ISTS42 Program/Agenda

An understanding of animals’ behavior is critical to determine their ecological roles and to inform conservation efforts. However, observing hidden behaviors is challenging, especially for animals that spend most of their time underwater. Using animal-borne devices, we investigated the fine-scale behavior of internesting hawksbill turtles nesting on the northeastern coast of Brazil. Our study analyzed data from ten turtles, including two that were tracked for two nesting seasons for a total sample size of 12 deployed FastGPS satellite transmitters equipped with time-depth recorders. We estimated latent behavioral states employing the mixed-membership method for movement (M4) and integrating dive variables (i.e. dive depth, dive duration and surface duration) with spatial components (i.e. on-land vs in-water, step lengths, utilization distribution). We identified five latent behavioral states: 1) pre-nesting, 2) transit, 3) quiescence, 4) area restricted search (ARS) within the residence, and 5) ARS near the residence of turtles. The last three states combined were categorized as “residency period”. Pre-nesting behavior (5.5% of internesting), characterized by shallower and remarkably long dives (up to 292 minutes), highlighted the turtles' preparation for egg-laying and lasted 22.7 hours on average. Transit behavior (13% of internesting), distinguished by the longest average step length, indicated active movement to the residence, and after a residence period, movement back to the nesting beach, and lasted 2.3 days on average. Quiescence (56.1% of internesting), the most predominant behavior, showed the lowest activity level and lasted 11.3 days on average. ARS within the residence (18% of internesting) showed relatively larger step lengths than quiescence, but was still considered a low activity level and lasted 5.8 days on average. ARS near the residence (7.4% of internesting) showed step lengths and dive duration proportions similar to quiescence and ARS within residence combined, but outside of the residence core area and lasted 3.6 days on average . We noted high fidelity to residence core areas and nesting beaches, within and between nesting seasons. Initial residence areas tended to be larger than subsequent ones within a season, likely reflecting more exploratory behavior at first, which then decrease over time as turtles adopt more quiescent behavior, conserving energy for reproduction. These behaviors offer detailed insights into turtle ecology and behavior during internesting periods, which is critical for marine turtle conservation and management. Quantifying these individual behavioral states improves estimates of turtles' spatially explicit susceptibility to various threats, such as vessel strikes and entanglement in fishing gears. For example, our analysis suggests that implementing a proposed Marine Protected Area (MPA) to safeguard a reef formation, currently under review, would significantly increase the protected area from the existing 0.4% to 30% of the turtles' internesting habitats in our study region. This study provides valuable guidance for the conservation and management of internesting marine turtles at a fine spatiotemporal resolution, enhancing national action plans for endangered species and the success of Marine Protected Areas. By incorporating biologically informative parameters, this approach can be applied broadly to study behavior beyond the hawksbill breeding season and other species.

Information yielded by satellite-tagged sea turtles is useful to generate understanding about multiple aspects of turtle ecology, like where they feed, the routes they follow during their nesting migrations, and how often they nest. This information can in turn be applied to answer many questions, like the type and severity of risks they face at feeding and nesting areas and on migration routes, and converting counts of tracks on beaches into estimates of turtle abundance. Data from green turtles Chelonia mydas at Ningaloo has shown that they move relatively little, other than for nesting or mating. The median displacement (distance between locations of capture and final transmission, transmitted over 72-416 days) of 19 “resident” turtles captured in the water (74-108 cm CCL) was 2.3 km, while the median displacement of 13 females tagged on the beach following nesting (94-104 cm CCL) was 179 km. Acoustic tagging provided further details about the locations and movement patterns of resident turtles. After nesting, females migrated either north or south, between the Kimberley and Shark Bay (a span of 10 degrees of latitude and ~1,500 km). Multiple turtles with vitellogenic (yolk-bearing) follicles identified with ultrasound were followed for their entire nesting migration. Each migrated >200 km from Ningaloo to nesting beaches on islands off the Pilbara coast, and then returned to Ningaloo (in one case the final transmission was 400 m from the location she was captured). Examination of GPS locations transmitted during the nesting period provided information on the likely number of successful nesting attempts (~4-7 clutches per individual).

Satellite telemetry is often used to track sea turtle movement patterns because transmitters can be configured to monitor and collect valuable information including environmental and behavioral (i.e., dive profiles) data of tagged individuals. These data can then be related to horizontal and vertical movement patterns to understand how tracked individuals interact with their environment at specific locations. Leatherback sea turtles (Dermochelys coriacea) nesting in southeastern Florida remain one of the few stable or increasing leatherback aggregations in the northwest Atlantic (NWA) RMU since 2000. The U.S. Northeast Shelf is one of the most globally productive marine ecosystems which may contribute to the success of this nesting aggregation, but declining NWA nesting stocks from other areas also use similar foraging grounds. Environmental conditions specific to this nesting aggregation may further contribute to the stability of southeastern Florida leatherbacks; however, no published studies have described movement patterns and environmental conditions encountered at high-use habitats for this aggregation since 2000–2002. Here, we aim to monitor female leatherback turtle movements with satellite transmitters to collect and relate environmental data with movement patterns and geographic locations. Thirty adult female leatherbacks were tagged after nesting with Wildlife Computers, Inc. SPLASH10-295I (N=26) and Sea Mammal Research Unit CTD/Fluorometer Oceanography Satellite Relay Data Logger (N=4) satellite transmitters during 2022–2023 on Juno Beach, Florida, USA (26.9433N, -80.0708W) using direct carapacial attachment. An optimized hot-spot analysis revealed that inter-nesting habitats in Florida covered approximately 12,694 km², ranging from Cape Canaveral southward to West Palm Beach. Post-nesting movements indicate that southeastern Florida leatherbacks tend to migrate north along the Gulf Stream before dispersing to four distinct foraging regions: Mid-Atlantic Bight (ranging from North Carolina to New Jersey), northeast Canada (ranging from Nova Scotia to the Gulf of St. Lawrence), northeast United States (ranging from Cape Code, Massachusetts to the Gulf of Maine), and pelagic waters (ranging from Sohm Plain to the New England Seamounts). Only one tagged turtle is currently foraging in the South Atlantic Bight. To date, one satellite transmitter deployed in 2022 remains active 513 days after initial deployment, and 18 transmitters deployed in 2023 remain active (mean ± SD: 183 ± 87 days). A space-state switch model allowed us to determine when and where leatherbacks switched from migratory to foraging behavior, and tracks will be compared to leatherbacks tagged on Juno Beach from 2006–2011, allowing us to see if inter-nesting and post-nesting movement patterns have changed over time. Results will provide insight on the movement patterns and current locations of high-use habitats of southeastern Florida leatherbacks. Future work will help establish baseline data regarding how environmental conditions influence leatherback distribution and behavior. These data can then be incorporated in future studies to examine and help predict how leatherback distribution and dive behavior will be affected by ongoing and future changing environmental conditions, most notably those related to climate change.

Loggerhead turtles, Caretta caretta, born on the nesting beaches of the Northwest Atlantic Ocean undertake a long transoceanic migration immediately after birth, traveling eastward in association with the Gulf Stream. Most of them reach the coasts of Europe and northwestern Africa when two or three years old and 20-30 cm in curved carapace length (CCL) and remain in the oceanic foraging grounds of the eastern Atlantic for several years, before migrating back to the western Atlantic when 40-60 cm CCL. On the other hand, some of these loggerhead turtles enter the Mediterranean Sea during their developmental migration, but the timing of that entry and the length of the period spent inside are poorly known. Here, we combined skeletochronology with the analysis of the stable isotope ratio of oxygen (δ¹⁸O), carbon (δ¹³C) and nitrogen (δ¹⁵N) in the cortical bone of the humerus of 31 juvenile loggerhead turtles of Atlantic origin found dead stranded in the Balearic Islands. We sampled different incremental layers (n = 74) to assess the timing of the entry into the Mediterranean basin and the existence of any ontogenetic change in the diet. In addition, we used the positive and linear correlation between sea surface salinity and δ¹⁸O to analyse the habitat use and identify individual movements between water masses. The surface waters of the Mediterranean Sea are salty and enriched in ¹⁸O due to a negative water balance, and hence, the entry into the basin could be traced by the δ¹⁸O values in the incremental layers of the cortical bone of skeletal elements such the humerus. The studied turtles measured between 28.00 and 80.64 cm CCL and the estimated age of the sampled layers ranged from 3 to 15 years old. Thus, the incremental layers corresponding to the first years of life have been reabsorbed and, as a result, the δ¹⁸O values corresponding to the years spent in the Gulf Stream were missing in all individuals. Nonetheless, the broad variability found in the δ¹⁸O values of the remaining incremental layers suggest that juveniles moved between water masses with different salinity levels before stranding in the Balearic Islands. The observed range of δ¹⁸O values encompassed those from the eastern Atlantic Ocean, the western Mediterranean basin, and the much saltier eastern Mediterranean basin, without any consistent temporal pattern. This suggests that the entry into the Mediterranean Sea may happen at any time during their stay in the eastern Atlantic Ocean and, once inside, these juvenile loggerhead turtles can follow a diversity of trajectories across the entire basin. Nevertheless, upon reaching ten years old they consistently settle in low salinity areas such as the southern Algerian Basin or the Alboran Sea, likely preparing for their return towards their natal beaches in the western Atlantic Ocean. Finally, the changes observed in the δ¹³C and δ¹⁵N values of the analyzed incremental layers were small, and did not indicate any ontogenic change in the diet of juvenile loggerhead turtles during their journey through the eastern Atlantic Ocean and the Mediterranean Sea.

Advances in satellite telemetry and remote sensing have greatly improved our understanding about the spatial ecology of adult sea turtles. However, a significant gap remains regarding the early sea turtle life history, commonly referred to as "The Lost Years”. Neonates, measuring just a few centimeters in length, enter the ocean and are rarely observed until they reappear years later as subadults and adults near their natal beaches or within foraging habitats.

Numerical models have emerged as powerful tools to shed light on these cryptic years. The Sea Turtle Active Movement Model (STAMM) significantly advances our understanding about the dispersal of sea turtle hatchlings and juveniles, by simulating their movements under the influence of oceanic currents (passive drift) and habitat-driven swimming movements (active dispersal).

STAMM has been previously calibrated to simulate the dispersal of leatherback turtles, but our research is dedicated to adapting it for loggerhead populations. Our calibration effort aims, through sensitivity studies and parameter estimates, to consistently simulate the general dispersal patterns and features of loggerhead sea turtles by successively studying passive and parameterized active dispersal, allowing us to separate the impacts of the local oceanic configuration from the intrinsic impact of swimming activity.

This presentation seeks to unveil preliminary findings from the ongoing STAMM loggerhead modeling efforts, emphasizing our first results on passive dispersal of juvenile loggerhead turtles originating from southeastern Florida (USA), and Cape Verde nesting sites. Our early investigations highlighted differences between passive dispersal patterns from both nesting sites but also already identified new potential dispersal areas in the Sargasso Sea, the Caribbean Sea and the Gulf of Mexico, specific to juveniles from the poorly understood Cape Verdean population.

By employing STAMM, we aim to fill a critical gap in sea turtle research and contribute to the delineation of loggerhead Regional Management Units, incorporating the juvenile stage dispersal information. This knowledge is pivotal for informing decision-making processes related to implementation of effective sea turtle conservation and management measures.

As rapid environmental changes continue to affect marine ecosystems, understanding the distribution of juvenile marine turtles in their dispersal stage is crucial for conservation and management. The "lost years" life history stage of green turtle development is difficult to study due to logistical constraints of accessing and studying small, pelagic individuals. While theoretical and captive-reared approaches have contributed to our understanding of dispersal stage green turtle behavior, wild-caught individuals provide essential insights into natural dispersal and habitat utilization that cannot be gained by other methods. To address this knowledge gap, we must develop a comprehensive understanding of the drivers behind movement and habitat selection in dispersal stage juvenile turtles. We focused on the Gulf of Mexico (GOM), where the sea surface temperature is warming at a rate roughly double that of the global oceans. Satellite tracks from 77 juvenile green turtles were used to plot turtle presence in the GOM from 2011-2022. The average tracking duration for turtles in this study was 31.3 (±13.61) days providing 7,394 green turtle locations in the GOM. Environmental variables associated with sea surface convergence zones (including sea surface temperature, eddy kinetic energy, etc.) were used to inform species distribution models (SDMs) created in the R package Biomod2. Area under the curve (AUC) and the true skill statistic (TSS) were used to determine which model performed the best. We identify previously undocumented habitat for juvenile green turtles in the Northern and Eastern GOM based on the predicted distribution. The results of our SDM provide a template for identifying essential habitat in the GOM for juvenile green turtles. Future models can be used to determine the distribution of dispersal stage green turtles elsewhere, or extend to other life-stages. Identifying the distribution of this cryptic life stage is essential for understanding how juvenile turtles interact with their environment and for predicting how they may respond to rapid climate change.

Hawksbill sea turtles (Eretmochelys imbricata) inhabiting the Hawaiian Islands (known as honuʻea or just ʻea in Hawaiian) are genetically isolated and represent one of the smallest distinct sea turtle populations on the planet. Despite their endangered status and recent advances in understanding their biology, many data gaps remain. The post-nesting migrations of this population were previously presented during ISTS41. Here, we attempt to further describe their migratory behavior by incorporating inter-nesting data, along with insights into their foraging behaviors throughout the Hawaiian Islands. We deployed satellite tags on a total of 14 post-nesting hawksbills that collected information on Argos and GPS locations, as well as dive parameters (time at depth, maximum dive depth, and dive duration), with an emphasis of deploying tags early in the nesting season. Additionally, we collected tissue samples from 4 nesting females and conducted bulk tissue and amino acid (AA)-specific stable carbon (δ¹³C) and nitrogen (δ¹⁵N) isotope analyses. We also collected gastrointestinal tracts from 14 necropsied hawksbill turtles for gut content analyses to gain further insights into the diet of this population. The majority of turtles migrated to the coast of Maui (n = 7; 50%) with only two off the coast of Molokaʻi (14.3%), one remained within Hawaiʻi Island (7.1%), and one off the coast of Oahu (7.1%). No final location was transmitted for two turtles, and the transmitter failed on the remaining turtle. Nevertheless, they all demonstrated relatively short-distance migrations and remained within the Hawaiian archipelago. Throughout their inter-nesting intervals, they also remained in close proximity (i.e., <2 km) to their nesting beach while exhibiting a U-dive pattern, utilizing deeper depths between nesting presumably to find a suitable resting place. Hawksbill turtles are considered spongivores in the Caribbean but data from the Eastern Pacific suggests they consume a wide variety of prey. Bulk stable isotope values from Hawaiian hawksbills ranged from -15.0‰ to -13.9‰ and 9.9‰ to 11.6‰ for δ¹³C and δ¹⁵N, respectively. These findings align with those in the Eastern Pacific, suggesting that hawksbill turtles in the Hawaiian Islands likely have a diverse diet, consuming a range of prey items. AA-specific δ¹³C and δ¹⁵N isotope analyses as well as gut content evaluations are currently being analyzed and will help verify these preliminary results, as well as additional potential insights into diet. The results of this work will be shared during the symposium. This study represents the first attempt to provide insights into the inter-nesting movements and foraging habits of Hawaiian hawksbill turtles, filling an important knowledge gap in their ecology and behavior. Determining actions to mitigate threats and designate protected areas is essential, particularly for a population that relies on a limited geographic area throughout its entire lifecycle. We aim to gain a comprehensive understanding of this highly threatened population, including the identification of movement patterns, foraging habits, and habitat preferences.