Biological Conservation 261 (2021) 109235

Contents lists available at ScienceDirect

Biological Conservation

journal homepage: www.elsevier.com/locate/biocon

Distribution and temporal trends in the abundance of nesting sea turtles in the Red Sea

Takahiro Shimada a,b,c,d,*, Mark G. Meekan a, Robert Baldwin e, Abdulaziz M. Al-Suwailem f, Christopher Clarke f, August S. Santillan f, Carlos M. Duarte b a Australian Institute of Marine Science, Crawley, Western Australia 6009, Australia b Red Sea Research Center and Computational Bioscience Research Center, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia c UWA Oceans Institute and School of Biological Sciences, University of Western Australia, Crawley, WA, Australia d Department of Environment and Science, Queensland Government, GPO Box 2454, Dutton Park, QLD 4001, Australia e Five Oceans (Environmental Services) LLC, Box 660, PC 131, Oman f Beacon Development Company, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia

ARTICLE INFO ABSTRACT

Keywords: Mobile species often aggregate at predictable places and times to ensure that individuals findmates and breed in Population ecology suitable habitats. Sea turtles demonstrate this life history trait, which can make these species highly susceptible Coastal development to population declines if nesting habitats are lost or degraded. Conservation management thus requires knowl­ Climate change edge of where and when turtles nest and changes in abundance in these habitats through time. Here, we Red Sea compiled new and published data and used a novel analysis to describe seasonality, annual abundance and Nesting seasonality Sea turtle spatial distribution of nesting green (Chelonia mydas) and hawksbill (Eretmochelys imbricata) turtles in data- deficient populations that inhabit the Red Sea. Major new rookeries were identified for green turtles at Jazirat1 Mashabah (113 and 179 nesting females in 2018 and 2019) and for hawksbill turtles at Jazirat Al Waqqadi (79 nesting females in 2018), both of which are located on nearshore islands of the Kingdom of Saudi Arabia in an area subject to industrial, residential and ecotourism developments. An upward trend in annual abundance of nesting sea turtles was estimated at some sites including Ras Al Baridi (Saudi Arabia), a major rookery of green turtles in the Red Sea, where the annual numbers increased from 14–110 individuals in 1982–1995 to 178 and 330 individuals in 2018 and 2019. This integrative work provides the most up-to-date, comprehensive information on nesting sea turtles in the Red Sea and documents a critical baseline for sea tur­ tle conservation and future management effort.

1. Introduction development, pollution and climate change (Cristofari et al., 2018; Venter et al., 2016). Mobile species, such as birds, fishes, large mammals and reptiles Sea turtles provide good examples of such species. These animals often aggregate predictably in the same place and time for breeding over aggregate to breed at certain places and times across generations, and years and generations (Baker et al., 2013; Groot and Margolis, 1991; strong fidelityto breeding habitats has resulted in distinct genetic stocks Miller, 1997; Wheelwright and Mauck, 1998). This spatio-temporal fi­ within the species range (Jensen et al., 2013; Miller, 1997). Nesting is delity to particular breeding sites ensures that individuals, which might focused on sandy beaches where females deposit eggs. Anthropogenic otherwise be widely distributed for foraging across an environment, can threats in nesting beaches include loss or modification of suitable nest­ findmates and reproduce in suitable habitats (e.g. Shimada et al., 2020). ing beaches, light pollution due to industrial or residential de­ However, the loss or degradation of these habitats can have significant velopments, and climate change driven extreme storm surges, sea level repercussions for the viability of populations, which is a situation faced rise and warming (Fuentes et al., 2010; Lalo¨e et al., 2017; Pendoley and by many species today, due to anthropogenic threats such as Kamrowski, 2016). Combined with incidental boat strikes, by-catch in

* Corresponding author at: Department of Environment and Science, Queensland Government, GPO Box 2454, Dutton Park, QLD 4001, Australia. E-mail address: [email protected] (T. Shimada). 1 ‘Jazirat’ is the Arabic word for ‘island of’. https://doi.org/10.1016/j.biocon.2021.109235 Received 5 January 2021; Received in revised form 11 June 2021; Accepted 27 June 2021 Available online 3 August 2021 0006-3207/© 2021 Elsevier Ltd. All rights reserved. T. Shimada et al. Biological Conservation 261 (2021) 109235 fisheries, targeted hunting of adults, harvesting of eggs, and predation present, we lack any estimates of recent and long-term patterns in on hatchlings and adult females by feral animals (Campbell, 2003; abundance that might be used to determine trajectories of these pop­ Gronwald et al., 2019; Shimada et al., 2017), these threats have resulted ulations (Wallace et al., 2010). Additionally, turtles in the Red Sea breed in the long-term decline of major sea turtle populations throughout on islands and coastlines that are now undergoing rapid change through species ranges. Although some populations of sea turtles have shown very large developments, most notably three projects in the Kingdom of signs of recovery (Chaloupka et al., 2008; Mazaris et al., 2017), today, Saudi Arabia (Saudi Arabia hereafter) that encompass many dozens of many species are still categorised as Endangered or Critically Endan­ nearshore islands and hundreds of kilometres of the mainland coast (PIF, gered by the International Union for the Conservation of Nature Red List 2017). For these reasons, there is an urgent need for data on breeding (IUCN, 2020). patterns to support appropriate conservation strategies for sea turtles in The spatial concentration of nesting sea turtles, and the vulnerability the region. of adults, eggs, and hatchlings during this phase offers an obvious focal Our study aimed to address this issue by reporting the outcome of point for cost-effective conservation and management strategies that large scale (several hundreds of kilometres) surveys of nesting sea turtles seek to halt or reverse ongoing declines in populations (Hamann et al., along the coastlines and islands of the north-eastern Red Sea conducted 2010). Of particular concern are those stocks that are poorly docu­ since 2018. To provide a comprehensive review of sea turtle nesting in mented and/or facing imminent potential threats from developments the Red Sea, we consolidated this new data and published information driven by growing populations of humans in coastal environments. The on seasonality, distribution and abundance in nesting patterns. We also Red Sea contains populations of the Vulnerable green (Chelonia mydas) examined trends in abundance of nesting green and hawksbill turtles at and Critically Endangered hawksbill (Eretmochelys imbricata) turtles some locations where repeated surveys had been conducted. This work (IUCN, 2020; Mancini et al., 2019) that are thought to be genetically provides a revised baseline for sea turtle conservation in the region and distinct from others in the wider Indian Ocean (Jensen et al., 2019). At contributes to global assessments of sea turtle population status such as

Fig. 1. Study sites across the north-east Red Sea. Each symbol is a beach with evidence of nesting by green turtles (green circle), hawksbill turtles (purple triangle), or both green and hawksbill turtles (orange square). Beaches that were surveyed but no evidence of nesting was found are shown by black points. Grey polylines show the marine boundaries of three development projects (NEOM, Amaala, TRSP) and the general area of Ras Al Baridi. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)

2 T. Shimada et al. Biological Conservation 261 (2021) 109235 the IUCN Red List. months between July and February except for September. The TRSP – Ras Al Baridi region was surveyed at least once each month for a year. 2. Materials and methods The peaks of the nesting seasons were only identifiedat sites where more than one survey was conducted within the respective nesting season 2.1. New data (Supplementary Material – Section D).

2.1.1. Study area and dataset 2.1.3. Abundance of nesting turtles Field work was conducted between 2018 and 2020 along the Saudi At most sites, our surveys likely captured the representative number Arabian coast of the north-eastern Red Sea, where little prior informa­ of clutches and emergences that occurred up to the last survey. For this tion on use by sea turtles was available (Supplementary Material – analysis, we used all the clutches and emergences that were estimated to Sections A and B). Location and timing of surveys was dictated by lo­ have occurred within the respective nesting seasons. The cumulative gistics, resulting in unequal coverage and survey effort across time and number of clutches (Clutchesk) laid by a species up to the last survey date space (Supplementary Material – Section A). Our surveys included Ras was calculated as:

Al Baridi (a known major rookery of green turtles in the Red Sea) in Al ∑k Clutchesk = (Clutchesi + Emergences ⋅NS) Madinat al Munawwarahas province as well as islands and coastal areas i=1 i (1) designated for industrial, residential and ecotourism development in Tabuk province (Fig. 1). The latter are officially named as NEOM, where k was the number of surveys conducted at a site in the year, Amaala and The Red Sea Project (TRSP) from north to south (PIF, 2017). Clutchesi and Emergencesi were the number of clutches and emergences In 2018, we surveyed all potential nesting beaches of 49 islands and recorded at the ith survey, and NS was the nesting success rate (proba­ extensive lengths of the mainland coast (>500 km) over 24 days be­ bility of egg laying per emergence). The mean NS rate in this region was tween February and August. In 2019, we revisited three islands (An estimated to be 0.628 in the satellite tracking study of 20 green turtles = = Numan, Mashabah, Al Waqqadi) and mainland beaches at Amaala and that nested at Jazirat Mashabah (n 13) and Ras Al Baridi (n 7) Ras Al Baridi over 15 days between June and December. In 2020, three during the 2019 nesting season (Shimada et al., 2021). This constant NS islands (An Numan, Shaybarah, Al Waqqadi) and Ras Al Baridi were rate was used in Eq. (1) across the sites because the NS rate did not differ surveyed over 5 days between January and March. Details of site and between Jazirat Mashabah and Ras Al Baridi despite the contrasting timing of each survey are described in Supplementary Material – Sec­ environmental settings; Jazirat Mashabah is uninhabited whilst at Ras tions A and B. Al Baridi anthropogenic impacts are possibly greatest among all north- At each site we recorded the nests and tracks of turtles on beaches. eastern Red Sea rookeries (Pilcher, 1999; Shimada et al., 2021; this Entry to and exit from a beach were recorded as a single track. We study). The rates of NS calculated for green turtles were also applied to closely examined nests and tracks to identify the outcome of nesting emergences of hawksbill turtles as the mean NS rates of both species are activity based on signs such as the presence or absence of egg chambers, very similar when they nest on the same beaches (Kameda and Wakat­ the shapes of body pits, and nest camouflaging.A nest with evidence of suki, 2011 and Sea Turtle Association of Japan, unpublished; Mortimer egg deposition was categorised as a clutch, whereas a nest and a track et al., 2011; Okuyama et al., 2020). without a clear evidence of egg deposition was categorised as an If the total number of clutches (Clutches) and individual’s clutch emergence. The age of each clutch and emergence was also estimated as frequency per nesting season (Clutch frequency) are known, the annual ≤1 day, ≤2 weeks, ≤1 month, >1 month old, or last season. The age abundance of nesting turtles (Turtles) at a site can be calculated as: categories within a breeding season were assigned by observation of Clutches Turtles = physical evidence including debris, footprints, crab mounds, tidal wash Clutchfrequency (2) and vegetation. As very little rain falls between spring and autumn (nesting season) at our study area (<5 mm in total; Mashat and Abdel This simple method could only be applied to 13% of our data, which Basset, 2011), we assumed few tracks were eroded by rain during a likely represent the annual total of clutches and emergences at the nesting season. However we assumed that tracks were eroded during a respective nesting sites. This small portion of the data was collected wet season, which occurs between breeding seasons (Mashat and Abdel during the last half of the nesting season at the sites where nests and Basset, 2011). This assumption was validated by the observation of tracks remained visible for several months (Supplementary Material – several marked tracks across a nesting season. More details of how we Section C). The remaining data (87%) represent the partial count of examined and aged each nest and track are provided in Supplementary clutches and emergences in the respective seasons because the surveys Material – Section C. were conducted only up to the middle of the nesting season, and thus the subsequent clutches and emergences that might have occurred within 2.1.2. Detection of nesting season the same nesting season were not recorded. The seasonal trend of nesting activities (nesting seasonality) was It is possible to estimate the annual abundance of clutches from examined based on the dates of our surveys and estimated age of each partial count data if the proportion of the collected data relative to the clutch and emergence. Due to the large latitudinal differences and un­ total annual abundance is known. For a breeding population that only balanced frequency of surveys and timing among sites, we grouped the has one cohort per nesting season, this estimation can be relatively data into two regions (NEOM – Amaala and TRSP – Ras Al Baridi) and simple. For example, at Bramble Cay in Australia (Limpus et al., 2001), determined seasonality per region and species (Fig. 1). The date of each the number of available nesting green turtles since the beginning of a clutch and emergence was calculated by subtracting half of the esti­ nesting season (Days), calculated as the number of turtles that arrived at mated age from the date of the survey as the most parsimonious rule. For the nesting ground minus those departed, closely follows a normal dis­ example, if a ≤1 month old emergence was recorded on 16 August, it tribution with the mean (μ) and standard deviation (σ) of Days (Sup­ was associated with 1 August (16 minus 30/2 days). For this analysis, we plementary Material – Section E), and so can be modelled as: ) ≤ 2 only used the clutches and emergences that were estimated to be 1 Days ∼ N μ, σ (3) month old, so that estimated dates could be assigned to each record without prior knowledge of nesting seasonality (i.e. start and end of This means the expected proportion (Proportionk) of the clutch nesting season). End of nesting season was further confirmedby satellite number up to the last day of survey (Dayk) relative to the total annual tracking data collected from 30 green turtles that nested at Jazirat number can be calculated from the cumulative density function of a Mashabah and Ras Al Baridi in 2019 (Shimada et al., 2021). Collectively, normal distribution as: the NEOM – Amaala region was surveyed at least once per month for 8

3 T. Shimada et al. Biological Conservation 261 (2021) 109235 ∫ Dayk Due to variation in the technique and timing of data collection in the Proportion = f (x)dx, for Day ≤ Day k k last (4) literature, we standardised historical estimates of the annual abundance Dayfirst of nesting female turtles using the above procedures (2.1.3). We fine- where Dayfirst and Daylast are the firstand last days of a nesting season at tuned estimates case by case, based on timing and duration of each a given site for the species and when Dayk > Daylast, Daylast was replaced survey and the nesting seasonality of the species at each site. Details of with Dayk. We verified that the annual total clutches and emergences historical data, each adjustment, and standardised estimates are pro­ known (13% of our data) followed the cumulative distribution functions vided in Supplementary Material – Section F. of the normal distribution (Fig. 2). For each region (NEOM – Amaala, Where estimates were available for more than one season, annual TRSP – Ras Al Baridi) and species (green, hawksbill), the mean for the abundance was averaged over the three most recent years of surveys. normal distribution was the mid-point of the estimated nesting season Worldwide, individual female green and hawksbill turtles breed on because, similar to the example of the Bramble Cay green turtles, there average once every three years (Miller, 1997). was only one nesting season cohort for each species at our sites (Sup­ plementary Material – Section D). Standard deviations and 95% confi­ 2.3. Abundance trend of nesting turtles dence intervals (CI) were estimated from the data by maximum likelihood estimation using the R packages stats and stats4 (R Core To compare trends among sites and years, data were standardised to Team, 2020). We used those probability density functions, prepared for estimate annual abundance of nesting turtles as described above and in each region (NEOM – Amaala, TRSP – Ras Al Baridi) and species (green, Supplementary Material – Section F. hawksbill), to calculate the expected proportion (Proportionk) of the cumulative number of clutches (Clutchesk) up to Dayk relative to the total 3. Results annual abundance. The annual clutch numbers (Clutches) of a species at each site and nesting season was then: Between 2018 and 2020, we recorded a total of 4613 and 1329 clutches and emergences of green and hawksbill turtles respectively Clutches Clutches = k along the Saudi Arabian coast of the northern Red Sea (Supplementary Proportion (5) k Material – Section F). Among these we identified 3158 green and 713 Finally, we estimated the annual abundance of nesting turtles (Tur­ hawksbill clutches and emergences that had occurred within the tles) per site and species by including the estimated annual clutch respective nesting seasons, and these were used to estimate the annual numbers (Clutches) in Eq. (2). For Clutch frequency in Eq. (2), we used the abundance of nesting females at each site. This new evidence of nesting global means, which are 5.9 clutches for green and 2.74 clutches for was collected at 26 sites (23 island and 3 mainland) for green turtles and hawksbill turtles (Esteban et al., 2017; Miller, 1997). These values are 50 sites (46 island and 4 mainland) for hawksbill turtles (Supplementary likely to be accurate approximations of the true mean clutch frequencies Material – Section F). Earlier studies had previously identified47 green of the nesting turtles in this region (Pilcher et al., 2014; Shimada et al., and 62 hawksbill nesting sites in the entire Red Sea (Supplementary 2021). Material – Section F). In total, we compiled data on clutches, emer­ We applied the above procedure to most datasets where all the gences, and individual nesting turtles from 78 green and 110 hawksbill clutches and emergences up to the last survey dates were presumed to be rookeries in the Red Sea where turtle nesting was confirmed between collected (Supplementary Material – Section C). Exceptions were the 1976 and 2020. From this data, we provide a revised overview of sea­ data collected only once during a nesting season of green turtles at Ras sonality, distribution, and abundance of the sea turtles in the Red Sea. Al Baridi in 2018, and at two islands (Mashabah, Al Waqqadi) in 2019. Details of new and published data are summarised in Supplementary At these sites most tracks only remain apparent for no more than a Material – Section F. month because of the high density of nesting, and in the case of Ras Al Baridi, a large amount of wind-blown cement dust from a nearby factory 3.1. Seasonality that accumulates over the nesting beaches (Pilcher, 1999; Supplemen­ tary Material – Section C). As these surveys only captured snapshots of 3.1.1. Green turtles clutches and emergences, the estimation method described above was In NEOM – Amaala, nesting of green turtles was confirmedbetween not applicable. Instead, we contrasted the number of emergences of 22 May and 9 October with the peak estimated in early August (Fig. 2a, these snapshot data to those of frequent surveys at Ras Al Baridi in 2019 Supplementary Material – Section D). In TRSP – Ras Al Baridi, green (Supplementary Material – Section A). In 2018 at Ras Al Baridi, the turtles nested between 21 April and 27 November with the peak esti­ – survey was conducted on 20 August (Emergences20Aug2018) and the same mated around mid-August (Fig. 2c, Supplementary Material Section beaches were surveyed again on 21 August in 2019 (Emergence­ D). Green turtles appear to nest slightly earlier with a peak in July/ s21Aug2019). The annual abundance of green turtles that nested at Ras Al August at Zabargad Island, Egypt (Hanafy and Sallam, 2003; cited in Baridi in 2018 (CMRB2018) was estimated proportionally from the annual Hanafy, 2012), which is another major rookery of green turtles in the – abundance estimated for 2019 (CMRB2019) as: Red Sea located approximately 150 230 km south-west of the Saudi Arabian sites (Fig. 3). Emergences CM = 20Aug2018⋅CM RB2018 Emergences RB2019 (6) 21Aug2019 3.1.2. Hawksbill turtles Similarly, the single surveys conducted at two islands (Al Waqqadi, In NEOM – Amaala, nesting of hawksbill turtles was observed be­ Mashabah on 7, 8 September 2019 receptively) were contrasted with the tween 22 May and 30 June with the peak likely around late May to early data collected at Ras Al Baridi on 4 September 2019 to proportionally June (Fig. 2b, Supplementary Material – Section D), implying their – estimate the annual abundances from CMRB2019. nesting season starts in early May or possibly in April. In the TRSP Ras Al Baridi region, we found evidence of hawksbill turtles nesting between 13 April and 29 July. Additionally, both adult male and female 2.2. Published information hawksbills were seen frequently in early March in the shallow waters adjacent to the nesting beaches within TRSP, although no surveys were We examined literature cited by seminal reviews and reports (Man­ possible between late March and mid-April due to logistical difficulties. cini et al., 2015; Phillott and Rees, 2019) and 524 papers published after This combined evidence suggests that in TRSP – Ras Al Baridi, hawksbill 2015 as identified by a Google Scholar search (14 March 2021) using a turtles likely start nesting in early April through July with the peak in combination of key words “Red Sea”, “nesting”, and “turtle”. late May and early June (Fig. 2d, Supplementary Material – Section D).

4 T. Shimada et al. Biological Conservation 261 (2021) 109235

Fig. 2. Cumulative distribution functions of Normal distribution (line) with the 95% confidence intervals (yellow bands), and cumulative nest abundance as pro­ portion to the total clutch counts (points) for green and hawksbill turtles at (a, b) NEOM – Amaala, and (c, d) TRSP – Ras Al Baridi. Ticks along the x axes show the first date of each month. Note clutch counts and the associated dates were estimated from count data of clutches and emergences as described in Supplementary Material – Section C and main text. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Nesting seasons of green and hawksbill turtles in the Red Sea. Lighter colours indicate nesting activities in each month (shown by a capital letter above each box) with darker colours denoting the peak periods. Months in bold mean that the presence or absence of nesting activities was confirmedby our field survey during the current study and from satellite tracking data from Shimada et al. (2021). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

5 T. Shimada et al. Biological Conservation 261 (2021) 109235

Published studies on nesting seasonality of hawksbill turtles in the than 50 individuals (Fig. 4, Supplementary Material – Section F). Red Sea at Giftun Islands in Egypt (Hanafy and Sallam, 2003; cited in Hanafy, 2012), Mukawwar Island and Suakin Archipelago in Sudan 3.2.3. Other sea turtles (PERSGA/GEF, 2007), Juzur2 Farasan in Saudi Arabia (PERSGA/GEF, There is only one reported incidence of nesting by an olive ridley 2007), and the Dahlak Archipelago in Eritrea (Eritrean Department of turtle (Lepidochelys olivacea) in the Red Sea, on the southern coast of Environment, 2014), together with our new data, suggested latitudinal Eritrea (Pilcher et al., 2006). All other confirmedcases of nesting in the effects on seasonality of the nesting events (Fig. 3). Hawksbill turtles Red Sea were of green and hawksbill turtles (Supplementary Material – began nesting in December with a peak from February to April on the Section F). southern rookeries at Dahlak Archipelago, whereas nesting did not begin until May with a peak in June on the northern rookeries at Amaala, 3.3. Abundance trends Giftun Islands, and NEOM, showing a clear delay in nesting activity with increasing latitude (Fig. 3). From the current study and the literature, we synthesised data collected at 10 green turtle rookeries and 12 hawksbill turtle rookeries 3.2. Distribution and abundance in the Red Sea, where annual abundance data are available for more than one year. 3.2.1. Green turtles Major rookeries of green turtles are aggregated in the northern Red 3.3.1. Green turtles ◦ ◦ Sea between 24.60 N and 25.63 N (Fig. 4). The largest aggregation was Ras Al Baridi is one of the most surveyed nesting sites of green turtles found at Ras Al Baridi with annual estimates of 178 (95% CI = 121–362) in the Red Sea. The abundance of nesting females was firstestimated in and 330 (95% CI = 225–675) nesting individuals in 2018 and 2019 1983 (Ormond et al., 1984), followed by more comprehensive studies respectively (Fig. 4, Supplementary Material – Section F). The second between 1987 and 1995 (Al-Merghani et al., 2000). The current study largest aggregation occurred within TRSP, where a total of 185 turtles provides the most recent estimates from the 2018 and 2019 nesting (95% CI = 101–604) were estimated to have nested in 2018 across 16 seasons. This combination of historical and new data provides abun­ islands (Fig. 4, Supplementary Material – Section F). Approximately dance estimates from Ras Al Baridi across 11 nesting seasons between 61% of the nesting within TRSP occurred at Jazirat Mashabah with the 1983 and 2019. Additionally, six other sites that Ormond et al. (1984) abundance estimated at 113 (95% CI = 80–219) and 179 (95% CI = visited in 1983 were also monitored during the current study (Supple­ 122–367) nesting turtles in 2018 and 2019 respectively. At NEOM, mentary Material – Section F). Nesting was also reported for 12 seasons across six island and one mainland sites combined, a total of 58 green at Zabargad Island between 2001 and 2014 (El-Sadek et al., 2016; turtles (95% CI = 18–321) were estimated nesting in 2018 (Fig. 4, Hanafy, 2012). Supplementary Material – Section F). This is probably a conservative The annual abundance of nesting green turtles appears to have estimate since two potential rookeries at Jazirat Thiran and Jazirat increased since 1980–90s at Ras Al Baridi. We estimated that, on Sanafircould not be surveyed. From El-Sadek et al. (2016) we estimated average, 43 green turtles nested between 1982 and 1995, whereas 254 that 62–168 (mean = 110) green turtles nested annually between 2009 individuals nested annually between 2018 and 2019 (Fig. 5, Supple­ and 2014 in Zabargad Island (Fig. 4, Supplementary Material – Section mentary Material – Section F). An increase was also apparent at Jazirat F). On other rookeries in the Red Sea, the mean number of nesting green Mashabah, where the mean annual abundance estimate of 10 in­ turtles appears to be less than 50 individuals per annum (Fig. 4, Sup­ dividuals in 1983 increased to 146 turtles between 2018 and 2019 plementary Material – Section F). (Fig. 5, Supplementary Material – Section F). At other rookeries, the trend in nesting suggests stable or slightly increasing numbers (Fig. 5, 3.2.2. Hawksbill turtles Supplementary Material – Section F). Aggregations of >50 nesting hawksbill turtles occur in both the north and south of the Red Sea. In the northern Red Sea, the largest aggre­ 3.3.2. Hawksbill turtles gation was found at TRSP where we estimated 183 hawksbill turtles Nine hawksbill turtle rookeries in Saudi Arabia were surveyed in (95% CI = 152–217) nested in 2018 across 37 islands, with 79 of these 1983 (Ormond et al., 1984) and then again in 2018–2019 (this study). (95% CI = 77–84) or 43% nesting at Jazirat Al Waqqadi. At NEOM, a Along the western side of the Red Sea, several years of abundance data total of 65 hawksbills (95% CI = 62–69) were estimated nesting in 2018 were reported from Giftun Islands, Egypt (Hanafy, 2012), and two years across 10 sites (8 island and 2 mainland), with approximately 67% of of data from Mojeidi Island, Eritrea (Teclemariam et al., 2009). nesting occurring at Jazirat Shushah and Jazirat Walah (Fig. 4, Sup­ There was an apparent increase in the annual abundance of nesting plementary Material – Section F). Similar to green turtles, the actual hawksbill turtles at Jazirat Al Waqqadi, Saudi Arabia from 14 in­ abundance of nesting hawksbill turtles in NEOM is likely greater because dividuals in 1983 to 79 individuals in 2018 (Fig. 5, Supplementary two potential nesting sites were inaccessible for survey (Jazirat Thiran Material – Section F). Surveys at Big Giftun Island also showed a small and Jazirat Sanafir). Other large aggregations of nesting hawksbill tur­ increase in numbers during the 2000s with 6 individuals nesting in 2001 tles have been reported to occur in the southern part of the Red Sea. and 31 nesting in 2007 (Fig. 5, Supplementary Material – Section F). At Moore and Balzarotti (1977; cited in Groombridge and Luxmoore, 1989) Ras Al Baridi, where no hawksbill turtles had been recorded from 1983 estimated that approximately 330 hawksbill turtles nested at Suakin to 1995 (Al-Merghani et al., 2000; Ormond et al., 1984), four fresh Archipelago in 1976. In 1983 Ormond et al. (1984) surveyed the Saudi tracks of hawksbill turtles were recorded in June 2019. A decline in Arabian coast of the Red Sea and found nests of hawksbill turtles on 48 nesting was suggested at Jazirat Bargan, Saudi Arabia, with the esti­ islands. The largest of these were at Jazirat Marrak and Jazirat Dohrab mates of 41 individuals in 1983 but only 3 individuals in 2018 (Fig. 5, (part of Juzur Farasan), where we estimated from nest count data that 73 Supplementary Material – Section F). Overall, at some locations nesting (range 37 to 110) hawksbill turtles nested at each island that year. In by hawksbill turtles appear to be stable or have increased since the Eritrea, at least 47 and 96 hawksbills nested on the rookeries within 1980s, whereas other nesting aggregations may be in decline (Fig. 5, Dahlak Archipelago in 2006 and 2007 respectively (Teclemariam et al., Supplementary Material – Section F). 2009). On other islands and mainland rookeries of the Red Sea, the number of hawksbill turtles nesting annually was estimated to be less 4. Discussion

Using long term data sets and a novel approach to analysis, we have 2 ‘Juzur’ is the Arabic word for ‘islands of’. provided a comprehensive overview of the seasonality, distribution, and

6 T. Shimada et al. Biological Conservation 261 (2021) 109235

Fig. 4. Distribution and estimated abundance of nesting (a) green and (b) hawksbill turtles in the Red Sea. The size and colour of each bubble are relative to the estimated annual number of nesting females at each site. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.) abundance of nesting sea turtles in the Red Sea. Importantly, we have individuals) and 2018 (79 individuals) were beyond the expected annual identified40 new sites used for nesting (38 island and 2 mainland) and fluctuationsfor this species, which typically do not vary more than two- found evidence that, for at least at some rookeries, abundance of nesting fold across proximate years (Bell et al., 2020). Additionally, our esti­ females is likely to have increased over the last three decades. Addi­ mates of annual abundance are likely to be conservative measures, since tionally, the timing of nesting has shifted in some rookeries, possibly as a some nests and tracks at these major rookeries may have become response to climate change. obscured prior to the surveys. Although our results must be interpreted Annual patterns in the abundance of nesting sea turtles must be with caution, particularly since many of the surveys in 1980/90s were interpreted with care. As shown at Ras Al Baridi and elsewhere in the short-term (sometimes involving only brief visits to beaches that may world, numbers in proximate years almost always fluctuateregardless of not have detected intermittent nesting), overall the weight of evidence the overall trend in the population (Al-Merghani et al., 2000; Chaloupka suggests that populations at Ras Al Baridi, and potentially at two islands et al., 2008). The breeding biology of females and their foraging envi­ (Mashabah, Al Waqqadi), have indeed increased in abundance over the ronment are two important elements that dictate these changes in last three decades. This may be the result of management measures put numbers. Sea turtles are capital breeders and an adult female typically in place in other areas of the population range across the Red Sea, since requires more than one year to attain a body condition suitable to pro­ no formal protection had been given to nesting sites in Saudi Arabia duce eggs (Miller, 1997). The time required to accumulate these energy (Mancini et al., 2015). reserves depends on the accessibility and quality of food, which in turn is The occurrence of females at sites where nesting activities have not largely influenced by environmental conditions. As a result, there is been reported previously (e.g. hawksbills at Ras Al Baridi) might be often synchronicity between the timing of breeding and climatic events interpreted as further evidence for increases in population size. How­ that drive patterns of productivity. For example, in the Pacific Ocean, ever, it is important to note that except for Ras Al Baridi, many sites were breeding of green turtles is correlated with El Nino-Southern˜ Oscillation surveyed very infrequently in the past, so that evidence of nesting ac­ events (Limpus and Nicholls, 2000; Santidrian´ Tomillo et al., 2020). For tivities might not have been detected. It is also possible that some turtles any breeding population, the cycle of annual fluctuationsin abundance might have temporarily shifted their nesting sites, although such of nesting females roughly equates to the average interval between two movements are infrequent (Limpus and Miller, 2008; Shimada et al., consecutive breeding seasons, which in general is 3–6 years for green 2021). While bearing these caveats in mind, the new nesting sites and hawksbill turtles (Miller, 1997). However, the changes in numbers identified in this study are located within the area of long-established of nesting females that we observed across decades at some of the major rookeries of the species, suggesting that if indicative of increasing rookeries in the Red Sea (Ras Al Baridi, Jazirat Mashabah, Jazirat Al populations, recolonisation into parts of former ranges is occurring Waqqadi) were still much larger than might be expected for typical rather than expansion of nesting into new habitats. Hawksbill and green annual fluctuations. For example, annual abundance of green turtles at turtles have been severely exploited for shells, eggs and meat across the Ras Al Baridi was much greater in 2018 and 2019 (178 and 330 in­ globe including the Red Sea (Mancini et al., 2015), but since the dividuals, respectively) than any fluctuations observed between 1983 implementation of conservation strategies in many places, some pop­ and 1995 (range 17 to 105 individuals), implying an increase in abun­ ulations have shown signs of recovery in abundance (Chaloupka et al., dance over the two decades. Similarly, at Jazirat Al Waqqadi, the dif­ 2008; Hanafy, 2012) and may be recolonising former ranges. Such ference in estimated abundances of hawksbill turtles between 1983 (14 recolonisation is a phenomenon common to population recovery of both

7 T. Shimada et al. Biological Conservation 261 (2021) 109235

Fig. 5. Annual abundance of nesting (a) green and (b) hawksbill turtles since 1983 in the Red Sea. Dashed lines connect data points of each site. See the Materials and Methods and Supplementary Material – Section F for details of how the abundance of nesting females were estimated. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) aquatic and terrestrial fauna (Chapron et al., 2014; Lafferty and Tinker, Environmental Information, 2016). These warming trends may have 2014). driven turtles to now commence nesting earlier in a breeding season Our results show that the nesting seasons of sea turtles begins some than a few decades in the past, a pattern consistent with changes in the months earlier in the south than in the north of the Red Sea, likely due to timing of breeding seasons in response to changes in temperatures across latitudinal-driven changes in temperatures. The Red Sea extends from a wide range of species (Mazaris et al., 2009; Visser et al., 2009). ◦ 12.5 to 30 N and the difference in water temperatures across this range Relocation of nesting grounds to cooler environments could be an ◦ of latitudes could be >5 C (Agulles et al., 2020). Low temperatures can alternative response to warming temperatures but is rare in species that slow down or cease the development of turtle embryos, whereas expo­ have fixed breeding sites, presumably because it is riskier to breed in a ◦ sure to very high temperatures (>~35 C) will be lethal (Howard et al., new, unknown habitat than to shift the timing of breeding at the same 2014). In the warmer southern Red Sea, the optimal incubation tem­ site. This emphasises the importance of the conservation of long-term ◦ perature (~29 C; Howard et al., 2014) likely occurs in late winter and nesting beaches for sea turtles in the Red Sea. spring, whereas in the cooler northern Red Sea, optima may not be Spatially explicit management strategies can be highly effective for achieved until the summer arrives some months later. Such differences the conservation of sea turtles given their strong fidelity to nesting may have implications for the timing (and potentially spatial distribu­ beaches and inter-nesting habitats (Jensen et al., 2013; Shimada et al., tion) of breeding under global warming. 2021). We identified nesting beaches that are critical for the sustain­ The effects of rising temperatures may also explain the apparent ability of these endangered species in the Red Sea. Although the trend of shifts in seasonality of nesting by sea turtles over decades. In 2019, increasing abundances at some nesting aggregations are encouraging, nesting by green turtles at Ras Al Baridi started as early as April with the other rookeries had declining numbers of females and most populations peak in August, whereas in the 1980s to early 1990s nesting only began are likely facing new and persistent anthropogenic threats including in July/August with a peak in September/October (Al-Merghani et al., climate change, coastal development and beach armouring, pollution, 2000; Pilcher and Al-Merghani, 2000; this study). This change has been both targeted and incidental catch in fisheries, and tourism (Hamann accompanied by warming of the Red Sea over the last half century, et al., 2010; Phillott and Rees, 2019). For this reason, it is important that ◦ which has trended upwards in temperature at a rate of 0.045 ± 0.016 C the key nesting beaches receive comprehensive protection and that in­ per decade at 15 m depths (Agulles et al., 2020). Just off the coast of Ras formation from older surveys of important aggregation sites of nesting ◦ ◦ Al Baridi (24.125 N, 37.625 E), the mean sea surface temperature was females (e.g. Suakin Archipelago, Juzur Farasan) is updated to assess the ◦ ◦ 26.1 C in 1983 but averaged 27.7 C in 2019 (National Centers for potential impacts of these escalating threats.

8 T. Shimada et al. Biological Conservation 261 (2021) 109235

The results from our 2018–2019 surveys have already had a positive References influence on conservation management of sea turtles. For example, ` Jazirat Al Waqqadi, one of the major rookeries of the hawksbill turtles in Agulles, M., Jorda, G., Jones, B., Agustí, S., Duarte, C.M., 2020. Temporal evolution of temperatures in the Red Sea and the Gulf of Aden based on in situ observations the Red Sea, has been entirely exempted from development and receives (1958–2017). Ocean Sci. 16, 149–166. https://doi.org/10.5194/os-16-149-2020. greater protection by the managing body (The Red Sea Development Al-Merghani, M., Miller, J.D., Pilcher, N.J., Al-Mansi, A., 2000. The green and hawksbill Company). Developments occurring nearby on other significant rook­ turtles in the Kingdom of Saudi Arabia: synopsis of nesting studies 1986–1997. Fauna of Arabia 18, 369–384. eries such as Jazirat Mashabah, are also being carefully planned to avoid Baker, C.S., Steel, D., Calambokidis, J., Falcone, E., Gonzalez-Peral,´ U., Barlow, J., loss of important habitats and to mitigate impacts such as light pollution Burdin, A.M., Clapham, P.J., Ford, J.K.B., Gabriele, C.M., Mattila, D., Rojas- on nesting turtles and hatchlings. Other major rookeries, notably Ras Al Bracho, L., Straley, J.M., Taylor, B.L., Urban,´ J., Wade, P.R., Weller, D., Witteveen, B. H., Yamaguchi, M., 2013. Strong maternal fidelity and natal philopatry shape Baridi, still have no formal systems of management in place to protect genetic structure in North Pacific humpback whales. Mar. Ecol. Prog. Ser. 494, breeding populations of turtles. Given existing and predicted anthro­ 291–306. https://doi.org/10.3354/meps10508. pogenic pressures (Pilcher, 1999; Shimada et al., 2021), it is highly Bell, I.P., Meager, J.J., Eguchi, T., Dobbs, K.A., Miller, J.D., Madden Hof, C.A., 2020. desirable that Ras Al Baridi and other important sea turtle nesting Twenty-eight years of decline: nesting population demographics and trajectory of the north-east Queensland endangered hawksbill turtle (Eretmochelys imbricata). habitats in the Red Sea are monitored regularly and receive compre­ Biol. Conserv. 241, 108376 https://doi.org/10.1016/j.biocon.2019.108376. hensive protection. Campbell, L.M., 2003. Contemporary culture, use, and conservation of sea turtles. In: Lutz, P.L., Musick, J.A., Wyneken, J. (Eds.), The Biology of Sea Turtles, vol. II. CRC Press, Boca Raton, FL, pp. 307–338. CRediT authorship contribution statement Chaloupka, M., Bjorndal, K.A., Balazs, G.H., Bolten, A.B., Ehrhart, L.M., Limpus, C.J., Suganuma, H., Troeeng, S., Yamaguchi, M., 2008. Encouraging outlook for recovery Takahiro Shimada: Conceptualisation; Methodology; Software; of a once severely exploited marine megaherbivore. Glob. Ecol. Biogeogr. 17, 297–304. https://doi.org/10.1111/j.1466-8238.2007.00367.x. Validation; Formal analysis; Investigation; Data curation; Writing - Chapron, G., Kaczensky, P., Linnell, J.D.C., Arx, M. von, Huber, D., Andren,´ H., Lopez-´ ´ Original draft preparation; Writing - Reviewing & Editing; Visualization; Bao, J.V., Adamec, M., Alvares, F., Anders, O., Balˇciauskas, L., Balys, V., Bedo,˝ P., Bego, F., Blanco, J.C., Breitenmoser, U., Brøseth, H., Bufka, L., Bunikyte, R., Supervision; Project administration; Funding acquisition. ¨ & Ciucci, P., Dutsov, A., Engleder, T., Fuxjager, C., Groff, C., Holmala, K., Hoxha, B., Mark G. Meekan: Resources; Writing - Review Editing; Supervi­ Iliopoulos, Y., Ionescu, O., Jeremi´c, J., Jerina, K., Kluth, G., Knauer, F., Kojola, I., sion; Project administration; Funding acquisition. Kos, I., Krofel, M., Kubala, J., Kunovac, S., Kusak, J., Kutal, M., Liberg, O., Maji´c, A., ¨ Robert Baldwin: Conceptualisation; Methodology; Investigation; Mannil, P., Manz, R., Marboutin, E., Marucco, F., Melovski, D., Mersini, K., ł & Mertzanis, Y., Mys ajek, R.W., Nowak, S., Odden, J., Ozolins, J., Palomero, G., Writing - Original draft preparation; Writing - Review Editing; Su­ Paunovi´c, M., Persson, J., Potoˇcnik, H., Quenette, P.-Y., Rauer, G., Reinhardt, I., pervision; Project administration. Rigg, R., Ryser, A., Salvatori, V., Skrbinˇsek, T., Stojanov, A., Swenson, J.E., ´ˇ Abdulaziz M. Al-Suwailem: Resources; Supervision; Project Szemethy, L., Trajçe, A., Tsingarska-Sedefcheva, E., Vana, M., Veeroja, R., Wabakken, P., Wolfl,¨ M., Wolfl,¨ S., Zimmermann, F., Zlatanova, D., Boitani, L., 2014. administration; Funding acquisition. Recovery of large carnivores in Europe’s modern human-dominated landscapes. Christopher Clarke: Investigation; Writing - Review & Editing; Science 346, 1517–1519. https://doi.org/10.1126/science.1257553. Supervision; Project administration. Cristofari, R., Liu, X., Bonadonna, F., Cherel, Y., Pistorius, P., Maho, Y.L., Raybaud, V., Stenseth, N.C., Bohec, C.L., Trucchi, E., 2018. Climate-driven range shifts of the king August S. Santillan: Investigation. penguin in a fragmented ecosystem. Nat. Clim. Chang. 8, 245–251. https://doi.org/ Carlos M. Duarte: Conceptualisation; Resources; Writing - Review & 10.1038/s41558-018-0084-2. Editing; Supervision; Project administration; Funding acquisition. El-Sadek, I., Ahmed, M.I., Aamer, M.A., Mancini, A., Hanafy, M.H., 2016. Nesting activities of green turtles (Chelonia mydas) on the beaches of Zabargad Island, southern Egyptian Red Sea. Egypt. J. Aquatic Biol. Fish. 20, 29–37. https://doi.org/ Ethics 10.21608/ejabf.2016.11175. Eritrean Department of Environment, 2014. The 5th National Report on the Implementation of the UNCBD. The State of Eritrea, Ministry of Land and No ethics approval was required to conduct the research as data Environment. Department of Environment, Asmara, Eritrea. collection did not involve interactions with the animals. Esteban, N., Mortimer, J.A., Hays, G.C., 2017. How numbers of nesting sea turtles can be overestimated by nearly a factor of two. Proc. Roy. Soc. B: Biol. Sci. https://doi.org/ 10.1098/rspb.2016.2581. Funding Fuentes, M., Limpus, C.J., Hamann, M., Dawson, J., 2010. Potential impacts of projected sea-level rise on sea turtle rookeries. Aquat. Conserv. Mar. Freshwat. Ecosyst. 20, This research was funded by The Red Sea Development Company 132–139. https://doi.org/10.1002/aqc.1088. Gronwald, M., Genet, Q., Touron, M., 2019. Predation on green sea turtle, Chelonia (TRSDC), Amaala, Beacon Development Company (BDC), King Abdullah mydas, hatchlings by invasive rats. Pac. Conserv. Biol. 25, 423–424. https://doi.org/ University of Science and Technology (KAUST), and Australian Institute 10.1071/PC18087. of Marine Science. Groombridge, B., Luxmoore, R., 1989. The green turtle and hawksbill (Reptilia: Cheloniidae): world status, exploitation and trade. CITES (Convention on International Trade in Endangered Species) Secretariat, Lausanne, Switzerland. Declaration of competing interest Groot, C., Margolis, L. (Eds.), 1991. PacificSalmon Life Histories. UBC Press, Vancouver. Hamann, M., Godfrey, M.H., Seminoff, J.A., Arthur, K., Barata, P.C.R., Bjorndal, K.A., Bolten, A.B., Broderick, A.C., Campbell, L.M., Carreras, C., Casale, P., Chaloupka, M., No conflict to declare. Chan, S.K.F., Coyne, M.S., Crowder, L.B., Diez, C.E., Dutton, P.H., Epperly, S.P., FitzSimmons, N.N., Formia, A., Girondot, M., Hays, G.C., Cheng, I.J., Kaska, Y., Lewison, R., Mortimer, J.A., Nichols, W.J., Reina, R.D., Shanker, K., Spotila, J.R., Acknowledgments Tomas, J., Wallace, B.P., Work, T.M., Zbinden, J., Godley, B.J., 2010. Global research priorities for sea turtles: informing management and conservation in the – We thank staff and numerous volunteers of TRSDC, Amaala, NEOM, 21st century. Endanger. Species Res. 11, 245 269. https://doi.org/10.3354/ esr00279. BDC and KAUST for their support, in particular Reny Devassy, Abhish­ Hanafy, M., 2012. Nesting of marine turtles on the Egyptian beaches of the Red Sea. ekh Vijayan, Raied A. Ajahdali, James Massey, Tito P. Pancho, and Zaki Egypt. J. Aquatic Biol. Fish. 16, 59–71. https://doi.org/10.21608/ejabf.2012.2125. Al Jahdali. We also thank Jeffrey D. Miller and Nicolas Pilcher for their Hanafy, M.H., Sallam, A., 2003. Status of Marine Turtles Nesting on the Egyptian Beaches of the Red Sea. PERSGA. valuable advice and sharing key literature, M. Hamann for sharing raw Howard, R., Bell, I., Pike, D.A., 2014. Thermal tolerances of sea turtle embryos: current data from Limpus et al. 2001, and Sea Turtle Association of Japan for understanding and future directions. Endanger. Species Res. 26, 75–86. https://doi. sharing unpublished data. org/10.3354/esr00636. IUCN, 2020. The IUCN red list of threatened species. Version 2020-3. [WWW Document]. URL https://www.iucnredlist.org (Accessed 12.22.20). Appendix A. Supplementary material Jensen, M.P., FitzSimmons, N.N., Dutton, P.H., 2013. Molecular genetics of sea turtles, in: Wyneken, J., Lohmann, K.J., Musick, J.A. (Eds.), The Biology of Sea Turtles, Volume vol. III. CRC Press, Boca Raton, FL, pp. 135–161.doi:https://doi.org/10.12 Supplementary material of this article can be found online at https 01/b13895. ://doi.org/10.1016/j.biocon.2021.109235.

9 T. Shimada et al. Biological Conservation 261 (2021) 109235

Jensen, M., Miller, J., FitzSimmons, N., Al-Merghani, M., 2019. Identificationof Chelonia PERSGA/GEF, 2007. Regional action plans for the conservation of marine turtles, mydas populations in the Kingdom of Saudi Arabia through regional genetic seabirds and mangroves in the Red Sea and Gulf of Aden (No. 12). In: PERSGA analyses. Mar. Turt. Newsl. 16–20. Technical Series. PERSGA, Jeddah, Saudi Arabia. Kameda, K., Wakatsuki, M., 2011. Reproductive biology of the hawksbill turtle, Phillott, A.D., Rees, AL.F., 2019. Sea Turtles in the Middle East and South Asia Region Eretmochelys imbricata, on Kuroshima Island of Yaeyama group, Ryukyu archipelago. MTSG Annual Regional Report 2019 (Report of the IUCN-SSC Marine Turtle Umigame News Letter 12–14. Specialist Group). Lafferty, K.D., Tinker, M.T., 2014. Sea otters are recolonizing southern California in fits PIF, 2017. The Public Investment Fund Program (2018–2020). Public Investment Fund. and starts. Ecosphere 5, art50. https://doi.org/10.1890/ES13-00394.1. Pilcher, N.J., 1999. Cement dust pollution as a cause of sea turtle hatchling mortality at Lalo¨e, J.-O., Cozens, J., Renom, B., Taxonera, A., Hays, G.C., 2017. Climate change and Ras Baridi, Saudi Arabia. Mar. Pollut. Bull. 38, 966–969. https://doi.org/10.1016/ temperature-linked hatchling mortality at a globally important sea turtle nesting S0025-326X(99)00110-1. site. Glob. Chang. Biol. 23, 4922–4931. https://doi.org/10.1111/gcb.13765. Pilcher, N.J., Al-Merghani, M., 2000. Reproductive biology of green turtles at Ras Baridi, Limpus, C.J., Miller, J.D., 2008. Australian Hawksbill Turtle Population Dynamics Saudi Arabia. Herpetol. Rev. 31, 142–147. Project. Environmental Protection Agency, Queensland. Pilcher, N., Mahmud, S., Howe, S., Teclemariam, Y., Weldeyohannes, S., 2006. An update Limpus, C., Nicholls, N., 2000. ENSO regulation of indo-Pacificgreen turtle populations. on Eritrea’s marine turtle programme and firstrecord of olive ridley turtle nesting in In: Hammer, G.L., Nicholls, N., Mitchell, C. (Eds.), Applications of Seasonal Climate the Red Sea. Mar. Turt. Newsl. 16. Forecasting in Agricultural and Natural Ecosystems - the Australian Experience, Pilcher, N.J., Antonopoulou, M., Perry, L., Abdel-Moati, M.A., Al Abdessalaam, T.Z., pp. 399–408. Albeldawi, M., Al Ansi, M., Al-Mohannadi, S.F., Al Zahlawi, N., Baldwin, R., Limpus, C.J., Carter, D., Hamann, M., 2001. The green turtle, Chelonia mydas, in Chikhi, A., Das, H.S., Hamza, S., Kerr, O.J., Al Kiyumi, A., Mobaraki, A., Al Queensland, Australia: the bramble cay rookery in the 1979-1980 breeding season. Suwaidi, H.S., Al Suweidi, A.S., Sawaf, M., Tourenq, C., Williams, J., Willson, A., Chelonian Conserv. Biol. 4, 34–46. 2014. Identification of Important Sea turtle areas (ITAs) for hawksbill turtles in the Mancini, A., Elsadek, I., El-Alwany, M.A.N., 2015. Marine turtles of the Red Sea. In: Arabian region. J. Exp. Mar. Biol. Ecol. 460, 89–99. https://doi.org/10.1016/j. Rasul, N.M.A., Stewart, I.C.F. (Eds.), The Red Sea: The Formation, Morphology, jembe.2014.06.009. Oceanography and Environment of a Young Ocean Basin, Springer Earth System R Core Team, 2020. R: A Language and Environment for Statistical Computing. R Sciences. Springer, Berlin Heidelberg, Berlin, Heidelberg, pp. 551–565. https://doi. Foundation for Statistical Computing, Vienna, Austria. org/10.1007/978-3-662-45201-1_31. Santidrian´ Tomillo, P., Fonseca, L.G., Ward, M., Tankersley, N., Robinson, N.J., Mancini, A., Phillott, A., Rees, A.F., 2019. Chelonia mydas North Indian Ocean Orrego, C.M., Paladino, F.V., Saba, V.S., 2020. The impacts of extreme El Nino˜ subpopulation. In: IUCN Red List of Threatened Species. https://doi.org/10.2305/ events on sea turtle nesting populations. Clim. Chang. 159, 163–176. https://doi. IUCN.UK.2019-2.RLTS.T142121108A154845002.en. org/10.1007/s10584-020-02658-w. Mashat, A., Abdel Basset, H., 2011. Analysis of rainfall over Saudi Arabia. JKAU: Met., Shimada, T., Limpus, C., Jones, R., Hamann, M., 2017. Aligning habitat use with Env. & Arid Land Agric. Sci 22, 59–78.doi:https://doi.org/10.4197/Met. 22–2.4. management zoning to reduce vessel strike of sea turtles. Ocean Coast. Manag. 142, Mazaris, A.D., Kallimanis, A.S., Tzanopoulos, J., Sgardelis, S.P., Pantis, J.D., 2009. Sea 163–172. https://doi.org/10.1016/j.ocecoaman.2017.03.028. surface temperature variations in core foraging grounds drive nesting trends and Shimada, T., Limpus, C.J., Hamann, M., Bell, I., Esteban, N., Groom, R., Hays, G.C., 2020. phenology of loggerhead turtles in the Mediterranean Sea. J. Exp. Mar. Biol. Ecol. Fidelity to foraging sites after long migrations. J. Anim. Ecol. 89, 1008–1016. 379, 23–27. https://doi.org/10.1016/j.jembe.2009.07.026. https://doi.org/10.1111/1365-2656.13157. Mazaris, A.D., Schofield, G., Gkazinou, C., Almpanidou, V., Hays, G.C., 2017. Global sea Shimada, T., Duarte, C.M., Al-Suwailem, A.M., Tanabe, L., Meekan, M.G., 2021. Satellite turtle conservation successes. Sci. Adv. 3, e1600730 https://doi.org/10.1126/ tracking reveals nesting patterns, site fidelity, and potential impacts of warming on sciadv.1600730. major rookeries of green turtles in the Red Sea. Front. Mar. Sci. https://doi.org/ Miller, J.D., 1997. Reproduction in sea turtles, in: Lutz, P.L., Musick, J.A. (Eds.), The 10.3389/fmars.2021.633814. Biology of Sea Turtles, Volume vol. I. CRC Press, Boca Raton, FL, pp. 51–81.doi: Teclemariam, Y., Giotom, M., Mengstu, T., Abraha, H., Mahmud, S., 2009. An update on https://doi.org/10.1201/9780203737088. marine turtles in Eritrea, Red Sea. Indian Ocean Turtle Newsletter 6–10. Moore, R.J., Balzarotti, M.A., 1977. Report of 1976 expedition to Suakin Archipelago Venter, O., Sanderson, E.W., Magrach, A., Allan, J.R., Beher, J., Jones, K.R., Possingham, (Sudanese Red Sea). Results of marine turtle survey and notes on marine and bird H.P., Laurance, W.F., Wood, P., Fekete, B.M., Levy, M.A., Watson, J.E.M., 2016. life. Sixteen years of change in the global terrestrial human footprint and implications for Mortimer, J.A., Camille, J.-C., Boniface, N., 2011. Seasonality and status of nesting biodiversity conservation. Nat. Commun. 7, 12558–12558.doi:https://doi.org/ hawksbill (Eretmochelys imbricata) and green turtles (Chelonia mydas) at D’Arros 10.1038/ncomms12558. Island, Amirantes group, Seychelles. Chelonian Conserv. Biol. 10, 26–33. https:// Visser, M.E., Holleman, L.J.M., Caro, S.P., 2009. Temperature has a causal effect on avian doi.org/10.2744/CCB-0830.1. timing of reproduction. Proc. R. Soc. B Biol. Sci. 276, 2323–2331. https://doi.org/ National Centers for Environmental Information, 2016. GHRSST level 4 AVHRR_OI 10.1098/rspb.2009.0213. global Blended Sea surface temperature analysis (GDS version 2) from NCEI. Ver. 2. Wallace, B.P., DiMatteo, A.D., Hurley, B.J., Finkbeiner, E.M., Bolten, A.B., Chaloupka, M. PO.DAAC, CA, USA [WWW Document]. URLdoi:10.5067/GHAAO-4BC02 (accessed Y., Hutchinson, B.J., Abreu-Grobois, F.A., Amorocho, D., Bjorndal, K.A., Bourjea, J., 7.24.20). Bowen, B.W., Duenas,˜ R.B., Casale, P., Choudhury, B.C., Costa, A., Dutton, P.H., Okuyama, J., Ishii, H., Tanizaki, S., Suzuki, T., Abe, O., Nishizawa, H., Yano, A., Fallabrino, A., Girard, A., Girondot, M., Godfrey, M.H., Hamann, M., Lopez-´ Tsujimura, M., Ishigaki, Takakazu, Ishigaki, Takashi, Kobayashi, M., Yanagida, H., Mendilaharsu, M., Marcovaldi, M.A., Mortimer, J.A., Musick, J.A., Nel, R., Pilcher, 2020. Quarter-century (1993–2018) nesting trends in the peripheral populations of N.J., Seminoff, J.A., Tro¨eng, S., Witherington, B., Mast, R.B., 2010. Regional three sea turtle species at Ishigakijima Island, Japan. ccab 19, 101–110. https://doi. management units for marine turtles: a novel framework for prioritizing org/10.2744/CCB-1428.1. conservation and research across multiple scales. PloS One 5, e15465–e15465.doi: Ormond, R.F.G., Dawson-Sheppard, A., Price, A., Pitts, R.G., 1984. Report on the https://doi.org/10.1371/journal.pone.0015465. Distribution of Habitats and Species in the Saudi Arabian Red Sea. International Wheelwright, N.T., Mauck, R.A., 1998. Philopatry, natal dispersal, and inbreeding Union for Conservation of Nature/Meteorological and Environmental Protection avoidance in an island population of Savannah sparrows. Ecology 79, 755–767. Administration/PERSGA, Kingdom of Saudi Arabia. https://doi.org/10.1890/0012-9658(1998)079[0755:PNDAIA]2.0.CO;2. Pendoley, K., Kamrowski, R.L., 2016. Sea-findingin marine turtle hatchlings: what is an appropriate exclusion zone to limit disruptive impacts of industrial light at night? J. Nat. Conserv. 30, 1–11. https://doi.org/10.1016/j.jnc.2015.12.005.

10