ISTS Symposium43 Program/Agenda

Juvenile green turtles are generally considered to occupy narrow ranges in neritic areas, where they predominantly consume marine algae and seagrass. In recent years, green turtle overgrazing has led to a decline in seagrass resources, thereby reducing the amount of seagrass which is available to them. Investigating how strongly they exhibit fidelity to foraging sites is important for understanding their foraging ecology and conserving seagrass habitats, but few studies have examined this topic. In this study, we evaluate green turtle preference for specific marine algae and seagrasses, and fidelity to foraging sites.

During 2018-2024, field studies were conducted at Kuroshima (24°14’13” N, 124°00’35” E), a small subtropical island with a diameter of about 4 km and surrounded by coral reefs. Green turtles were captured using an entanglement net set in foraging areas in the inner reef on the northern and southern sides of Kuroshima. Identification of food items ingested at the capture site was carried out by esophageal lavage. Animal-borne data loggers were attached to 18 turtles, with four turtles released near the capture site and the other 14 turtles relocated 2-4 km away from their capture site in marine algae or seagrass habitats around the islands. The devices were retrieved after 3-4 days. Food items and feeding time were analyzed from video data, and homing routes were investigated from behavioral data. The feeding time ratio was calculated as the total feeding time relative to the total video-recorded time.

All four turtles released near capture sites stayed there and 12 out of 14 translocated turtles returned to their original capture sites. The feeding time ratio of turtles released near the capture sites ranged from 17.1 to 70.8%. For turtles that returned from the release site, the feeding time ratio before or during their return was low (0-1.9%) but increased near the capture site (0-31.8%). Five of six turtles, from which esophageal lavage samples were collected, consumed the same species of algae that they had consumed at the capture sites, and they rarely consumed other species of algae before or during their return. Of the two turtles that did not return, one consumed the same species at the release site as it consumed at the capture site. The other was released in site where seagrass consumed at the capture site was not growing and did not feed. The returned turtles moved through the inner reef on the east side of the island and arrived at the capture sites in about 4-7 hours, avoiding the west side, where rocky exposures occur at low tide in the inner reef. Turtles translocated to the western area returned through the outer reef on the west side, taking about half a day to one day. The reasons for choosing to return through the inner reef may include familiarity with the terrain and/or escaping predators such as sharks outside the reef. Green turtles indicated strong fidelity to foraging sites. One reason may be their preference to consistently feed on the same species.

Sea turtles are highly migratory species, with nesting sites and foraging areas often separated by vast distances. Individuals from different breeding populations may utilize the same foraging grounds. Consequently, accurately determining the natal origin of individuals in feeding areas is crucial to unveiling the effect of bycatch on effective management and conservation. Genetic techniques offer a powerful solution to these challenges, providing exceptional resolution for evaluating foraging grounds. Over the years, population genetic studies on loggerhead sea turtles have predominantly relied on Mixed Stock Analysis (MSA) based on the mitochondrial DNA (mtDNA) control region, D-loop. However, this approach only estimates the proportion of individuals in a foraging area originating from different nesting sites with large confidence intervals. Contrarily, genomic methodologies, such as 2bRAD, have proven particularly effective for individual assignments, offering a more detailed understanding of population structure. In this study, using 2bRAD we sequenced 232 individuals from nine foraging grounds: one in the Atlantic: Canary Islands (CAN), two in the transition between the Atlantic and Mediterranean: Andalusia (AND1 and AND2) separated by the Straits of Gibraltar, and six in the Mediterranean: Valencia (VAL), Balearic Islands (BAL), Catalonia (CAT), Lampedusa (LAM), western Aegean Sea (WAG), and eastern Aegean Sea (EAG)). We used a baseline previously built from data collected at well-established nesting sites, representing the potential origins of these individuals. We genotyped the baseline and the foraging ground individuals together and performed individual assignments for the unknown individuals. Our approach followed a hierarchical structure, starting with a global perspective with the three regional management units (RMUs) that could contribute to the assayed foraging grounds: West Atlantic (WAT), East Atlantic (EAT), and Mediterranean (MED). Then, we narrowed it down to a subregional level for the Mediterranean nesting individuals. We successfully assigned each individual to one of the three RMUs. We observed a decrease in individuals assigned to Atlantic origins in the nine foraging grounds following the entrance of Atlantic water in the Mediterranean. At the Mediterranean subregional level, we observed different contributions of the three subRMUs (Greece, Libya and the Levantine Region) and the presence of admixed individuals among them. Additionally, we compared our newly assigned samples using 2bRAD with previous foraging grounds assignments derived from MSA on mtDNA, both at RMU and subRMU levels. We observed a decrease in individuals assigned to Atlantic origins and an increase in those linked to Mediterranean nesting populations, particularly from Levantine sites contributing to Western Mediterranean foraging areas. While some differences in assignment data may reflect methodological variations due to the low resolution of the mtDNA, the significant discrepancies observed suggest potential temporal and spatial shifts in the composition of Mediterranean foraging grounds. However, additional research is needed to discard any methodological influences and to fully understand these shifts and their implications.

Telemetry and stable isotope analysis (SIA) are two complementary tools that can be used to gain insights into sea turtle foraging location and ecology. Satellite and acoustic telemetry allow for the tracking of sea turtle movements; however, the high cost of transmitters often results in limited sample sizes. Stable isotope analysis provides a cost-effective method when compared to biotelemetry to gather information on feeding ecology and trophic position by analyzing stable isotope ratios of light elements, but prior knowledge of geographic locations and related isoscapes is required for analysis. Combining both methods can help describe the diet and foraging strategy of an individual at its known forging areas. If isotopic signatures from distinct foraging areas are significantly different, individuals can then be assigned to putative foraging grounds based on SIA alone. Leatherback sea turtles (Dermochelys coriacea) primarily feed on gelatinous zooplankton, have broad-ranging migrations, and exhibit fidelity to discrete foraging regions, making them ideal candidates for using SIA to infer foraging grounds prior to nesting events. Previous SIA studies of northwest Atlantic leatherbacks only infer foraging grounds by grouping similar δ¹³C and δ¹⁵N values, which does not give information about specific locations, and none have paired telemetry with SIA. Here, we assign isotopic signatures to known foraging locations and suggest the best tissue to sample to infer foraging areas for leatherbacks. In 2023, we tagged 25 nesting leatherbacks with satellite (N=20) and acoustic (N=5) transmitters and sampled 30 individuals for SIA on Juno Beach, Florida, USA. Whole blood, plasma, serum, red blood cells, skin, and egg samples were collected for the analysis of δ¹³C and δ¹⁵N values. Mean values of δ¹³C and δ¹⁵N varied among tissue type and location of foraging grounds. Red blood cells had the lowest δ¹⁵N value (mean: 10.95‰; range: 6.92–13.10‰) and skin had the highest δ¹⁵N value (mean: 13.16‰; range: 9.53–17.99‰). Unhatched eggs had the lowest δ¹³C value (mean: -23.40‰; range: -25.09–-21.76 ‰) and skin had the highest δ¹³C value (mean: -17.40‰; range: -18.83– -14.90‰). Results indicate that skin samples are the most enriched in nitrogen and carbon, suggesting that leatherbacks primarily forage in coastal shelf habitats and feed on jellyfishes rather than lower-level zooplankton (i.e., holoplanktonic salps). Skin also has a long turnover rate (> 6 months), thus is commonly used to identify previous foraging grounds for other sea turtle SIA studies. A principal components analysis will be used to illustrate relationships between tissues sampled, and a PERMANOVA with Pillai’s trace test will be run to determine differences in isotopic value and known foraging grounds identified from satellite tracked leatherbacks. A wards hierarchical cluster analysis will be run to delineate foraging clusters based on δ¹³C and δ¹⁵N signatures. Then, a discriminative function analysis will be performed to determine if untracked leatherbacks (N=5) can be assigned to foraging areas. Results will allow us to suggest the best tissues to analyze for leatherback SIA studies, contribute to isotopic values of δ¹³C and δ¹⁵N of northwest Atlantic leatherbacks, and identify foraging grounds for Florida leatherbacks.

Learning the structure and function of ecosystems is crucial for understanding the dynamics of nutrient flow, energy transfer, and species interactions within these systems. This knowledge is not only essential for effective conservation and management, but also for assessing potential impacts of climate change on marine biodiversity. In particular, sea turtles play key roles in marine food webs, serving as indicators of ecosystem health and shifts in community dynamics. However, the trophic strategies of sea turtles remain poorly understood, especially in highly dynamic marine ecosystems such as off the coast of Peru. In this study, we investigated the ecological strategies of four sympatric sea turtle species in Peru (C. caretta, C. mydas, D. coriacea, and L. olivacea) using stable isotope analysis of carbon (d¹³C) and nitrogen (d¹⁵N) from 133 turtle skin samples collected between 2003 and 2009. We found a significant latitudinal trend in d¹⁵N for C. mydas and L. olivacea, likely attributed to baseline isotopic differences, confirming a similar regional trend observed in other taxa. No latitudinal effects were observed for d¹³C. To further explore how these isotopic patterns vary spatially, we examined the three distinct geographic areas in Peru with distinct biotic and abiotic features used by sea turtles: North, Central and South. Our results revealed that both d¹³C and d¹⁵N values characterized these areas for C. mydas and L. olivacea, suggesting region-specific foraging patterns. In contrast, isotopic values did not differentiate these areas for C. caretta and D. coriacea, indicating that these species may be moving among regions. To compare isotopic niche spaces, we focused on the Central region of Peru, where we had a robust sample size from all species. We found that C. caretta and D. coriacea exhibited broader trophic niches, while L. olivacea showed more specialized resource use, and C. mydas exhibited intermediate niche size. C. caretta also exhibited the highest d¹⁵N values, which suggests a higher-trophic level diet and/or use of areas with high denitrification rates (i.e., baseline influence). Additionally, we found moderate niche overlaps for all turtles except C. caretta. Niche overlap among C. mydas, D. coriacea, and L. olivacea ranged from 14% to 45%, indicating greater resource sharing among these species. C. caretta exhibited minimal overlap, confirming its distinct resource use. These findings provide new insights into the trophic relationships and niche differentiation of sea turtles in the southeastern Pacific high seas and Peruvian waters, with important implications for their conservation and management in a changing environment.

The stable isotope composition in the scutes of green turtles (Chelonia mydas) preserves dietary information for up to 10 years, allowing insights into long-term feeding patterns. This study investigates the diet of 44 nesting. female, green turtles in the Red Sea using stable isotope analysis of δ¹³C and δ¹⁵N values from turtle scutes, combined with macrophyte isotope data collated from the literature. We categorised macrophyte isotopic values by habitat and taxa, and applied a mixing model to estimate the relative contributions of various habitats and taxonomic groups to green turtle diets. Our results show that green turtles rely heavily on seagrass habitats, with seagrass species being the primary components of their diet. Notably, Cymodocea nodosa was the most frequently consumed species, followed by Enhalus acoroides. The high consumption of Cymodocea, a genus not typically identified as a primary food source, may reflect its elevated nitrogen content, which could help compensate for the nutrient-poor conditions of the Red Sea. In addition to dietary analysis, we explored the relationship between the curved carapace length (CCL) (mean; 100.73 cm, range; 91.9 - 111.7 cm) of turtles and the distance between their nesting and foraging sites. Using a Gaussian mixture model, we identified two distinct clusters of foraging behavior: “Close” foragers that remain near nesting areas and “Distant” foragers that travel farther to forage. Our analysis found that “Distant” foragers are significantly smaller in CCL, with slightly lower δ¹³C values and significantly higher δ¹⁵N values than “Close” foragers. These isotopic differences may reflect regional nutrient gradients, as nitrogen levels increase in the southern Red Sea near the Bab al Mandeb Strait where it connects to the Indian Ocean. The elevated δ¹⁵N levels in “Distant” foragers could indicate exposure to different nutrient baselines, possibly due to a shift in diet or feeding in distinct isotopic environments. The patterns observed may also relate to the remnants of dietary habits from earlier life stages when juvenile turtles consume invertebrates and plankton. These findings provide valuable insights into green turtle foraging ecology in the Red Sea, identifying key seagrass species that support their diet. The distinct clusters of foraging behavior highlight variability in individual foraging ranges, likely shaped by resource availability and habitat distribution. This research underscores the importance of protecting seagrass meadows and conserving critical foraging habitats that sustain green turtle populations across the Red Sea.

Sea turtles are highly migratory, shifting between various habitats across life-stages and reproductive cycles. Understanding the long-distance movements of sea turtles is therefore key to understanding their ecology as well as developing effective conservation strategies for protecting these endangered species. One potential, yet currently under- utilized tool, for studying sea turtle movement patterns is to use epibiotic organisms as indicators of habitat use. This relies on the concept that some epibiotic organisms only settle on hosts in specific habitats, and so a turtle is colonized with these species must have recently occupied the same habitat. Recent research on loggerhead turtles in the Mediterranean has shown that the barnacle genus Chelonibia likely suggests that the host turtle was recently in coastal habitats, while the barnacle genera Conchoderma, Lepas, and Platylepas are indicative of recent residence in offshore waters. Nevertheless, it is not known if these trends are applicable outside of the Mediterranean and across all sea turtle species. Here, we utilized a global database of over a century of records to examine how the frequency of occurrence of these four barnacle species (Chelonibia, Conchoderma, Lepas, and Platylepas) differ between turtles in different habitats and life stages. Our results reveal that, across the four sea turtle species with sufficient data (loggerhead, green, hawksbill, and olive ridley turtles), Chelonibia spp. were significantly more associated with turtles inhabiting coastal areas (< 10 km) (83%), while Conchoderma spp. and Lepas spp. were significantly more associated with turtles inhabiting offshore areas (> 10 km) (70 and 60%, respectively). In contrast, our results indicated Platylepas spp. were not significantly more associated with turtles inhabiting offshore areas (30%). Notwithstanding all this, most data came from loggerhead (N = 16) and green turtles (N = 19), confirming Chelonibia spp., Conchoderma spp. and Lepas spp. as indicators of habitat use for this species, although this may also apply to most species. Some barnacles can serve as ecological indicators of habitat use in sea turtles on a global scale, although further data is required to confirm this hypothesis for all sea turtle species.