The Sultanate of Oman is located on the eastern tip of the Arabian Peninsula (Fig. 1; Table 2 for mapping information) with coastal borders to the Persian/Arabian Gulf, hereafter referred to as the Persian Gulf, the Gulf of Oman, and the Arabian Sea to the northwest Indian Ocean (NWIO). Although the Persian Gulf and Gulf of Oman are relatively small compared to the expanse of the Indian Ocean, they include vital nesting and foraging grounds for loggerhead (Caretta caretta) and green sea turtles (Chelonia mydas), with occasional hawksbill (Eretmochelys imbricata) and olive ridley (Lepidochelys olivacea) sea turtles (see, for example, Pilcher et al. 2014; Tollab et al. 2015; Mobaraki et al. 2020; Willson et al. 2020b).

View Figure

• Download as Powerpoint
• Download Figure

Oman has the largest number of nesting green sea turtles of any single Indian Ocean nation (Salm 1991), with the majority of nesting occurring within the Ras al Hadd Turtle Reserve, with approximately 6000 nesting females recorded every year (Ross and Barwani 1982; Ross 1985; Salm and Salm 1991). Nesting population estimates in Ras al Hadd Reserve have increased to 10,000 annually according to (Mobaraki et al. 2020) with peaks of 361 nesting turtles per km per night on Ras Al Jinz south beach (Mendonca and Abi-Aoun 2009). These estimations need to be confirmed with recent monitoring surveys. Green turtles also nest in lower numbers on the southern shores of Oman, where nesting beaches host both green and loggerhead turtles (Salm 1991; Mendonca et al. 2001), on some beaches of Muscat (Salm and Salm 1991), at the Ad Dimaniyyat Islands Nature Reserve (DINR) (Ross 1980; Ross and Barwani 1982; Salm 1991; Salm and Salm 1991), and on Masirah Island (Salm et al. 1993). There are reports of green turtles nesting sporadically in Musandam (Salm and Jensen 1991), noting that nesting opportunities on the peninsula are limited due to the scarcity of sandy beaches.

Masirah Islands provides one of the most important nesting aggregations for the loggerhead turtles in the world, once hosting up to 30,000 nesting females per year (Ross 1979; Ross and Barwani 1982; Salm 1991; Salm and Salm 1991; Baldwin 1992). In recent years, a 79% decline in the nesting activity on Masirah was noted over a 20-yr period from data collected between 1985 and 1996 (historical data) and 2008–2016 (recent data, Willson et al. 2020b). Loggerheads also nest on the southern beaches of Oman, with majority on Al Hallaniyat Islands and beaches along the mainland (Salm 1991). The DINR is considered one of the most important nesting aggregations for hawksbill turtles in the region with 109–350 nests per year along 2 km of beaches (Ross 1980; Ross and Barwani 1982; Salm 1991; Salm and Salm 1991). Minor nesting sites for hawksbills have been reported along some beaches of Muscat (Salm and Slam 1991), Masirah Island (Salm et al. 1993), and Barr Al Hikman (Whittington-Jones et al. 2023). Olive ridley turtles are only known to nest on Masirah Island along 30 km of beach with a published record of around 150 nesting females per year (Ross and Barwani 1982). Rees and Baker (2006) confirm olive ridleys continue to nest on Masirah, though no population estimates were provided due to low direct interception (n = 6) of females.

Flipper tagging of nesting females, and their subsequent resightings on nesting beaches, was conducted on Masirah Island and at Ras Al Hadd Turtle Reserve by the Environment Authority up until the mid-2000s. This technique has confirmed migrations from Masirah Island and Ras al Hadd to the United Arab Emirates, Yemen, Saudi Arabia, Ethiopia, Somalia, and Pakistan (Salm 1993). Tag loss or corrosion beyond readability of tags can disrupt long-term studies, impacting accuracy of population estimates (Reisser et al 2008; Schofield et al. 2008).

More recently satellite tracking efforts on nesting females were applied to all 4 nesting species across Oman and showed the importance of Oman’s neritic waters as foraging and internesting habitats. Green turtles were satellite tagged near Masirah Island (Rees et al. 2012a, 2018), Ras Al Hadd (Pilcher et al. 2019), and the Al Qurm Protected Area, Sharjah UAE (Pilcher et al. 2021b) the Persian coast of the Musandam Peninsula. Whittington-Jones et al. (2023) have shown both juvenile and adult green turtles with long coastal migration routes, ultimately passing through the Strait of Hormuz past the Musandam Peninsula to Masirah Island. They also have the capacity to conduct long-distance movement into the Red Sea and the Gulf of Kutch in India (Rees et al. 2012a, 2018). Data from 40 satellite tags deployed on nesting loggerhead females on Masirah Island (2006 and 2010–2012) showed that most turtles remained within a 100-km corridor north of the island for up to 14 wks posttagging, and later migrated south across the continental shelf, reaching offshore waters of Oman and Yemen (Tiwari et al. 2015). This aligns with evidence of behavioral polymorphism in this population, where satellite tracking of 10 loggerheads revealed distinct migratory strategies, with some individuals remaining near nesting sites while others undertook extensive oceanic loops before returning to nest (Rees et al. 2010). Similarly, 25 hawksbill turtles tagged between 2007 and 2012 from Masirah Island and DINR migrated steadily south toward Quwayrah, Oman, staying near the mainland and continental shelf (Pilcher et al. 2014). Nine olive ridleys tagged in 2008 on Masirah Island preferred coastal neritic waters (Rees et al. 2012a, 2012b).

Extensive foraging grounds for loggerhead and green sea turtles extend across Oman, with aerial survey records documenting > 100 sea turtles foraging per kilometer (Salm et al. 1993). Antonopoulou and Pilcher (2018) identified, with satellite telemetry, hawksbill foraging grounds in the DINR and in Sawqarah Bay, with additional foraging identified between Salalah and Barr Al Hikman (Whittington-Jones et al. 2023). Juvenile green sea turtles exhibit a propensity for foraging within neritic habitats of shallow coastal zones, occupying distinct developmental habitats as they progress toward adulthood (Lenz et al. 2017; Pilcher et al. 2021a). The Musandam Peninsula is a mountainous rocky headland with many fjords. The Musandam National Park Nature Reserve was declared in 2022 by Royal Decree No. 54/2022. The reserve spans 1150 km² covering most of the peninsula's outer reaches. The Khasab region in the north of Musandam has been reported rich in Halodule unniervis and Halophila ovalis seagrass species (Salm and Jensen 1991), both described as being eaten by grazers such as green turtles, from juvenile to adult individuals (Aragones and Marsh 1999; Kuiper-Linley et al. 2007; Ballorain et al. 2010).

While satellite telemetry provides critical insights into spatial movements, habitat use, and migration patterns, challenges in deploying and processing satellite tag data can limit its feasibility for early-phase population assessments. Substantial costs are associated with refining and miniaturizing satellite tag technology, and despite advances in sensor design through human use, sports, and health applications, these developments have not fully translated to sea turtle satellite tags, particularly concerning depth capabilities and battery recharge efficiency (Hays and Hawkes 2018). Furthermore, data transmission is restricted by the narrow bandwidth available for satellite communication. Additionally, tags have finite lifespans and may become dislodged over time due to various factors such as material fatigue, growth of the animal, or environmental exposure. Ethical considerations must also be addressed, as the process of capturing, attaching, and retaining these tags poses potential risks, including injury, increased predation vulnerability, and physiological stress, all of which may impact the individual’s overall fitness (Hamelin and James 2018). While satellite telemetry excels at answering spatial ecology questions, its feasibility can be constrained by cost, technical limitations, and ethical concerns. Recognizing these trade-offs is essential when determining the most appropriate method for conservation and management goals. Given these challenges, complementary approaches such as Photo-ID combined with GIS can serve as valuable tools for long-term, cost-effective conservation monitoring, particularly when population-level assessments are the primary focus.

Photo-ID is an alternative noninvasive capture-mark-recapture (CMR) technique, which uses the natural variation in scale pattern on the sea turtle facial profile (Reisser et al. 2008; Schofield et al. 2008; Jean et al. 2010), top of the head (Lloyd et al. 2012), and flippers (Mills et al. 2023) and their unique characteristics to identify hard-shelled sea turtles. This technique has been proven stable over time in sea turtles and is considered a valuable tool for short- and long-term population monitoring in green turtles (Carpentier et al. 2016), loggerheads (Papafitsoros et al. 2023), hawksbills (Hudgins et al. 2023), and olive ridley turtles (Stelfox et al. 2020). Photo-ID provides significant benefits and increased population-monitoring applications because of the noninvasive nature (De Urioste et al. 2016; Hancock et al. 2020) and species adaptability of Photo-ID capture mark recapture algorithms (Dunbar et al. 2014). Applications of Photo-ID include behavioral studies (Chassagneux et al. 2013; Schofield et al. 2017), monitoring foraging individuals (Schofield et al. 2008; Lloyd et al. 2012; Dunbar et al. 2014; Neves-Ferreira et al. 2023), analysis of population distribution and trends (Williams et al. 2017; Hudgins et al. 2023), examining breeding rates, periodicity, and mating patterns (Hays et al. 2010; Schofield et al. 2020; Papafitsoros et al. 2022; Witzmann et al. 2023), growth rates (Long and Azmi 2017; Araujo et al. 2019), health monitoring, (Bennett et al. 1999; Schofield et al. 2013; Papafitsoros et al. 2021; Hancock et al. 2023), and measuring the impact of ecotourism on individual turtles (Papafitsoros 2015; Hayes et al. 2017; Köhnk and Stelfox 2022).

The incorporation of Photo-ID as a CMR tool in monitoring sea turtle encounters serves to address the significant gap in knowledge of in-water sea turtle activity and spatial distribution, especially for understudied life stages such as juveniles and adult male turtles. The broad scope of applications for Photo-ID can improve our understanding of sea turtle distribution, habitat usage, and population dynamics.

In this study sea turtle Photo-ID encounters from the Musandam Peninsula and Ad Dimaniyyat Islands from 2019 to 2023 are presented. Photo-ID is used to determine the population dynamics observed from sea turtles encountered during in-water surveys and by-catch incidents. Musandam is located within an area of interest with sea grass beds and intermittent nesting, making encountering migrating adults or foraging juveniles a strong probability. Studying in-water sea turtle encounters serves to provide an important view into sea turtle marine habitat usage.

METHODS

Study Site. —

The operational base for this study is Six Senses Zighy Bay Resort 25°42′25″N, 56°16′25″E, where a sea turtle biologist from the Olive Ridley Project (ORP) was stationed. The site Zighy Bay is a directly accessible coral reef located within 200 m of the coastline. A standardized survey protocol was followed, consisting of in-water surveys conducted 5 d per week. Transects were set at a minimum of 500 m and lasted approximately 1 hr; transect distances and times were reduced if necessary to accommodate guests and dependent on water conditions. All other dive and snorkel sites are as distinguished by the operator (Haffah, Lima Rock North, Lima Rock South, Octopus Rock, Ras Hamra, Ras Lima, Ras Marovi, Sanat Bay, Sanat Corner, Stingray Bay, The Cave, Wonderwall, and Ad Dimaniyyat Islands) and were accessed by boat. These are opportunistic surveys taken with dive and snorkel operators with capacity for 8 though often groups are smaller; dive time ranged 45–60 mins. Photos were primarily taken by the sea turtle biologist using a digital camera in underwater housing, though they could be provided by anyone participating in the dive or snorkel with a camera. The biologist was primarily responsible for capturing images; guests with an interest in the topic were also encouraged to contribute to data collection, as part of a broader citizen science project. All survey participants were introduced to and encouraged to closely adhere to the ORP Code of Conduct (Olive Ridley Project 2018), which outlines behavioral guidelines for safe wild turtle encounters limiting disturbance to the animals. Survey efforts were discontinued in March 2020 due to the impact of COVID-19 and were resumed in 2022. To provide a comparable measure of sea turtle encounter numbers capture per unit effort (CPUE) was calculated for each quarter, using survey hours and total of identified encounters.

Photo-ID Protocol. —

For Photo-ID it is important to note that the scales of each side of a turtle’s head are distinct from each other; therefore to ensure identifications of sea turtles are not replicated, they are not assigned a full National ID (NID) code until at least the right side of the sea turtle’s head (and preferably both sides) was captured during an encounter. Any encounters of individuals with only the left profile photo are kept as Candidate ID (CID) profiles for future comparison purposes to avoid assigning 2 NID codes to 1 individual, in case the left and right profiles are captured independently. This strict processing protocol served to ensure Photo-ID results was robust and accurate. Nomenclature of NID codes G denotes species green, in the case of hawksbill this is H, M denotes the country of Oman, and 065 reflects when the sea turtle was first documented. CID codes use lowercase species and country rules.

Photo-ID was applied to all sea turtles encountered during the study period; the additional information collected during an in-water encounter provides context for the encounter (Table 1). The study is predominantly focused on in-water sightings, though there are some instances of bycatch documented. In the instance of a bycatch sea turtle, the curved carapace length (CCL) from nuchal scute to supracaudal notch and curved carapace width (CCW) at the widest part of the carapace were directly measured with a flexible measuring tape.

Table 1. Additional data attributes collected during a sea turtle encounter.

View Fullscreen

Life Stage and Sex Assignment. —

The size of a sea turtle's carapace can be used to assign a life stage and the presence or absence of other physical attributes; namely, the tail length can be used to assign sex. In the Arabian Gulf, the minimum CCLs of nesting females was 65 cm for hawksbills (Pilcher et al. 2014) and 78.7 cm for green turtles (Hirth 1980). CCL can be converted into SCL for green and hawksbill sea turtles (Bjorndal et al. 2016, 2017; Lenz et al. 2017). Therefore, hawksbills of a minimum size of 60 cm SCL, and greens of 70 cm SCL, with the presence or absence of an elongated tail, were assigned as adults of the respective sex. Size estimates of SCL were collected by sea turtle biologists, following established methodology (Hudgins et al. 2023). The juvenile and adult categories with size definitions are more accessible to citizen scientists, though size estimates are documented if provided. Sea turtle activity categories have been defined and were implemented as standard practice for all sea turtle biologists’ encounters. This was further expanded to incorporate substrate categories of the encounter from October 2023 (Table 1).

Image Review. —

To ensure image quality is maintained, each image was reviewed and scored for 0–5 points for each of the following categories: focus, lighting, and angle according to (Hudgins et al 2023) methodology. Only images scored higher than 10 were considered suitable for Photo-ID. All identifications were initially completed manually by a sea turtle biologist. The Internet of Turtles (IoT, https://iot.wildbook.org/) Hotspotter algorithm (Dunbar et al. 2021) was used as secondary confirmation of manual identifications. The algorithm detects an area of interest from the image, upon which it generates potential matches, and the match is verified by trained users as a visual inspection. In cases of difficult image quality a second manual identification was performed by one of the coauthors. In the instance of an unmatched encounter, potential new sea turtles are reviewed by a senior ORP team member, who confirms the ID before assigning a NID or CID code.

The GIS datasets (Table 2) used in this study were sourced from multiple authoritative repositories. Country borders of Oman and Musandam were obtained from the Global Administrative Areas database (GADM), while sites of interest names (Fig. 1) were derived from the DIVA-GIS gazetteer. Basemap and hillshade data were provided by ArcGIS Pro. Protected areas, including nature reserves and marine protected areas (MPAs), were sourced from the World Database on Protected Areas (WDPA) maintained by UNEP-WCMC and IUCN (2024). Waterway data were extracted from the Humanitarian OpenStreetMap Team Open Database License (ODbL). Dive and snorkel site names (Fig. 2) were recorded using ORP and Scuba Schools International (SSI) coordinates, while turtle population data were collected by ORP.

View Figure

• Download as Powerpoint
• Download Figure

Table 2. GIS data description and sources.

View Fullscreen

RESULTS

Since 2019, ORP in-water surveys recorded a total of 828 sea turtle encounters, of which 147 individual green turtles and 9 individual hawksbill turtles were assigned a NID code (Fig. 2; Table 2 for mapping information). An additional 79 green turtles and 1 hawksbill turtle were assigned a CID code profile. In 5 instances Photo-ID did not yield either a CID or NID designation due to insufficient image quality for Photo ID purposes, specifically the extended distance of the turtle focus of the camera in poor visibility; in a further 23 sea turtle encounters, no photos were obtained. Surveys were recorded across 22 different sites, with sea turtles encountered at 17 of them (Fig. 2).

Survey effort has varied across the years. Initial exploratory surveys during 2019 covered 17 dive and snorkel sites with 50 sea turtle encounters (Fig. 3), yielding identification of 14 green turtles, consisting of 4 adults and 10 juveniles, and 7 hawksbill encounters, from 4 individuals consisting of 2 adults and 2 juveniles. Thirty-two sea turtle encounters were recorded with the identification of an additional 2 juvenile green sea turtles. The year 2020 was the peak year for identifying hawksbill turtles with 12 encounters from 4 NID individuals: 1 adult male, 1 adult of unknown sex (tail not visible), and 2 juveniles. Only 1 hawksbill was encountered in Zighy Bay; the remaining turtles were mostly located in Sanat Bay (n = 8), Sanat Corner (n = 2), and The Cave (n = 1) (Fig. 2, top right). With the exception of Zighy Bay, these sites were not surveyed again until 2023.

View Figure

• Download as Powerpoint
• Download Figure

Regular surveys resumed in January 2022, following the reopening of the border between the United Arab Emirates and the Musandam Peninsula of Oman, with survey efforts predominantly focused in Zighy Bay. There were 2 excursions to Wonderwall and 1 to Ras Lima in the first quarter of 2022 with 5 green turtle encounters, including 2 adult males, 1 adult female, and 2 juveniles. Ultimately 1 new individual green turtle was encountered at Wonderwall. The remaining 302 green turtle encounters were recorded at Six Senses Zighy Bay, with 65 new juvenile and 1 adult green sea turtles identified. There were 3 juvenile hawksbill encounters, 1 new adult hawksbill individual identified, and 1 additional hawksbill candidate identified.

During 2023, a notable increase was seen in the number of sites surveyed with a total of 15 locations. Sixty-four new juvenile green turtles identified, 8 of which were found in the first surveys conducted at the Ad Dimaniyyat Islands (Fig. 2, bottom right).

Survey hours by quarter have been used to calculate the Capture Per Unit Effort; the figures reported reflect the CPUE rates of NID turtles only. Increased survey efforts from 32 hrs per quarter to 60–70 hrs per quarter have resulted in an increase in the number of sea turtles encountered and the number of NID sea turtles (Fig. 4).

View Figure

• Download as Powerpoint
• Download Figure

Of the 156 NID sea turtles, 1 green turtle was only encountered washed up and deceased, 80 individuals were only encountered once, and 56 individuals were resighted on 1 to 9 occasions (Table 3). The remaining 19 individuals account for 46% of the total encounters during 2019–2023; these individuals were resighted on 10 or more occasions (Fig. 5). The most sighted NID-coded individual (GM065) was observed on 43 separate encounters over 594 d. The greatest resighting span was for GM004, with a duration of 1676 d (13 February 2019 to 16 September 2023) and 14 total sightings. The time between the 2 most recent sightings of the 19 most-encountered turtles was a minimum of 3 d, a maximum of 155 d, and a median of 21 d. All sea turtles resighted on 10 occasions or more were resighted only within Zighy Bay. They have not been encountered at other studied sites.

View Figure

• Download as Powerpoint
• Download Figure

Table 3. Resighting counts of lower 56% of encountered NID turtles.

View Fullscreen

Only 3 sea turtles were sighted at more than 1 site. HM005 was sighted at The Cave (20 January 2020), then 600 m southwest at Sanat Corner (1 February 2020), and inside Sanat Bay 800 m southeast from The Cave (4 February 2020 and 12 March 2020). HM011 was first sighted in Zighy Bay (19 November 2022), then later ∼ 1.2 km north at Stingray Bay (19 April 2023). GM010 was sighted on both reef faces from the rocky peninsula outcrop at Wonderwall on the south face (19 February 2019-02) and Ras Hamra at the north face (20 February 2019) ∼ 1 km around the corner. Site fidelity is not a measure that can be definitively inferred at this time, due to the survey's close proximity of sites and a survey bias toward Zighy Bay.

Turtle size estimates were made for nearly all encounters (820 of 828 turtles; Fig. 6). Eighty-nine percent of sea turtles encountered were estimated to be 30–50 cm SCL. The smallest green turtle encountered (GM191 on 15 September 2023) was estimated to be 25 cm SCL, and the largest encountered turtle (GM007 on 19 February 2019) was an estimated 90 cm SCL. Two turtles have been repeatedly sighted with recorded size estimates and were subsequently encountered as bycatch. First, GM056 was encountered in-water a total of 21 times, first encountered 24 March 2022, with an estimated straight carapace length of 40 cm; this estimate increased to 45 cm 5 July 2023. The same turtle was encountered as accidental bycatch on 6 December 2023, and sea turtle biometrics were directly obtained. At the time GM056 had a CCL of 40.4 cm. Second, GM004 was encountered in-water 17 times over the study period. The first encounter with a size estimate was on 13 February 2019 at a SCL of 40 cm, followed by 5 January 2020 at 50 cm SCL, and 19 March 2022 at 60 cm SCL. On 2 June 2023, GM004 was recovered from accidental bycatch and measured at 60 cm CCL prior to release. Of the 10 NID adult sea turtles sighted, only 3 have been resighted (GM018 with 1 resighting 100 d later, HM008 with 1 resighting a day after the initial sighting, and HM005 with 3 resightings within 52 d but not seen since 12 March 2020). An additional 10 CID green turtle adults and 2 encounters were without a photo and 1 encounter with a photo of insufficient quality.

View Figure

• Download as Powerpoint
• Download Figure

Furthermore, 270 encounters were accompanied with sea turtle activity information: 176 were encounters with swimming sea turtles, 47 resting sea turtles, and 42 foraging sea turtles observed over different substrate types (including mixed hard and soft coral [n = 34], rock [n = 7], and open water [n = 1]. In total, 173 encounters were documented in a mixed hard and soft coral environment, 11 on rocky substrate, 8 encounters in open water, and 3 over sand.

DISCUSSION

This is the first study of the in-water population of sea turtles on the Musandam Peninsula using Photo-ID. This study highlights the simplicity and efficiency of employing Photo-ID CMR and highlights its applicability and potential value for sea turtle research in Oman. Over 4 yrs, the method successfully documented 147 individual green and 9 hawksbill turtles with NID status. By restricting Photo-ID status of a sea turtle with only the left image available as a CID, it serves as an important quality control and prevents duplicate identifications. This protocol step is necessary, as the left and right profiles of a sea turtle’s face are different (Carpentier et al. 2016) and assigning NID status to either the left or right side might otherwise lead to an overestimation of population size.

This CMR method has consistently yielded successful Photo-ID results, from many individuals which may have been observed only once to date. Photo-ID of individual sea turtles has been demonstrated as stable for green turtles for least 11 yrs (Carpentier et al. 2016). Extended longevity estimates of turtles via successful Photo-ID confirmations is plausible as existing datasets mature. The integration of computer assisted identification creates the nonsubjective process required for efficient accurate Photo-ID (Jean et al. 2010). The evolution of Photo-ID techniques has facilitated efficient analysis of large databases, reducing processing time and enhancing accuracy (Dunbar et al. 2021; Buteler et al. 2022).

The connectivity of sea turtle populations from within the Persian Gulf out into the Gulf of Oman and the Arabian Sea (Pilcher et al. 2021a) means that the Strait of Hormuz and the Musandam Peninsula are specific places of interest to potentially encounter migrating female sea turtles. Despite this, 94% of NID turtles were juveniles, with only 6 adult green turtles and 4 hawksbill adult sea turtles. With nesting females passing the peninsula (Pilcher et al. 2021b; Whittington-Jones et al. 2023), we expected to encounter more adults or more seasonality in sea turtle sightings due to their migratory behavior. It is assumed that adults migrating between their primary foraging grounds and nesting location tend to swim through an area without significant foraging activity (Pilcher et al. 2021b). This may explain why there are only 14 encounters from 10 NID adults since 2019; the adult sea turtles are using the region for a much more transient purpose. Continuing to survey existing sites and expansion into a wider geographic region will increase the probability of resightings and the addition of new sea turtle sightings, including adults.

Over the long term, the potential for detecting seasonal variation may be possible, once adult sea turtles are consistently located in the region. However, a large population of previously undocumented juvenile sea turtles has been identified. Posthatchling sea turtles have an initial pelagic open ocean phase before they recruit to neritic coastal habitats. Turner Tomaszewicz et al. (2022) document neritic recruitment CCL sizes from approximately 35 cm in hawksbill turtles, with juveniles recruiting to neritic habitats at around 30 cm CCL and remaining there until reaching 50 cm CCL (Lenz et al. 2017). This is very much in line with the SCL carapace estimates for 89% of the observed juvenile population in this study, which is indicative of the presence of suitable foraging and resting habitats capable of supporting them in the Musandam region. Nearly half (46%) of the encounters recorded were of only 12% of the turtles with a NID (n = 19 individuals), indicating high side fidelity and a small regular resident population. While 71.9% turtles (n = 112) were resighted only once or twice, representing 18.1% of all recorded encounters, the majority of the identified turtles seem to have a more transient presence in the area. This is indicative of either a population turnover, meaning that the sea turtles we are sighting do not stay in the area long term or that the reef habitat is not at the carrying capacity for individual sea turtles. Although there are some instances of sea turtles sighted at different locations, site fidelity is not a measure that can be definitively inferred at this time, due to the survey bias in Zighy Bay. As the most easily accessible and the most heavily surveyed site, Zighy Bay has recorded 116 NID sea turtles; increasing survey efforts to cover all sites more evenly in the future will help to alleviate the current survey bias.

Incidences of bycatch have provided opportunistic direct measurements for CCL and curved carapace width (CCW) measurements. This provides some direct insight into the accuracy of the in-water SCL calibration and methodology in deriving in-water SCL estimates and creating a basis for further informative calibration training The SCL estimates collected for GM056 during in-water surveys are within ± 5 cm of the actual CCL obtained upon bycatch encounter. It would not be possible to obtain this detail in the data without the collaboration of fishing communities.

When the Ad Dimaniyyat Islands were surveyed, all sea turtles encountered were new and distinct from our Musandam Photo-ID population dataset. By continuing Photo-ID CMR efforts in Musandam we may encounter some of these individuals equally; they may be more inclined to migrate south and west or simply into the wider NWIO. This provides further support for the theory of site fidelity suggesting that turtles in these regions may exhibit localized residency patterns rather than frequent long-distance movements between sites (Broderick et al. 2007; Rees et al. 2010). However, given the relatively small sample size and the considerable distances involved, further research is necessary to determine the extent of connectivity between these populations.

To effectively protect sea turtles throughout all their life stages requires protection of critical nesting grounds and foraging habitats (Ferreira et al. 2003). Studying and protecting sea turtles while they are at sea is more complex than at their natal beaches due to limited access to a spatially vast area. While at sea, fundamentally sea turtles require stable access to food sources and safe resting habitats such as those identified (Salm and Jenson 1991). In the north of Musandam there are 2 seagrass species, H. unniervis and H. ovalis, that are globally recognized as foraging species for green sea turtles (Aragones and Marsh 1999; Kuiper-Linley et al. 2007; Ballorain et al. 2010). The incorporation of substrate metadata of sea turtle encounters documents the habitat type the sea turtle is encountered in at each dive and snorkel site. In the absence of detailed benthic surveys this method builds a picture of the different substrate types found at a given dive and snorkel site. During this study none of the individuals from the predominantly juvenile green turtle population were observed in a seagrass habitat, which, on a wider geographical scale, is considered a primary food source for this species (Bjorndal 1997). Esteban et al. (2020) classify nutritional green turtle food sources as seagrass, macroalgae, animal matter, and terrestrial plants (e.g. fruits and leaves from mangroves). The exact benthic species composition and algae cover present on the reef and rock habitats in Zighy Bay remain to be investigated. Nevertheless, this benthic habitat might provide sufficient nutrition for green turtles with a more omnivore and/or less seagrass specialized diet. This may be due to site selection bias; surveys outside of Zighy Bay that are opportunistic trips with the dive operator providing dive and snorkel trips to guests may also favor coral substrates over seagrass.

The primary obstacle in tagging sea turtles lies in ensuring the tags ability to remain durable and readable throughout the turtles’ lengthy lifespan, which can be decades and entail considerable growth. Losing tags can skew population size estimates, survival rates, and other demographic analyses. Notably, corrosion in the metal flipper tags, especially within the locking mechanism, seems to significantly contribute to tag failure (Bellini et al. 2001). Integrating flipper tag and Photo-ID records reduces the impacts of tag loss on population estimates (Reisser et al 2008). Satellite tagging proves to be a costly method, often with instruments having a short lifespan and potential adverse effects on animal welfare (Hays and Hawkes 2018). This technique may be considered invasive and may affect the natural swimming behavior of the animals involved (Hamelin and James 2018). Photo-ID has broad applications to sea turtle encounters in-water, for documenting multiple bycatch events of the same turtle, and even for repeated nesting events, the last facilitated with the use of red-light photography. This importantly aims to balance the current biases that generally focus on the easy-to-access nesting populations, and provides important insight into sea turtle distribution and population dynamics in their primary habitat.

Photo-ID is a method suitable for large-scale population estimates (Hudgins et al. 2023). Longevity of scale patterns has been proven in green sea turtles (Carpentier et al. 2016), loggerheads (Papafitsoros et al. 2023), hawksbills (Hudgins et al. 2023), and olive ridley turtles (Stelfox et al. 2020). Photo-ID has the capacity to significantly increase the geographical range that we are able to study. It is important to consider the sea as interconnected cross-border marine habitats (Ferreira et al. 2006), and collaboration and data sharing will be key to the success of monitoring in-water sea turtle populations. This can be achieved through the creation of collaborative networks of dive operators throughout Oman and the Fujairah region of UAE with more than 100 named dive sites.

Photo-ID as a monitoring method that is widely accessible to an environmentally conscientious tourist. Engaging sustainable tourism and in particular the dive community, which are often well equipped including personal underwater cameras, allows for the expansion of data collection, with the potential to significantly increase the geographical range monitored for sea turtle activity in the future. Key foraging and nesting habitats have been identified through citizen scientist survey contributions in the Maldives (Hudgins et al. 2017). Studies in Oman have traditionally relied on conventional tagging methods, including flipper and satellite tagging. Moreover, these methods require significant effort and human resources, such as trained marine biologists, and specialized equipment. In contrast, Photo-ID requires comparatively minimal effort. Anyone passionate about marine life and sea turtles can participate in the process. The only equipment necessary is an underwater camera, and the associated costs are negligible.

A comprehensive conservation strategy must integrate long-term monitoring of understudied life stages and in-water activity, which remain poorly understood in Oman. This study addresses these gaps by applying Photo-ID in Musandam, an understudied region of Oman’s sea turtle populations. The effectiveness of an MPA depends on data-driven monitoring and adaptive management. In-water Photo-ID provides valuable population data, complementing bycatch, nesting, and satellite tracking programs to enhance regional knowledge. Additionally, this stepwise approach in Musandam enables targeted selection of turtles suitable for satellite tagging. The 54% of individuals identified in this study as transient with fewer than 2 resightings are of particular interest and emphasize the need for further tracking to determine broader movements and destinations within the NWIO and Persian Gulf. Expanding citizen science engagement, particularly through dive networks in Oman and the UAE, strengthens regional habitat connectivity and supports collaborative conservation efforts.

Replication of this study throughout Oman would be successful in expanding our knowledge of sea turtle populations, in particular in-water encounters, which without Photo-ID CMR remain poorly studied and undocumented. Engaging with the existing dive operators to promote sea turtle conservation and monitoring will be an important next step in progressing sea turtle research in Oman.

CONCLUSION

This study demonstrates Photo-ID-based CMR as an efficient tool for sea turtle population monitoring. The advantages include its noninvasiveness, cost effectiveness, and broad range of applications for sea turtle ecology. There is significant potential for expansion of in-water sea turtle monitoring efforts with the incorporation of stakeholders and citizen scientists as data sources. Continued monitoring of existing sites and wider geographical expansion in the Musandam Peninsula supports opportunities of resighting resident sea turtles and increases the potential of encountering new individuals including migrating adults. Like many conservation projects, COVID-19 impacted data collection, with no data collected between April 2020 and December 2021. Continued survey efforts in 2022 and 2023 have demonstrated Musandam to be an important area for in-water sea turtle activity, warranting a continuation and expansion of research efforts. A specific expedition exploring the peninsula fjords may help to identify specific habitats of interest; including potential seagrass beds and to elucidate any spatial connectivity of sea turtles by reducing the distances between sites surveyed. This may also be achieved by increasing the survey efforts for the more northern remote dive and snorkel sites that may yield results connecting sea turtles encountered in sites with close proximity to one another. This study expands on existing research on Photo-ID techniques and highlights their limited use in Oman. Photo-ID–based CMR offers an efficient, noninvasive approach to sea turtle monitoring that can be quickly and affordably scaled to maximize stakeholder engagement despite resource constraints.