Methods

We conducted the present study between March 2011 and October 2012, covering all management sections except Zayun, where no villages are adjacent to the reserve boundary. To investigate the current diversity and distribution of freshwater turtles, we interviewed collectors of non-timber forest products and ex-hunters living in villages adjacent to YGL. We showed photos of native and nonnative freshwater turtles to interviewees to identify which species were historically and currently found in the area. To verify the identification, we asked the interviewees to describe distinguishing characters of the species. We compiled data on the locality and date of each capture or sighting to document the distribution of each species present in the 6 surveyed management sections.

Based on information from the interviews, we conducted forest and river surveys in each section. For forest surveys, we targeted species that use the terrestrial habitat and forest streams: Cuora galbinifrons, Cuora mouhotii, Cuora trifasiata, Platysternon megacephalum, and Sacalia quadriocellata. We focused our effort in 3 sections that interviews identified as having high turtle diversity: Gaofeng, Nankai, and Fanyang. We installed traps made of wire mesh, measuring 45 × 20 × 15 cm, on the forest floor and placed partially submerged traps on the edge of streams. Traps were baited with salted fish, meat, or cattle viscera and checked at least every 3 d. The trap design and bait choice have proven to be effective throughout southern China (Gong et al. 2005b; Sung et al. 2013).

For river surveys, we targeted species inhabiting deeper water: Pelodiscus sinensis, Palea steindachneri, and Pelochelys cantorii. We focused on 3 sections—Yuanmen, Wanchong, and Maoyang—because of recent sightings of P. steindachneri and P. cantorii revealed in interviews. We installed large, floating cage traps made of wire mesh, measuring 2.5 × 1.5 × 0.7 m, in river regions with deep pools or large streams (cross-sectional width > 10 m; Som et al. 2006). Traps were baited with dead chickens. Because of high levels of human disturbance along large rivers, traps were checked every 4 hrs to prevent vandalism and theft. We used binoculars to see whether turtles were captured in traps. Sampling was conducted during suitable weather (i.e., not during heavy rain), and each sampling sessions was between 4 and 22 consecutive days.

Results

We interviewed a total of 30 people in the 6 management sections of YGL. Eight of the 11 species native to Hainan were reported from YGL (Table 1). Geoemyda spengleri, Mauremys mutica, and Mauremys sinensis were unknown to the interviewees; thus, we believe them to be absent from YGL. Cuora mouhotii and P. megacephalum were the most widely reported species in the 6 sections.

Table 1. Interview data reporting the occurrence of turtle species in the six surveyed management sections of Yinggeling Nature Reserve (P  =  present at site). The conservation status of each species on the IUCN Red List (IUCN 2013) and China Species Red List (Wang and Xie 2004) is included also: CR  =  critically endangered; EN  =  endangered; VU  =  vulnerable.

View Fullscreen

For field surveys, we conducted a total of 7999 and 531 trap days for forest and river surveys, respectively. We captured 18 individuals of 4 species, including C. mouhotii, P. sinensis, P. megacephalum, and S. quadriocellata (Table 2). We do not disclose detailed information on where turtles were captured to ensure the safety of the populations (Stuart et al. 2006).

Table 2. The species captured during field surveys in Yinggeling Nature Reserve, Hainan.

View Fullscreen

Discussion

The results of this study suggested that turtle populations in YGL are highly depleted, possibly attributable to intensive human exploitation. Most interviewees described turtle harvest in the last 20 yrs that caused drastic decline in turtle populations. From forest surveys, we captured 1 turtle for every approximate 666 traps days. By comparison, in Jianfengling Nature Reserve (another reserve in Hainan), it took ~ 200 trap days to capture an individual (Gong et al. 2006a). Despite considerable field effort (over 8000 trap days), we found only 4 species: C. mouhotii, S. quadriocellata, P. megacephalum, and P. sinensis.

Cuora mouhotii and S. quadriocellata were the 2 most common species found in the forest survey—6 individuals per species—yet their densities were low. These are the only 2 species in Hainan with relatively high population densities, allowing for detailed ecological studies (Gong et al. 2005b; Wang et al. 2011a). Cuora mouhotii was and continues to be harvested in large volume for food and pet markets (Cheung and Dudgeon 2006; Gong et al. 2006b). The harvesting pressure for S. quadriocellata for food is relatively low because local people say it tastes bad, yet S. quadriocellata is still threatened by harvesting and is common in pet and food markets in cities (Cheung and Dudgeon 2006; Gong et al 2006b).

Platysternon megacephalum, based on interviews, was common in all 6 sections of YGL, but populations have declined because of intensive hunting targeting this species. The common hunting technique for P. megacephalum is using baited hooks along stream pools. The results of field surveys—only one individual found—support the conclusion that populations in the reserve have been decimated.

Pelodiscus sinensis is the only softshell turtle species that is relatively common in Hainan. It was reported that P. sinensis is found in rivers and irrigation ditches of farms around YGL. In our river surveys, we captured 5 adults and found 2 egg clutches on sandy banks adjacent to agricultural fields. From interviews, the density of P. sinensis has been in marked decline possibly attributable to hunting and habitat destruction; P. sinensis is a common food item, and individuals are eaten whenever they are found. In addition, increasing sand mining activities are reducing habitat quality and availability of nesting sites.

Three Hainan species on the Turtle Conservation Fund Priority Species List, C. trifasciata, C. galbinifrons, and P. cantorii (Turtle Conservation Fund 2014), were not found in our surveys. From interviews, we identified several areas where C. trifasciata was captured in the last 10 yrs. Similar to most areas across its range, the density of wild C. trifasciata is extremely low (Horne et al. 2012). Cuora trifasciata are disappearing from the wild because of tremendous harvesting pressure—from interviews, the price of C. trifasciata in local villages has skyrocketed from around US$2500 10 yrs ago to over US$8000 in the last 3 yrs. The density of C. galbinifrons is also low in YGL, similar to elsewhere in Hainan (Wang et al. 2011b). According to interviews, there were no captures of C. galbinifrons from the reserve in the last 10 yrs; thus, we speculate that the population is very close to extirpation in YGL. There is a chance that P. cantorii still exists in YGL. In 1995, 2000, and 2009, there were reports of P. cantorii captured near the reserve. It is likely that individuals survive in the deepest pools along the rivers, but they are undoubtedly very rare (Gong et al. 2006b). Priority has been placed on conserving these three species, therefore, it is imperative to locate and protect the remaining wild individuals of these species to prevent extirpation.

Of the species on the freshwater turtle and tortoise checklists of Hainan, 4 additional species are not in YGL: M. mutica, M. sinensis, G. spengleri, and Manouria impressa. Both Mauremys species are typically found in rice paddies, marshes, and ponds (Yasukawa et al. 1996; Lovich et al. 2011), and their absence in the mountainous YGL is not surprising. In Hainan, G. spengleri remains a highly restricted species (Shi 2005), yet large numbers have been reported being sold in markets (Gong et al. 2005a). The current status of wild populations remains largely unknown. Also, we paid particular attention to M. impressa during our interview survey because its presence in Hainan is questionable; the only record is based on a specimen purchased from an animal trader in Haikou City (Buskirk 1989). We found no evidence of any large-sized tortoise present at any time in Hainan and similar investigations throughout Hainan over the last 12 yrs yielded similar results (B.P.-L.C., unpubl. data, 2014). Therefore, we believe that including M. impressa on species lists of Hainan is incorrect and that this species should be removed.

Identification of key conservation areas with the highest turtle diversity or abundance is crucial to maximize the effectiveness of patrols and, thus, can be an effective measure to tackle illegal turtle hunting (Sung et al. 2013). In situ conservation actions, including setting up protected area with effective enforcement, have been recommended as crucial measures for conservation of freshwater turtles (Horne et al. 2012). However, antipoaching patrols may be ineffective, because of the limitation of manpower and the large size of protected areas. Since 2006, a team of almost 200 wardens has regularly patrolled YGL against illegal hunting. One job of wardens is to find and destroy turtle traps, and over the years, there seems to be a decreasing trend in the number of turtles traps found. Although promising, we believe it is difficult, if not impossible, to eliminate hunting completely because the reserve is large and incentive to hunt is very high. To safeguard the remaining turtle populations, we believe efforts should be concentrated in areas with relatively high turtle abundance and diversity. From the results of this study, we identified 2 key turtle conservation areas and started an intensive patrol and monitoring plan to halt the illegal trapping activities and aid the recovery of remnant populations.