Green sea turtles typically undertake long-distance migrations between nesting sites and foraging areas (Seminoff et al. 2008; Southwood and Avens 2009; Godley et al. 2010). Nesting sites may host turtles that originate from multiple foraging areas (Limpus et al. 1992; Bjorndal et al. 2005), and foraging areas may be home to turtles from multiple nesting sites (Casale et al. 2008; Carreras et al. 2011). The linkages between foraging areas and nesting sites have been documented via genetic studies (Bjorndal et al. 2005; Bowen and Karl 2007), stable isotope studies (Bradshaw et al. 2017), and satellite tracking (Schofield et al 2013; Hays et al. 2014). Understanding this connectivity between nesting sites and foraging areas is a key component of conservation planning as it allows for the development of effective conservation agendas (Martin et al. 2007; Hamann et al. 2010; Hays et al. 2014; Dunn et al. 2019).

The Arabian region is an important global sea turtle hub, with a vast loggerhead turtle nesting site in Oman (Ross and Barwani 1982) alongside large green turtle nesting sites in Yemen (Nasher and Al Jumaily 2015), Oman (Ross and Barwani 1982), and Saudi Arabia (Miller 1989; Pilcher 2000). These large green turtle nesting sites host turtles from a number of regional foraging grounds, and flipper tags from green turtles in Oman have been recovered in the United Arab Emirates, Yemen, Saudi Arabia, Ethiopia (now Eritrean waters), Somalia, and Pakistan (Salm et al. 1993). More recently, satellite tracking has augmented this regional connectivity data in the Arabian region. Two postnesting turtles have been tracked from Masirah Island in Oman (Rees et al. 2012) and both traveled to the southern Red Sea. Two postnesting and 4 rescued green turtles from Kuwait stayed in the northwest Arabian Gulf (Rees et al. 2018). Another 16 postnesting turtles tracked from Saudi Arabia (Maneja et al. 2018) stayed within the Arabian Gulf, and 8 rehabilitated turtles were tracked from the United Arab Emirates (Robinson et al. 2017). Six of these remained in the Arabian Gulf, one traveled toward Pakistan, and one crossed the Indian Ocean toward Thailand.

Here, using tracking data from postnesting green turtles, we explore the connectivity between Ras Al Hadd, possibly the most important green turtle nesting site in the entire northwest Indian Ocean (NWIO), and foraging areas in the Arabian region. The findings from this work broaden our general understanding of sea turtle life cycles and ecology within the Arabian region, and will help with the design of effective conservation strategies for green sea turtles nesting in Oman and at the foraging destinations. Our results may also be used to refine or affirm our delineation of regional management units (RMUs; Wallace et al. 2010) for the Northern Indian Ocean green turtle RMU and the current designation of the green turtle as Vulnerable (Mancini et al. 2019) on the International Union for the Conservation of Nature (IUCN) Red List.

METHODS

We deployed 3 ST-14 (Telonics, Inc, Mesa, AZ) transmitters in October 2007 and 6 SPOT-352B (Wildlife Computers, Redmond, WA) transmitters in November 2016 on postnesting green sea turtles at Ras Al Hadd in Oman. Turtles were restrained in open wooden boxes once nesting was completed, and transmitters were attached using a combination of epoxy resin and fiberglass following manufacturer recommendations. We sourced satellite signals from Argos (www.argos-system.com) and downloaded these via the Satellite Tracking and Analysis Tool (Coyne and Godley 2005). We filtered data for location fix qualities 3, 2, 1, A, and B, speeds of ≤ 5 km/hr between fixes and implausible data such as points over land (Witt et al. 2010). We included A and B data because of the low latitude that limits the number of locations via fewer Argos satellite passes (Pilcher et al. 2014). To standardize data across turtles, we chose the highest quality signal fix close to midday (Zbinden et al. 2008), and also the highest quality fix close to midnight because of the potential differences between day and night behavior (Rice and Balazs 2008). These selected signals were then used to calculate total distance and travel speeds.

Internesting behavior was inferred from short distance movements with random heading changes in the vicinity of the nesting site, up until the point of departure from the nesting site, when more purposeful and uniform direction movements were identified as migration behavior. These were assessed visually for each turtle. Similarly, we assumed foraging behavior had commenced when the migration ended and the turtles again displayed short-distance movements with random heading changes. We calculated minimum travel distances from straight-line movements between the location fixes, taking into account the spherical shape of the planet. Where tracks crossed landmasses, we extrapolated the shortest route around the landmass using straight sectors. We determined average travel speeds by dividing the total of all distances between fixes during the migration phase by the time interval between start and end of the migration, and determined proximity of location fixes to the coast using the Nearest Feature function in Cartographica 1.5.1. We used telemetry data to infer nesting emergences based on the number of 10- to 14-d internesting intervals (Miller 1997) during the period the turtles remained in the vicinity of the nesting site and prior to undertaking purposeful postnesting migrations to their foraging areas.

RESULTS

The 9 postnesting green sea turtles traveled from the Turtle Reserve in Ras Al Hadd, a regionally important nesting site in Oman, to 8 different destinations, 4 of which were within Oman's territorial waters. One turtle migrated to a feeding area in southwest Abu Dhabi (Turtle 1; Figs. 1 and 2; Table 1), and another to a foraging area south of Ras Al Khaimah (Turtle 2), both in the United Arab Emirates. One turtle migrated to a foraging ground in the Gulf of Kutch, India (Turtle 3), via Pakistan coastal waters, while another also migrated to the same location in India but via an oceanic route (Turtle 4); another migrated over 1200 km to a feeding area in the Dahlak archipelago, Eritrea, in the southern Red Sea (Turtle 5). The last 4 turtles remained in Oman, with one taking up residence at a foraging site approximately 85 km southwest of Ras Al Hadd and some 15 km south of the small fishing town of Al Ashkharah (Turtle 6), and a second off the northeast coast of Masirah Island (Turtle 7); a third traveled farther to an area between the fishing villages of Qaysad and Sawkirah, some 600 km in a straight line from Ras Al Hadd (Turtle 8), and the last turtle (Turtle 9) migrated to Al Hallaniyah Island, some 660 km from Ras Al Hadd (Figs. 1 and 2; Table 1). These diverse and disparate postnesting migrations, all of which were deemed to have reached foraging areas, demonstrate the multiple and wide-scale contributions of the large nesting aggregation at Ras Al Hadd to regional sea turtle stocks.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Table 1. Satellite transmitter deployment and tracking metadata, postnesting migration swim speeds, and distances between Ras Al Hadd (RAH) and the foraging areas (FA). CCL = curved carapace length; NA = not applicable; RAK = Ras Al Khaimah.

View Fullscreen

Three of the turtles departed immediately after satellite transmitter attachment (Turtles 1, 6, and 7) and 1 (Turtle 3) departed 3 d after deployment. Turtle 2 remained at Ras Al Hadd for 22 d and Turtle 9 remained 23 d prior to departure during which, based on 10- to 14-d renesting intervals, we estimate they deposited another 2 clutches of eggs each. Location signals for turtles that remained at Ras Al Hadd were within 15–20 km of the nesting site. Turtle 5 departed after 16 d and Turtle 8 departed after 13 d, and we estimate these turtles deposited 1 additional clutch of eggs each prior to departure. Turtle 4 remained at Ras Al Hadd for 64 d, during which she could have deposited up to 5 to 6 additional clutches of eggs. Because the nesting season had commenced prior to tag deployment, it is not possible to determine overall clutch frequency for the season for these turtles, but the clutch counts estimated here fall within known clutch frequencies for the nesting site (mean = 4.4, range = 3–6; Rees et al. 2008).

The migrations were broadly disparate in geographical destination, but shared several similarities. Turtles generally remained close to the coast (mean distance from each filtered location to shore during the migration phase = 9.7 km, range = 0.1–83.6), with the exception of Turtle 3, which crossed the deep northern Indian Ocean and then continued along a coastal route past Pakistan and south to India, and Turtle 4, which traversed the Indian Ocean on a more direct route to India (mean distance to shore during the migration phase = 106.5 km, range = 0.1–512.2 km). Travel speeds among all turtles averaged 24.2 km/d during the migrations (SD = 11.0 km/d, range = 12.0–47.9 km/d; Table 1). There were differences in travel speeds between oceanic and coastal movements, with Turtles 3 and 4 travelling at an average 37.8 km/d (range = 26.6–47.9 km/d) during the oceanic crossings, faster than turtles (including the coastal migration phase for Turtle 3) that traveled at an average 20.8 km/d (range = 12.0–34.9 km/d) during the coastal migrations. The shortest migration (to the nearby Omani coast) lasted 5 d and covered ∼ 110 km, and the longest migration took 63 d and covered ∼ 1650 km (to Eritrea). All turtles arrived at discrete foraging areas with postarrival signals lasting from 26 to 152 d (Table 1).

DISCUSSION

Nesting sites can host sea turtles from multiple distant foraging areas (Luschi et al. 1996; Bolker et al. 2007; Bradshaw et al. 2017) and our findings, combined with past tracking efforts (Rees et al. 2012) and historical flipper tag returns (Salm et al. 1993), along with recent tracking of 45 foraging turtles from 2 foraging areas in the United Arab Emirates (Pilcher et al. 2020) reveal that Ras Al Hadd is a regionally important nesting site for green turtles.

Our findings revealed that turtles from Oman migrate to the United Arab Emirates, Eritrea, India, and other parts of Oman, generally remaining in coastal waters (< 10 km from shore) unless transiting to other landmasses. Flipper tag returns from Oman have also been recovered in Somalia (Schleyer and Baldwin 1999) and from Sudan, Saudi Arabia, and Eritrea in the Red Sea (Salm et al. 1993). Tags have also been recovered inside the Arabian Gulf in Saudi Arabia, Qatar, Bahrain, and the United Arab Emirates (Salm and Salm 1991; Salm et al. 1993; Ferreira 2002), and eastward in the northern Indian Ocean in Pakistan and the Maldives (Salm et al. 1993). The wide spatial dispersal of turtles from Ras Al Hadd underscores the site's importance of the site for sea turtles at the NWIO scale, hosting turtles from multiple foraging areas spread several thousand kilometers to the east and west across the NWIO. Of note, no tag returns have been reported from locations south of the Horn of Africa in Somalia along the mainland African coast, or from island groups such as the Seychelles or French Indian Ocean Territories, indicating green turtles from this nesting site remain in the NWIO.

Turtles 1 and 2 migrated to the United Arab Emirates and settled at known green sea turtle foraging grounds off the Abu Dhabi coast and Ras Al Khaimah (Hasbún et al. 2000; Miller et al. 2004; Environment Agency Abu Dhabi 2007). The connectivity between these foraging areas and Ras Al Hadd has also been demonstrated through a recent study that tracked turtles from foraging areas in the United Arab Emirates to identify their nesting destinations (Pilcher et al. 2020). The foraging area in Abu Dhabi along the Khor Al Bazim lies within the Marawah Biosphere Reserve, where fishery activities are controlled. However, the turtle that migrated to Ras Al Khaimah migrated to a zone where coastal beach seine fishing that interacts with sea turtles is common (Hasbún et al. 2000). Bycatch is also common in coastal waters of Iran (Kiaalvandi et al. 2012), through which these turtles move. The use of wire mesh fish traps, locally known as “ghargoors,” is also prevalent in the region and linked to bycatch of sea turtles (Yaghmour et al. 2018). The linkage of these sites to Oman demonstrates that this threat is of concern to the wider regional green turtle population.

Turtles 3 and 4 traveled to the southwest Gulf of Kutch (Kachchh) in India, where green turtles are also known to forage (Sunderraj et al. 2001), and where fisheries bycatch is also believed to be a major threat (Rajagopalan et al. 1996). Green turtles have been caught as bycatch in nets near Baydar Island (Bhaskar 1979), adjacent to the foraging area used by the Turtles 3 and 4, and multiple green turtle strandings have also been reported for this area (Sunderraj et al. 2006; Joshi et al. 2018). Fisheries in nearshore waters along the coast of Pakistan, through which Turtle 3 migrated, have also recorded turtle bycatch (Moazzam and Nawaz 2014; Shahid et al. 2015), highlighting susceptibility to fisheries both at the foraging area and along the migration route.

The movement of Turtle 5 to Eritrea mirrors a previous satellite track from Masirah Island (another green turtle nesting site in Oman) to a location just south of the one recorded in the present study (Rees et al. 2012), along with that of another turtle that went to Saudi Arabia in the southern Red Sea. The Dahlak archipelago, where Turtle 5 and the one tracked by Rees et al. (2012) took up residence, is a known green turtle foraging area (Howe et al. 2004), and satellite tracks from Egypt to this site (Attum et al. 2014) reveal that it hosts turtles from multiple nesting sites. This foraging area is subject to bycatch in local fisheries (Howe et al. 2004; Mebrahtu 2015) and illegal hunting of adult turtles (Frazier 1980). The migration routes and travel speeds of the 2 turtles tracked by Rees et al. (2012) and the 1 from this study were broadly coincident, with all 3 turtles staying close to the Arabian peninsula throughout the migration until entering the Red Sea through the Bab el Mandeb, travelling at ∼ 0.8–1.5 km/hr. The turtles did not stray into deeper waters, but rather traveled close the Omani and Yemeni coasts, where turtle hunting (Siddeek and Baldwin 1996; Hariri et al. 2000; Stanton 2008) and bycatch in fisheries (Hare 1991) have also been reported.

The last 4 turtles remained in Omani waters, with one settling approximately halfway between Ras Al Hadd and Masirah Island, a second settling off Masirah Island, a third off the mainland coast between Qaysad and Sawkirah fishing villages, and the last just south of Al Hallaniyah Island. These areas in Oman were identified as green sea turtle foraging habitats via coastal surveys (Salm et al. 1993), with the greatest accumulation or foraging green turtles (∼ 1000–3000) occurring between Masirah and the mainland (Ross 1985). One of the green turtles in this study did reach Masirah, but rather than take up residence in the shallow seagrass habitat between the island and the mainland, the turtle spent 73 d inhabiting an area within 10 km of the northeastern corner of the island, fronting the Indian Ocean. The 3 remaining turtles were tracked for 26–84 d within a 10-km radius of the mainland or island sites. Bycatch is of concern in some parts of Oman (Hare 1991; Al-Mamary et al. 2015; Brownell et al. 2019), as is illegal hunting (Sideek and Baldwin 1996), but encouragingly, local authorities in Oman have made substantial progress to address conservation concerns through several key initiatives: 1) benthic trawling was banned by Royal Decree 20/2009, 2) regulations have been introduced to control the use of drift gillnets (Ministry of Agriculture and Fisheries [MAF] 2019a), and 3) natural marine resources were further protected via Royal Decree 20/2019 on Promulgating the Marine Living Resources Law (MAF 2019b). In addition, the MAF plans to introduce a new law to ban the catch of cetaceans, which is likely to also be beneficial to turtles.

We suggest that the widespread fisheries bycatch threat could be addressed through regional bodies such as the Regional Organisation for the Protection of the Marine Environment and the Regional Organization for the Conservation of the Environment of the Red Sea and the Gulf of Aden, the Indian Ocean Tuna Commission, the Convention on Migratory Species, and the Indian Ocean–Southeast Asia Sea Turtle Memorandum of Understanding (https://www.cms.int/en/legalinstrument/iosea-marine-turtles), along with bilateral discussions among Oman and those countries to which Omani nesting turtles migrate. Alongside this, engagement with local fishing communities that have a direct impact on sea turtles, particularly in the vicinity of important nesting areas such as Ras Al Hadd, is also required.

These findings enhance our knowledge of connectivity between the Ras Al Hadd nesting site in Oman and multiple foraging areas spanning from the southern Red Sea across to India, and of the regional ecology of green turtles in the NWIO. Our data may be used in revalidation of the extent of the NWIO Regional Management Unit for green turtles (Wallace et al. 2010) and contribute to IUCN regional species assessments. Our findings may also be used to support the development of fishery management guidelines to minimize capture of green sea turtles in trawl and gillnet fisheries in all those countries through which these sea turtles moved. Finally, the migrations documented in this study reveal a need for a greater understanding of fisheries bycatch in the NWIO. We suggest that these data would be enhanced through regional fishery observer programs, citizen science programs, and via direct engagement with local fishing communities.