ISTS Symposium43 Program/Agenda

The availability of sea turtle reference genomes and the price reduction of Whole Genome Sequencing (WGS) techniques coupled with the increased computing capabilities will lead to a near future where whole genome resequencing of hundreds of individuals will be available, even for species with large genomes. However, the budgets for population genomic studies are currently still limited and genome reduced representation sequencing techniques (RADseq) are generally used to unveil sea turtles’ population differentiation, origin of stranded individuals, or foundation of new nesting areas among others. Different genome reduction techniques are being used in different species and laboratories rendering cross-studies difficult to compare. Moreover, samples will not always be available for comparisons across space and time and thus there is a need to develop a methodology to allow merging the present and the future of sea turtle genomics. In anticipation of that day, we used data from 11 Caretta caretta individuals from four Regional Management Units (RMUs), two from the Pacific, one from the North West Atlantic, five from the North East Atlantic and three from the Mediterranean. We sequenced each individual using both 2b-RAD and WGS to assess the potential of combining genomic data obtained using different methodologies. We evaluated the population structure inferred by the two types of data and how to merge them using four different strategies. As a preliminary step, we genotyped and filtered each methodology separately and observed the same pattern of genetic differentiation among individuals with 2b-RAD (4,523 SNPs) and WGS (8,062,514 SNPs), with the two first axes separating individuals according to their RMU. In the first strategy, we jointly genotyped the raw reads from both methodologies and kept all filtered SNPs (16,877 SNPs) resulting in the same pattern of genetic differentiation observed above in the first axis but the second axis separating methodology. In the second strategy, we genotyped and filtered separately both types of data and kept the 2bRAD SNPs in the WGS dataset (3,669 SNPs). We observed an overlap of the genotypes obtained with both methodologies, and separation according to RMU. In the third strategy, we jointly genotyped the raw data from both methodologies and kept the 2bRAD SNPs (4,539 SNPs) and observed the same pattern as in the second strategy but keeping higher SNP number, demonstrating the feasibility of combining RADseq and WGS data. Finally, we evaluate the percentage of concordant and discordant loci in the three strategies and discuss the nature of the discrepancies. The results obtained demonstrate that it is possible to combine RADseq and WGS data and we provide a method for optimal merging of the data. The integration of RAD and WGS data will be key for combining genetic analyses across laboratories and merging data across oceans. Sea turtles' conservation will be more effective if its wide distribution can be analysed globally, since they can migrate over long distances during their feeding period. Our results lay the foundations for future genomic investigations and ensure that the samples analysed currently will help refining the origin of unknown individuals.

Assessing conservation status and pursuing applicable management priorities for marine megafauna across multiple scales pose significant challenges. Because marine turtles exemplify these challenges, the IUCN Marine Turtle Specialist Group (MTSG) developed the ‘conservation priorities portfolio’ (CPP) framework in 2011 to evaluate population risk and threats for regional management units (RMUs). Here, the MTSG updated the 2011 CPP framework through an inclusive assessment process. Expert elicitation results involving 145 individuals from 50 countries indicated that marine turtle conservation status appears to be improving, but significant challenges remain. Since the previous assessment, long-term trends increased on average, and threat impact scores improved for nearly twice as many RMUs (53%) as worsened (28%) (≥ 10% threshold for changes in numeric scores). While expert-assessed threat impacts have generally decreased, fisheries bycatch remains the highest scored threat across regions and species. Risk-Threat categories improved for a majority (54%) of RMUs. Over 40% of RMUs were scored as Low Risk-Low Threats, eight of which were green turtles (Chelonia mydas). Less than 20% of RMUs were scored as High Risk-High Threats, four of which were leatherback turtles (Dermochelys coriacea). Most High Risk-High Threats RMUs were in the Pacific Ocean, while most Low Risk-Low Threats RMUs were in the Atlantic Ocean. Eleven RMUs were evaluated as critical data needs. Our results—also provided through an interactive data dashboard—underscore the importance of context-specific planning to effectively target limited conservation resources. Future assessments should further prioritize inclusion of under-represented topics, researchers, and regions to better address multi-faceted conservation challenges.

Highly migratory species pose challenges in conservation and management due to the complexity of their life cycles. For these reason, different categories of conservation units have been defined for marine turtles at different geographical scales to cover the management needs. Management units (MUs) were first defined as population genetically isolated thus deserving an independent management. Mitochondrial DNA has been widely used to define MUs for all species, although the male component is not considered while using this maternally inherited marker. However, turtles from different MUs often share migratory routes and foraging habitats. Regional Management Units (RMUs) were thus defined by integrating habitat use with the already defined MUs to represent distinct regions for conservation and management purposes. Finally, an intermediate level of SubRegional Management Units (SubRMUs) has been proposed for the Mediterranean based on shared modelled hatchling trajectories among different MUs. To test the genomic signatures matching these potential levels of structuring for conservation in the loggerhead turtle, we sampled 278 individuals at 13 nesting populations from three regional management units: Mediterranean (MED RMU: 238 individuals from 3 SubRMUs (Greece, Libya and Levantine) including 11 populations), North-West Atlantic (NWA: 15 individuals from 2 populations) and North-East Atlantic (NEA: 25 individuals from 1 population). We genotyped all individuals using 2bRAD methodology resulting in 6,626 biparentally inherited SNPs distributed homogeneously across all chromosomes. All the individuals were clearly clustered in three groups using a DAPC that correspondes to the three RMUS. When plotting only the individuals from the Mediterranean RMU, three clusters were obtained with some overlap, corresponding to the three proposed SubRMUs. When performing pairwise comparison among populations, three different levels of F_st were detected from RMU (mean F_st 0.05294 ± 0.009, mean ± SD), SubRMU (mean Fst 0.01180 ± 0.005, mean ± SD) and MU (mean F_st 0.00438 ± 0.002, mean ± SD) confirming the hierarchical structure of the species. Within the Mediterranean, up to seven Management Units were defined 1) Zakynthos and Kyparissia, 2) Messara, 3) Rethymno, 4) Dalyan, Belek, Akamas and Alagadi, 5) Lebanon, 6) Israel and 7) Lybia. Considering that the RMU and SubRMU levels were initially defined using dispersal information on hatchlings, juveniles and adults, we can conclude that dispersal at these stages also influence the general patterns of gene flow leaving a footprint in the genome. Up to this point, we can confirm that despite the highly migratory behavior, the philopatry of the species is reflected in the genome of the species. Our findings demonstrate that genomic tools provide robust insights into the high genetic structuring of Caretta caretta at three different hierarchical levels: RMU, SubRMU and MU. These results have critical implications for conservation strategies, particularly in identifying individuals distributed across overlapping RMUs, ensuring more precise and effective management efforts.

The Southwest Indian Ocean (SWIO), defined here as the waters bounded by the eastern coast of Africa between Kenya and South Africa to 74°E, and from 1°S in the north to 30°S in the south, is recognized as a key breeding and foraging area for hawksbill turtles (Eretmochelys imbricata). This species, whose populations have declined significantly worldwide in the past, is now listed as Critically Endangered by the International Union for Conservation of Nature (IUCN). Current knowledge of the population genetic structure and connectivity in the SWIO is restricted to the northern part of the region, where major breeding populations are located. However, genetic data from other, less important, nesting sites and, importantly, from developmental foraging areas throughout the rest of the SWIO remain limited.

To fill this gap and understand population structure and connectivity within the SWIO and with the rest of the Indian Ocean, this study assessed genetic diversity within hawksbill rookeries and foraging aggregations and investigated the natal origin of juvenile foraging aggregations using mixed stock analysis (MSA).

We analyzed the mitochondrial (mt)DNA control region (CR; 766-769 bp) sequenced from females and hatchlings across five rookeries, and foraging juveniles from nine developmental areas, sampled between 2004 and 2022. Data was further combined with previously published sequences from the wider Indo-Pacific, leading to a comprehensive genetic dataset for hawksbills throughout the Indian Ocean.

Our analysis identified 32 haplotypes, including two new found in rookeries and six in foraging groups, and revealed a high haplotypic diversity. Results confirmed strong genetic differentiation between SWIO rookeries and those in the broader Indo-Pacific, with only two haplotypes shared between groups and the rest unique to the SWIO region. High genetic connectivity was also observed among SWIO rookeries, except for the small population in the northern Mozambique Channel, which showed low gene flow and slight genetic divergence from other rookeries. The MSA showed that juveniles (in all mixtures), originated mainly from the Southwest Indian Regional Management Unit (contributions > 75%), with minor contributions from other Indo-Pacific stocks, found particularly in the Mozambique Channel mixture. However, the presence of orphan haplotypes in our study suggests that there remain possible unknown or unsampled rookeries in the region, or incomplete sampling of known rookeries.

This study is the first to provide a detailed insight into the genetic structure of nesting hawksbill populations and juvenile foraging groups across the SWIO. We reveal distinct genetic patterns along the Mozambique Channel, highlighting the critical role of this region in hawksbill turtle conservation. Further monitoring and sampling should be carried out in these small, isolated, and vulnerable populations to refine our understanding of their genetic structure and inform targeted management strategies. Genetic connectivity studies remain essential to ensure that effective conservation strategies are designed and implemented, particularly to identify and protect isolated or vulnerable populations.

The discovery of nesting leatherback (Dermochelys coriacea) and loggerhead (Caretta caretta) turtles on the beaches of northern KwaZulu-Natal (KZN), South Africa, by the then Natal Parks Board (now Ezemvelo KZN Wildlife) in the early 1960s revealed that these turtles were being poached as they emerged to nest. A conservation monitoring programme was immediately implemented in 1963, with the primary objective to protect the nesting turtles, their nests, and progeny, and the secondary objective to monitor population recovery to evaluate the potential for sustainable harvesting. The secondary objective was abandoned following global declines in sea turtle populations, and the programme remained dedicated to protection and documenting the population response. It continued uninterrupted annually, having now entered its 61st year of implementation, taking its place amongst the longest-running conservation programmes of its kind in the world. Monitoring takes place annually from 15 October to 15 March of the following year along the 90km stretch of beach from Sodwana Bay to the South African/Mozambican border. Daily patrols (morning and night) are undertaken by teams of trained community monitors who document all nesting activity. Nesting turtles are measured and tagged when encountered. Long-term population trends are based on track counts in the 13km long index beach where consistent monitoring effort has been applied since programme inception. Conservation effort over this time has yielded variable success rates in the recovery of these nesting turtle populations. Both species exhibited rapid growth in the first decade of protection, with nesting trends deviating markedly for the 2 species thereafter. The nesting loggerheads exhibited exceptional recovery in the nesting trend, with individual numbers increasing from approximately 200 females to just under 800 individuals during the 2023/2024 nesting season. The nesting leatherbacks has shown a comparatively weaker yet positive response to conservation despite being subjected to identical beach protection protocols and possessing a reproductive strategy that theoretically surpasses that of the loggerheads. This population trend has stabilised over time, having recovered from an initial 5 individuals in the beginning to the current population size of approximately 80 individuals. The turtle conservation programme represents a complex interplay of successes and ongoing challenges. Despite this, South Africa ranks amongst the lowest for sea turtle poaching incidences in the region, thanks to robust legislation and the positive relationship between local communities and management authorities of the now iSimangaliso Wetland Park WHS which protects the turtle nesting beaches of South Africa. This programme is a model for sustainable, non-consumptive use of endangered species, having transformed the perception of turtles from a short-term food source to a long-term provider of economic opportunities via turtle-based tourism and employment opportunities as turtle monitors. This programme will continue to evaluate the latest population parameters for these two nesting species, their population trends and implementing adaptive management strategies where possible to enhance conservation outcomes. The IUCN statuses of these sub-populations demand a concerted effort to ensure their long-term survival, emphasizing the importance of both effective protected area management and broader collaborative conservation initiatives that extend beyond protected area boundaries.

Population abundance and trend analysis are essential to conduct conservation assessments at local and global scales, and for guiding conservation management strategies and policy. For sea turtles, monitoring population abundance is particularly challenging because of their scale of distribution, migratory nature, complex life cycles, long generation times and cryptic early life stages. For this reason, nesting beaches serve as a valuable focus, mainly because nests can be easily counted and evaluated, compared to the difficulty of counting individuals at sea. Even though nest numbers are an indirect representation of adult female abundance (only one part of the global population), any fluctuations observed in nest numbers are the cumulative result of events (threats, impacts, food abundance, etc.) that occurred prior, on the nesting beaches or in the water. These limitations notwithstanding, annual nest counts remain the most helpful index of sea turtle population abundance and trends, which can be widely and systematically obtained, compared among locations, and tracked over long time series.

The Northeast Atlantic RMU, breeding in Cabo Verde (West Africa), has been categorized as “endangered” in the last IUCN Red List status review, mainly because their area of occupancy is relatively small (< 500km²), and the rookery is subject to constant anthropogenic pressures causing a decline in habitat area, extent and quality. This RMU has been considered the third largest loggerhead nesting subpopulation in the world (after the Northwest Atlantic RMU (principally Florida, United States) and the Northwest Indian Ocean RMU (mainly on Masirah Island, Oman), both with more than 20,000 nest per year), based on first census conducted on Boa Vista Island (12,000 – 20,000 estimated nests per year from 2007 to 2009). Moreover, in the last 15 years different NGOs and local associations have been actively involved on sea turtle conservation in the entire archipelago, and relevant information about nesting female’s abundance and nest trends has been obtained, demonstrating that the Cabo Verde breeding colony is considerably larger than previously estimated.

Data from the four islands with more than 10 years of standardized beach surveys (Boa Vista, Sal, Maio and Santa Luzia) were exhaustively analyzed to obtain the first nest trend analysis of this important rookery, showing that the nest trend of the Cape Verde loggerhead colony has experienced a strong increase, unique in the world, turning Cabo Verde the nesting site with larger number of loggerhead nests in the world (average 150,000 estimated nest per year in the entire archipelago from 2019 to 2023). Different hypotheses were proposed and analyzed to try to understand and explain the main causes of this important increase.