Consequently, foraging grounds of nesting females could not be inferred using SIA of d13C and d15N values from epidermis or blood samples. It was hypothesised that the absence of the expected unique d13C and d15N values associated with foraging regions was related to the unique oceanic currents along the east coast of Australia and the influence of the Great Barrier Reef and coastal outflows on sea surface temperature and water residence times.nGreen turtles were sampled at two geographically and genetically distinct rookeries, Raine Island in the northern Great Barrier Reef (GBR) region and Heron Island in the southern GBR region. It was expected most individuals from these distinct rookeries would have latitudinally separated foraging grounds. Consequently, a hierarchical cluster analyses was used to determine if nesting individuals from these two rookeries could be differentiated by their tissue isotopic values. Nesting individuals from these geographically distant rookeries could not be distinguished based on d13C and d15N values from sampled bodily tissue. Consistent with this finding, there was no correlation between d13C or d15N and latitude in a cohort of 40 green turtles tracked via satellite telemetry from their Raine Island nesting beach to their foraging grounds. Hence, as with the east Australian population of loggerhead turtles, the foraging grounds of nesting green turtles could not be determined from SIA of sampled epidermis. It was theorised that the absence of identifiable clusters in d13C and d15N values for individuals from the same foraging grounds was related the lack of a prevailing oceanic current through the potential foraging regions. In addition, factors such as the influence of oceanic temperate and coastal outflow, which may confound the development of unique isotopic gradients or isoregions, were also attributed to the inability to resolve the foraging regions of the sampled cohort.nAn analysis of faecal material from loggerhead turtles foraging in Moreton Bay, southeast Queensland indicated a diet composed predominantly of benthic crustaceans, bivalves and gastropods, consistent with previous observations. However, SIA of blood and epidermis discovered large variations between sampled individuals suggesting that, either the loggerhead diet included a wider range of animals than the faecal sampling method provided, or that the dietary items themselves had highly variable d13C and d15N values such that they could not provide an consistent baseline to infer the diets or feeding areas of sampled turtles.nThe diets of foraging green turtles in Port Curtis, central Queensland were explored using oesophageal lavage and SIA of blood and epidermis to investigate trends in diet composition from different foraging habitats within this region. While oesophageal lavage samples indicated diets were consistent with the food items available within their foraging habitat, a MixSIAR hierarchical Bayesian mixing model, using d13C and d15N isotopic values, identified food items which were under-represented in lavage sampling that had a greater contribution to dietary composition. In addition, animal material appeared to make a significant contribution to the diets small (l 55 cm CCL) green turtles at this location. nElevated d15N, relative to the values observed in primary producers, in turtles l 55 cm in curved carapace length (CCL), suggested that these turtles continue to forage on macrozooplankton after recruitment to neritic foraging grounds. An elevated d15N was also observed in individuals g 90 cm CCL, a finding consistent with the increasing reports of omnivory in adult green turtles. Green turtles of distinct sizes were associated with different habitat types, indicating that individuals potentially exhibit ldevelopmental migrationr to different foraging grounds within Port Curtis as they grow. Small turtles (l 55 cm CCL) were mostly associated with shallow rocky reef and mangrove habitats, whilst medium (g 55 cm CCL) and large (g 90 cm CCL) turtles were commonly found throughout the subtidal sandbanks. A comparison of epidermis and plasma d13C values indicated ~ 40 % of turtles may have changed their diet to/from seagrass in the period between the synthesis of their plasma and epidermis tissues, a finding consistent with previous observations from this foraging region.nnImportantly, and contrary to findings from studies conducted on loggerhead turtles in the north-west Atlantic, SIA could not be used to confidently identify the foraging region of either loggerhead or green turtle individuals from the east coast of Australia. The reasons why this method failed to identify foraging regions along the east coast of Australia were (1) the large variation in SI values in individuals sampled in the same foraging regions, and (2) the lack of identifiable unique iso-regions or latitudinal gradients in d13C and d15N values along the east coast. While this study found the use of carbon and nitrogen stable isotope values in turtle tissue could not be used to identify their foraging region, the incorporation of different isotopic tracers, such as d 18O or d 34S which are less influenced by dietary composition, may better reflect regional differences in habitats and improve the ability to identify the foraging region of marine turtles nesting along the east coast of Australia.