Conservation Needs of the Critically Endangered Philippine Forest Turtle, Siebenrockiella leytensis, in Palawan, Philippines
Abstract
The Philippine forest turtle, Siebenrockiella leytensis, is endemic to Palawan and Dumaran islands. Its limited distribution, exploitation, and habitat destruction are the main threats to this Critically Endangered species. The populations of 5 sites in northern Palawan were assessed in terms of habitat, population size, density, and structure. Fieldwork was conducted from 28 January to 15 June 2007. Turtles were collected through visual encounters and with pit fall and baited funnel traps. All captured individuals were marked and released after standard measurements were taken. Considering the relatively short sample period, populations were considered “closed,” that is, with negligible birth, death, and migration. Population size was then estimated using statistical analysis. Information on exploitation was gathered through interviews. Surveys revealed various degrees of habitat destruction and different levels of exploitation among sites. Population size estimates varied from 10 to 110 individuals per site, and in the majority of sites, adults were most prevalent. Densities ranged from 4.4 to 121.7 (mean 39.3 ± 49.5) individuals/ha. The highest density and largest population were found at the least disturbed site. Low densities were recorded in the more disturbed habitats, presumably as a result of exploitation. In 3 of the 5 sites turtles are heavily exploited either for local food consumption or for the international pet trade. Findings justify its current status as Critically Endangered. In situ conservation of the most pristine population (Site I) is recommended in line with further research on the biology, ecology, and distribution pattern of the species.
The Philippine forest turtle, Siebenrockiella leytensis—previously known under its synonym Heosemys leytensis—is a Critically Endangered freshwater turtle species (IUCN 2010) that is endemic to Palawan and Dumaran islands (Diesmos et al. 2004b), and most probably restricted to central and northern parts of the province, delimiting its range to roughly 7000 km2. It is one of the key conservation species in Palawan (Anda and Tabangay-Baldera 2004). Internationally, its trade is regulated under Appendix II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES 2010). The Philippine Wildlife Act 9147 disallows the collection and trade of the species (Republic of the Philippines 2001); hence, the many specimens now known in collections outside Palawan—except those in authorized rescue centers—are illegally obtained.
Threats have been identified in several recent studies, and exploitation and habitat destruction rank first among them (Diesmos et al. 2004a, 2004b; Gavino and Schoppe 2004; Lopez and Schoppe 2004; Acosta 2006; IUCN 2010; Diesmos et al. 2008; Schoppe and Cervancia 2009). Threats seem to arise and worsen faster than knowledge on the biology and ecology of this species accumulates. Knowledge is restricted to distribution; some general descriptions of the habitat, feeding habits in the wild, and captive care; phylogeny; morphology; and first data on growth (Taylor 1920; Buskirk 1989; Diesmos et al. 2004a, 2004b, 2005; Fidenci 2004; Gavino and Schoppe 2004; Lopez and Schoppe 2004; Yuyek 2004; Acosta 2006; Schoppe 2008; Schoppe and Cervancia 2009). Virtually nothing is known about demographic trends, size and age at maturity, reproduction, surviving numbers in the natural populations, population sizes, and densities.
In situ conservation is believed to be the priority measure for S. leytensis at this point of time (Schoppe and Cervancia 2009). The main objective of this study therefore was to identify populations that have special conservation needs or potential. In view of the insufficient knowledge on the biology and ecology of the species it was further intended to assess populations of S. leytensis in terms of size, density, and composition.
METHODS
Habitat
Populations of S. leytensis were assessed at 5 sites (Sites I to V) of 4 municipalities (Puerto Princesa City, Roxas, Dumaran, and Taytay) in Palawan, Philippines (Fig. 1). Palawan is an island archipelago situated between Borneo and oceanic Philippines and separates the South China Sea from the Sulu Sea. A central mountain range forms the backbone of Palawan Island. Alluvial plains are mostly narrow with the exception of the southeastern portion (Widmann 1998). Because of the relatively short watersheds, only few permanent water flows exist; smaller rivers and rivulets dry up completely during the dry season. The climate in northern Palawan is classified under Type I with a dry season from December to May and a wet season from June to November (PAGASA 2001). The rest of Palawan falls under Type II with no pronounced maximum rainfall period but a short dry season of 1–3 months (PAGASA 2001). All sites are along small to medium size forest streams (Fig. 2a–e). The sizes of the survey sites depended very much on the topography, accessibility, and stream flow, and ranged from 0.7 to 3.2 ha. At each survey site, the length, width, depth, and current flow of the stream were measured, and the canopy cover was assessed at the start, the center, and the end of the survey area. Water and air temperature were measured daily between 0600 and 0700 hours. The dominant substrate types of the streams, the vegetation in the immediate vicinity of the streams, and human practices in or along the streams are described.
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0783.1
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0783.1
Population Size Surveys
Mark–recapture surveys following Schnabel's method (Krebs 1998) were conducted from 28 January to 15 June 2007. Turtles were trapped with baited funnel traps and pit fall traps and hand-captured during visual encounter surveys. Visual encounter surveys were conducted in the early morning and early evening hours. They covered the water bodies and 15 m right and left of the stream. Depending on the topography of the study site, 7–13 pit fall traps, and 10–22 funnel traps were used. Funnel traps were baited with bananas, fish, or chicken necks and set in the water bodies. Bait was replaced every 2–3 days after the morning sampling. Each captured individual was marked through carapace notching, measured, weighed, photographed, described, and released to place of capture. Surveys per site lasted between 4 and 6 weeks and were only terminated if no new individuals were captured. Considering the short survey period, populations were treated as “closed,” and data were analyzed according to the methods of Schumacher and Eschmeyer (Krebs 1998) to estimate population sizes. A closed population does not change in size during the study period: that is, the effect of births, deaths, and migration are negligible. Thus, populations are typically closed over only a short period of time, and the assumption of dealing with a constant population size is justifiable considering that the study was conducted over a short period of time. Results among survey sites were made comparable by calculating the estimated density as the number of individuals per 100 m2 (and per hectare) of the study site, where the size of the study site was computed as the surveyed stream length × mean width + 15 m right and 15 m left of the stream.
Population Structure
The composition of populations was analyzed in terms of the size of individuals and in terms of the relative composition of the different life history stages/size classes. Standard measurements were taken and of these, only median carapace length (MeCL) is discussed here. Size measurements up to 150 mm were taken with vernier calipers (to the nearest 0.1 mm) and then with tree calipers (to the nearest 1 mm). The classification of the life history stages relied on growth studies and on observations of reproductive behavior in captivity (Acosta 2006; Schoppe 2008; S. Schoppe 2008, unpubl. data). Accordingly, hatchlings (MeCL of ≤ 46 mm) were defined as those juveniles that still showed “birthmarks” such as presence of an egg tooth and/or umbilical region distinct from the rest of the plastron. Juveniles (47–170 mm MeCL) were immature individuals that did not show external sexual dimorphism (Figs. 3, 4), and adults (MeCL ≥ 170 mm) were sexually dimorphic individuals. However, sexual dimorphism is not very conspicuous in the S. leytensis, and sexes are difficult to tell apart for the inexperienced fieldworker, hence data were combined.
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0783.1
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0783.1
Distribution, Local Use, and Trade
Interviews with local residents of 18 municipalities (all mainland plus the island municipalities of Busuanga, Balabac, Coron, Culion, and Dumaran) were conducted to answer 2 main questions: 1) whether S. leytensis occurs in the area, and 2) if the species is collected for local use or trade. Information regarding the occurrence of the species was validated through visual encounter surveys and trapping surveys. The interviews and validation surveys were conducted in the second half of 2006. At the same time pet market surveys were conducted in Manila.
RESULTS AND DISCUSSION
Habitat
All 5 survey areas shared the characteristic of perennial streams (Table 1). By far the densest vegetation and canopy cover were found in Site I. In all 5 study sites, other freshwater turtle species (Cuora amboinensis and/or Cyclemys dentata) were found to share the habitat with S. leytensis to various degrees. Our findings suggest that water temperature and stream substrate are the primary factors delimiting the distribution of S. leytensis.
Population Size and Density
We found considerable variation in the number of individuals encountered per site. The total number of individuals caught per site ranged from 10 to 112 individuals. The calculated population size estimates were with 10, 24, 36, 39, and 110 individuals—very similar to the actual number of individuals encountered (Table 2). The highest number of individuals was found in Site I and subsequently decreased from Site II to V (Table 2). Density of S. leytensis ranged from 4.4 to 121.7 turtles/ha (mean 39.3 ± 49.5 turtles/ha) and was highest in Site I followed by Sites III, II, IV, and V (Table 2). No information is available regarding the minimum population density of S. leytensis needed for successful reproduction. At this time we cannot conclude whether the population densities encountered are low or high, enough to sustain the population, or representative of viable populations. The only other population size and density data for S. leytensis are from a 4-week mark–recapture survey conducted at about 2 km from our Site II in 2005. That survey found a total of 22 individuals, estimated a population size of 21.8 individuals, and calculated a density 22 turtles/ha (Acosta 2006). No other comparable data have been published. We believe that at this time it is important to monitor population trends, and we expect that a long-term study that assesses trends over 5 consecutive years (Schoppe 2009) will contribute significantly to an eventual updated IUCN status assessment.
Population Structure
Turtles at Site I obtained the largest mean medium carapace length (MeCL 195.6 mm) among the 5 sites (Table 2). The largest individual had a MeCL of 298.0 mm, which is close to the maximum size record of 302 mm by Diesmos et al. (2004a), and weighed 3220 g. The second largest mean MeCL was encountered at Site V where no small individuals were found at all (Table 2). The smallest individuals including hatchlings were found in Site I and III. In 4 of the 5 sites adult individuals were much more prevalent, ranging from 65% to 82% of the population (Table 2). Only in Site IV did immature individuals outnumber adults (60%). Only one set of comparative data is available. In 2005, Acosta documented a population with 54.5% immature to 45.5% adult individuals (Acosta 2006). Applying general knowledge on the traits of turtles we would expect a dominance of immature over mature individuals because turtles have a long reproductive life span, delayed sexual maturity, and high adult and low egg and hatchling survivorship. According to Congdon et al. (1993, 1994) these traits result in a stable age structure that should be dominated by juveniles and subadults. For that reason, the ratio of nonreproductive to reproductive individuals should greatly exceed one, and adults should represent only a small fraction of the population (Gibbs and Amato 2000).
Considering the small sample size and the little knowledge we have of the biology of S. leytensis, it might be possible that life history stages were not correctly assigned, traps were selective, or juveniles are more secretive, causing the apparent dominance of adults. These remaining questions need to be answered with additional research on the species. This is of special importance considering that “….any removal of reproductive adults, and to a lesser extent of juveniles, from one or a few small populations of animals with a limited annual reproductive potential and late maturity, would have significant effects on the population structure, recruitment and population genetics” (Anonymous 2002, p. 19). For example, Reed et al. (2002) showed that an annual increase in adult mortality in female alligator snapping turtles, Macrochelys temminckii, by less than 1% would result in population declines. Similarly, Congdon et al. (1994) found that common snapping turtle, Chelydra serpentina, populations would be negatively affected by an annual harvest of 10% of the adult females. Gamble and Simon (2003) found that painted turtles, Chrysemys picta, are susceptible to overharvest with an annual removal of only 4%–5% of the female population. We assume that S. leytensis is similarly affected by harvest and agree with Heppell (1998) that conservation efforts that reduce mortality and/or extraction of adult freshwater turtles are likely to stabilize declining populations.
Distribution, Local Use, and Trade
Interviews with 1026 local residents of 18 municipalities—the information of which had been validated—confirmed the occurrence of S. leytensis in Puerto Princesa City, Dumaran, San Vicente, Roxas, and Taytay. Locals from the northern mainland municipality of El Nido; the northern island municipalities of Coron, Culion, and Busuanga; and the municipalities of Aborlan, Narra, Quezon, Rizal, Bataraza, and Balabac in southern Palawan claimed that the species is present in their areas as well. Field validation surveys by Matillano (2008) in the north and Fidenci (2007) in the south could, however, not confirm the presence of S. leytensis in those areas (and only definitively confirmed the presence of C. dentata). We suspect that the practice of showing photographs to nonspecialist residents to confirm the presence of the 2 morphologically similar species (S. leytensis and C. dentata) is to blame for the suspect reports of S. leytensis in southern Palawan (see also Diesmos 2004b). This would confirm the earlier assumption (Schoppe and Cervancia 2009) that the species is restricted to central and northern Palawan. However, the species is likely to occur on other islands in the northern part of the province, such as Busuanga, Coron, Culion, and Linapacan of the Calamianes island group (Diesmos et al. 2004b).
Our survey revealed that S. leytensis is collected for local consumption to different extents in 3 of the 5 sites. In one site the species is also traded and the price for juveniles ranged from US$0.50–1.00 (Php 25–50.00) per individual while adults were sold by collectors for US$1–2 (Php 50–100.00) per individual in 2007. The price has been stable since 2005, but in 2005 mainly adult S. leytensis were traded while smaller sizes have been in demand since 2006.
Among the municipalities with confirmed occurrence of S. leytensis, our surveys recorded trade in the species in Taytay and Dumaran. Furthermore, we noticed trade activity in El Nido. Several middlemen had been identified in Taytay that supplied large volumes of S. leytensis to a Taiwanese trader in 2006. Yet another middleman used to transport the turtles to southern Palawan and exported them from there. In the municipality of Dumaran, the trade of S. leytensis was run by a local resident of Taytay as well. We also received a report that there was at least one foreign trader who shipped specimens abroad in 2006. Usually the turtles are transported together with other wildlife species using local public transportation. Identified exit points for S. leytensis are Puerto Princesa City, Liminangcong in Taytay, Taytay seaport, and the ports of Araceli and Dumaran. Based on our information, traded animals are then delivered to Manila, Cebu, Iloilo, and Zamboanga. In Manila they are traded as pets; our survey revealed that S. leytensis was sold for US$15.90–19.90 (Php 800.00–1000.00) per individual in 2006. But Manila, Cebu, Zambuanga, and Davao also serve as trans-shipment points to Asia and other regions of the world (Diesmos et al. 2004b). Pet trade activity has been confirmed by Diesmos et al. (2004b) for Europe and Japan. A pet shop survey in Japan in 2007 found the species in 2 of 40 shops, where it was sold for US$1624 (M. Auliya, Traffic Southeast Asia, pers. comm., April 2008). In Malaysia specimens were made available on an order basis, and it took about 2 days to deliver (Anonymous trader, pers. comm., October 2006). We assume that turtles are shipped together with other wildlife to Kudat and/or Sandakan. This trade route was identified for the Palawan pangolin as well as for the Palawan bird trade (Cruz et al. 2007; Schoppe and Cruz 2008).
CONCLUSIONS
Turtle population density was considerably lower in those sites where the species is exploited. Populations under the pressure from subsistence collection (Site II) and those harvested for the international pet trade (Site IV) had similarly low densities of 12 and 8.4 turtles/ha, respectively. We conclude that among the 2 main threats to the survival of S. leytensis, exploitation is more detrimental than habitat destruction. The combination of both threats, however, will undoubtedly lead to local extinction.
Knowledge of the specific habitat requirements of S. leytensis is still insufficient, and with the current understanding we conclude that the species prefers shaded shallow to deep cool forest streams that are slow flowing with soft bottoms and banks.
Siebenrockiella leytensis can be locally abundant. However, due to its very limited geographic range, ongoing exploitation, habitat destruction, and lack of information on population demographic and trend data we consider its current Critically Endangered IUCN Red List status justifiable.
RECOMMENDATIONS
Among the 5 sites, conservation of Site I within the buffer zone of the Puerto Princesa Subterranean National Park should be given priority. Here, the relatively pristine habitat and healthy population should be protected in situ. Since the site is already part of a nationally protected area, emphasis should be given to proper buffer zone management. Immediate conservation actions should focus on information education campaigns and informing local communities in the area about the existence, importance, and conservation needs of the species. A RARE Pride campaign that has been used successfully to change awareness and behavior at key levels of a community, to reduce threats to biodiversity, to create more effective conservation programs, to inform and inspire local constituencies, and to support long-term conservation for other focal species in the Philippines and elsewhere, could be a powerful tool in the conservation of the Philippine forest turtle.
In line with conservation measures there should also be research on the biology and ecology of the species to help better understand its management needs. Of utmost importance are habitat studies that will help us understand what delimits the distribution of the species to northern Palawan: what are its delimiting factors in the selection of habitat, and what determines its population density? Studies on home range, dietary requirements, and reproduction, especially age at maturity, location of nests, clutch size, reproductive season, egg and hatchling survival, and growth, are of major concern.
Of great interest and concern is answering the question of how many S. leytensis are left in the wild. To estimate the size of the remaining total population of S. leytensis in the wild, the quality and extend of available habitat needs to be assessed.
To determine trends in the conservation status of the species, populations should be continuously monitored, and exploitation and trade quantified. Information on illegal wildlife trade and environmental laws protecting the species should be broadcasted.
Priority research topics such as population assessments, demography, breeding biology and seasonality, assessment of the impacts of habitat degradation and exploitation, and additional field surveys to determine the extent of its distribution within the Palawan region (Diesmos et al. 2008) remain valid and should be urgently addressed.
Map of Palawan showing the municipal boundaries and distribution of S. leytensis. Blue dots indicate municipalities where S. leytensis has been confirmed to occur.
Habitat in the 5 different sites. Siebenrockiella leytensis occurs even in altered and disturbed parts of streams. a. Site I in Puerto Princesa City. b. Site II in Roxas. c. Site III in Dumaran. d and e. Site IV and V in Taytay.
Small juvenile S. leytensis.
Late juvenile S. leytensis in habitat.
