Population Status and Threats

In May 2004, a working group (WG) convened that included representatives from Papua (Indonesia), PNG, Solomon Islands, and Vanuatu (Kinan 2005). This WG consisted of researchers, managers, and tribal community leaders with extensive local knowledge. The WG reviewed collective knowledge and mapped out nesting sites that were either documented or believed to have more than 20 nests per season. The WG drew on several sources of information from internal reports, gray literature in local languages, field notes, and personal observations to compile a matrix of information on population size and threats (Kinan 2005). Where possible, the WG attempted to estimate the number of nests laid per season at each site. Given the inherent error in these estimates, in addition to the annual variability in numbers of nests laid characteristic of marine turtles, a range was given for numbers of nests laid each year based on observations since 1999. The intent was to provide a minimum estimate for the number of nests laid annually by leatherbacks in the western Pacific as a basis for comparison to the previous estimate reported in Spotila et al. (1996). To make the estimates comparable, we divided the number of nests by 5 to estimate the number of females per year as described in Spotila et al. (1996). This approach makes the assumption that the average number of clutches laid by leatherbacks in the western Pacific is five. However, the number of nests laid per females is unknown for these populations. It is important to emphasize that we are only using this simplistic approach to provide a conservative population estimate for comparative purposes.

Genetic Population Structure

Skin samples were collected from nesting females or salvaged from dead hatchlings from Jamursba-Medi (Wembrack, Warmamedi, and Batu Rumah; Fig. 1), and Wermon in Papua; from Kamiali in PNG; and from Solomon Islands (Sasakolo; Dutton et al. 1999; Fig. 1). Hatchling samples came from nests after emergence, taking care not to sample more than 1 nest per female. Skin samples were preserved in a 20% dimethyl sulphoxide solution saturated with laboratory grade salt, as described in Dutton et al. (1999). DNA was isolated by using either standard phenol/chloroform extraction techniques (Sambrook et al. 1989) or by using the Fast Prep DNA isolation kit (Bio101®). Amplification of mtDNA was performed by polymerase chain reaction (PCR) by using the primers HDCM2 and LTCM2, designed to target 496 bp at the 5′ end of the control region of the mitochondrial genome (see Dutton et al. 1999). Template DNAs were amplified in 50 μL PCR reactions on a Perkin Elmer 480 thermocycler by using the following profile: initial denaturation at 94°C for 2 minutes, followed by 36 cycles of 1) DNA denaturing at 94°C for 50 seconds, 2) primer annealing at 52°C for 2 minutes, and 3) primer extension at 72°C for 1 minute 30 seconds, concluding with a final primer extension for 5 minutes at 72°C. The size of the amplified products were determined by using electrophoresis in a 2% agarose gel stained with ethidium bromide. PCR products were then purified by using the Qiaquick PCR Purification Kit (Qiagen, Valencia, CA) and stored at 4°C. Direct cycle sequencing reactions of the light strand were performed on 2 μL purified PCR product combined with 2 μL ABI Prism dRhodamine Terminator Cycle Sequencing Kit, 3 μL primer LTCM2, and 5 μL purified water. The labeled extension products were then purified via ethanol precipitation and analyzed with an Applied Biosystems model 377 or 310 genetic analyzer. The sequences were edited by using Gene Codes Sequencher 4.1. Haplotype frequencies were compared between the sampled nesting sites at PNG and Papua, and with published data from Solomon Islands and Malaysia (Dutton et al. 1999) by using a χ² test to determine genetic homogeneity. Monte Carlo procedures were used to correct the probabilities obtained by the χ² test for multiple tests and small samples (Roff and Bentzen 1989).

Nesting Distribution and Abundance

Several previously undescribed nesting sites were identified. In most cases, there was local knowledge of leatherback nesting; however, no census work was carried out, and the number of nests could not be estimated (Table 1, Fig. 1). We estimate an approximate total of 5000–9100 leatherback nests are laid each year among all the beaches identified in the western Pacific (Table 1). The rookery at Papua, Indonesia, remains the largest and best studied (Hitipeuw et al. 2007), with 3 beaches at Jamursba-Medi containing the bulk of the nesting. However, there are several sites along the northwest coast of Papua that had historic reports of high-density nesting, where the status is now unclear. An aerial survey conducted in July 2005 only detected relatively low densities of nests along the stretch of coastline northwestward from Jamursba-Medi to Sorong (S.R. Benson, unpubl. data). In addition, Wermon appears to have much greater nesting activity than previously thought (Hitipeuw et al. 2007). Little is known about the vast, isolated coastline stretching southeastwards from these beaches to the border of PNG (Fig. 1), thereby underscoring the need for aerial surveys to identify unknown leatherback nesting sites along the coastline from the Bird's Head (Vogelkop) Peninsula of Papua to the PNG border. In PNG, there is scattered nesting along the north coast and around the Island of New Britain and Bougainville (Table 1; Fig. 1). This has been confirmed by recent aerial surveys (Benson et al. 2007) that identified Buang-Buasi and Kamiali as 2 areas of higher-density nesting. We recommend establishing these 2 areas as index sites for long-term monitoring to determine nesting abundance trends in PNG. The other sites listed for PNG in Table 1 can essentially be considered beaches that are used by the same nesting population that nests in the Huon Gulf. The estimates of total nests laid annually at all the sites in the Huon Gulf range from 500 to 1150 (Table 1). This range reflects the annual variability in nests and is based on preliminary data from 3 years of aerial surveys (S.R. Benson and V. Rei unpubl. data). There was sporadic monitoring of nesting within a 4-km stretch of beach demarcated at the Kamiali Wildlife Management Area (KWMA), and 40–89 nesters were tagged annually between 1999 and 2005 (Kisakao 2004; Kinan 2005; Benson et al. 2007; N. Pilcher, unpubl. data). Nest counts from this effort underestimate the total nesting activity for PNG and the Huon Gulf, and we consider the numbers provided in Table 1 from the aerial surveys as the most reliable estimates currently available. More recently, beach monitoring was initiated at the sites in Labu-Tale, Buang-Buassi, and Paiawa (Table 1; N. Pilcher, pers. comm.), and it is hoped that this expanded monitoring will provide more accurate estimates of annual abundance in the future.

Table 1. Western Pacific leatherback (Dermochelys coriacea) nesting sites identified as having more than 20 nests annually.^a

View Fullscreen

Satellite telemetry suggests that nesters tagged at Kamiali tend to lay subsequent clutches in the same season at other sites, as well as within and adjacent to the KWMA. There is also an indication of movement by nesters between Kamiali and Bougainville (Benson et al. 2007). However, Bougainville is closer to the Solomon Islands than the Huon Gulf, and it is possible that some leatherbacks nesting in the Solomon Islands tend to also lay clutches on Bougainville. This would account for the variability in nest counts reported for Bougainville (Table 1), however, the extent of within-season movement between nesting sites is unknown. The Solomon Islands are more important than previously thought, with scattered nesting reported from several sites in the western province of Isabel (Table 1; Fig. 1). Surveys have been incomplete, and it is likely that nesting activity described in Table 1 underestimates the true population size. Vaughan (1981) reported leatherback nesting at 61 beaches in Isabel and the western province but many with just a few scattered nests. In 1996, 40 leatherbacks were sampled over a 7-day period for the genetic study at Sasakolo (Dutton et al. 1999; D. Broderick, pers. comm.). Given the scattered nesting around the islands in the region, it is possible that the size of this nesting population is on the order of hundreds, rather than tens of females. Extensive aerial surveys in conjunction with complete monitoring and tagging at multiple sites should be a priority for a more accurate assessment of leatherback population abundance in the region.

By using the numbers given in Table 1, we estimated a total of 1113 females nesting annually (FNA) following methods reported in Spotila et al. (1996). This number might be larger, because there are still areas where undocumented nesting occurs throughout the Island of New Guinea and beyond, such as in Thailand and Vietnam. This minimum estimate is larger than that of Spotila et al. (1996) who estimated 700 females for the western Pacific rookeries. Use of these new estimates would have produced a regional population estimate of 2782 breeding females in the western Pacific by applying the same simplified methods Spotila et al. (1996) used (multiplying FNA by 2.5) to derive their 1996 estimates for the regional populations. This is likely a conservative estimate and depends on the assumption that the average number of nests laid per female is 5 (Spotila et al. 1996). If leatherbacks lay fewer nests on average then the estimated number of females derived from the nest counts will be greater and vice versa, so that estimates would range from 2100–5700 females based on estimates ranging from ca. 840-3200 FNA (Table 2). This illustrates the problem of drawing conclusions on population status from estimates of numbers of nesters derived from nest counts and the importance of further research to determine other demographic parameters, such as numbers of nests laid per female and the extent of movement between nesting beaches. Given the high level of uncertainty, we suggest evaluating the population size in terms of numbers of nests, rather than attempting to derive numbers of females.

Table 2. Estimates of minimum females nesting annually (bold) and total number breeding females (italics)^a in the western Pacific leatherback metapopulation by assuming 4, 5, or 6 nests laid on average by each nester per season. Estimates based on upper and lower range of nests given in Table 1.^b

View Fullscreen

Genetic Population Structure

A total of 6 haplotypes were identified among the 106 samples analyzed for Solomon Islands, Papua, and PNG (Table 3). All of these haplotypes were previously identified and described in Dutton et al. (1999). Haplotype I was the most common haplotype in all the rookeries sampled. There was no significant difference in haplotype frequencies among all 4 populations sampled (χ² = 7.363, p = 0.972). Given this lack of differentiation, further statistical analysis was not pursued. The lack of genetic differentiation between Papua, PNG, and Solomon Islands leatherback rookeries could indicate ongoing gene flow among this metapopulation or could also be the result of the inability of these mtDNA markers to resolve fine-scale population subdivisions (Dutton et al. 1999). Further analysis by using multiple nuclear markers (microsatellites) may help resolve this (Dutton et al. 2000). The demographic connection between PNG and Solomon Islands is plausible, given recent telemetry data that indicate leatherbacks tagged at Kamiali (PNG) may also nest on Bougainville during the same season. Further tagging and telemetry studies will help resolve this.

Table 3. mtDNA haplotype frequencies at major rookeries in the western and Indo-Pacific.^a

View Fullscreen

The common “I” haplotype for the western Pacific rookeries in our survey was not found in the eastern Pacific stock (Costa Rica and Mexico; Dutton et al. 1999), nor was it found in the limited sample from Malaysia (Dutton et al. 1999; Table 3). This haplotype, therefore, can be used to identify individuals of western Pacific stock origin encountered at foraging areas and in fisheries by-catch (Dutton et al. 2000, 2006). The published haplotype frequencies for Terengganu, Malaysia, were significantly different from the 4 western Pacific rookeries from this study (χ² = 49.346, p =  0.002), indicating that this Indo-Pacific stock was distinct from the western Pacific stock that consisted of the rookeries reported in this study.