Study Area

The Chin Hills are part of a contiguous chain of mountains linking the Rakhine Hills in southwestern Myanmar to the Himalaya Mountains in the north (Platt et al. 2012). These mountains are characterized by extremely rugged topography consisting of high ridges separated by deep, restricted stream valleys; plant communities correspond closely to elevational gradients, slope, and aspect, and consist of tropical evergreen, semievergreen, subtropical, and temperate forests, and extensive bamboo brakes (Salter 1983; Sayer 1983; Platt et al. 2012, 2013b). KPTWS (133 km²) is located in Paletwa Township of southern Chin State and encompasses Kyauk Pan Taung, an isolated massif rising to 1310 m (Beffasti and Galanti 2011). The Chin Hills are populated by the Chin, a Tibeto-Burman linguistic group of shifting agriculturalists (Diran 2001). The physical and human environment of the Chin Hills is described in greater detail elsewhere (Carey and Tuck 1896; Thom 1906; Sayer 1983; Saul 2005).

Methods

We conducted fieldwork in the southern Chin Hills from 2 to 25 November 2012, a period coinciding with the end of the annual wet season. We traveled by road to Matupi, walked overland to Sami, traveled by boat to Mi Gyaung Ba Wa, and from there walked into KPTWS; we later returned along a somewhat different route (Fig. 1). We stopped at each village along our route and conducted semidirected, opened-ended interviews (Martin 1995; Gilchrist et al. 2005) of hunters, farmers, and other knowledgeable individuals regarding the local occurrence and natural history of turtles, collecting methods, and levels of exploitation. While recognizing the limits of traditional ecological knowledge (Huntington 2000), our experience (Platt et al. 2004, 2005, 2007) and that of others (Thirakhupt and van Dijk 1995; Kuchling et al. 2004; Zhou et al. 2008) indicates such individuals are reliable sources of information on local chelonians and their natural history. In accordance with the format of an open-ended interview, we asked each informant a series of questions that included standard questions prepared in advance and others that arose during the course of conversation. Open-ended interviews are more of a conversation than a typical question-and-answer session, and rather than rigidly adhering to a set of prepared questions, the interview provides an opportunity for collecting and discussing unanticipated information (Huntington 2000; Gilchrist et al. 2005). Interviews were conducted by native Burmese speakers and later translated into English. We also asked to examine any living turtles or shells that might be available in villages. We measured plastron length (PL; measured from base of anal notch) with tree calipers and photographed each specimen. Morphometric measurements are presented as mean ± 1 standard deviation (SD). Sex was determined on the basis of plastron morphology; adult male H. depressa exhibit a plastral concavity, which is absent in females (Iverson and McCord 1997). Place names are in accordance with a national gazetteer currently being compiled. We determined geographic coordinates (India-Bangladesh Datum) and elevation with a Garmin® GPS12 (Table 1). Interview transcripts, voucher photographs of specimens, and field notes are archived in the Campbell Museum, Clemson University, Clemson, South Carolina.

Table 1. Geographic coordinates and elevation of locations mentioned in the text.

View Fullscreen

Results and Discussion

We obtained 10 plastrons of H. depressa from residents of Par Maung (7), Wet Ma (1), and Mi Gyaung Ba Wa (2) villages in southern Chin State (Fig. 1). These turtles were reportedly harvested within walking distance (2–5 km) of the villages where we obtained them. Our sample consisted of 4 males, 4 females, and 2 juveniles. Mean adult PL was 203 ± 20 mm SD (range  =  168–224 mm); females (PL  =  217 ± 8 mm SD) were considerably larger than males (PL  =  188 ± 17 mm SD), although our small sample precluded meaningful statistical comparisons. The plastrons of 2 juveniles measured 109 and 153 mm; neither could be reliably sexed. These records from southern Chin State extend the known distribution of H. depressa approximately 225 km north from previously reported populations near An in central Rakhine State (Fig. 1; Platt et al. 2003b). Moreover, our records from Chin State are only the third group of H. depressa specimens for which specific locality data are available (see also Platt et al. 2003b, 2010). Plastrons obtained from Mi Gyaung Ba Wa unequivocally establish the occurrence of H. depressa in KPTWS.

The elevation of villages where we obtained H. depressa plastrons ranged from 21 to 490 m. In addition to the specimens we examined, one hunter from Mi Gyaung Ba Wa found the remains of a female H. depressa at a well-known resting place along the trail to the summit of Kyauk Pan Taung, where we later determined the elevation to be 700 m. We found nothing to suggest H. depressa occurs at elevations > 700 m in the Chin Hills. Most tellingly, villagers living at higher elevations were unfamiliar with this turtle. Elsewhere, H. depressa has been found at elevations of 25–150 m and 139–251 m in central (Platt et al. 2003b) and southern (Platt et al., 2010) Rakhine State.

Villagers in the Chin Hills usually find H. depressa while searching for elephant yams (Amorphophallus paeoniifolius), a perennial understory plant with a commercially valuable tuber that is widely harvested. When found, H. depressa are typically concealed among leaves and other ground debris in what are best described as “forms.” Villagers consider H. depressa uncommon and difficult to find in comparison to yellow tortoises (Indotestudo elongata). Whether this reflects actual biological rarity or cryptic behavior remains unknown, although we suspect the latter based on our observations of captive H. depressa, which spend extended periods buried beneath leaf litter without moving. Villagers reported encountering H. depressa in secondary and primary forests, and dense brakes of regenerating bamboo. Furthermore, interviewees often mentioned H. depressa in association with steep terrain, consistent with our observations of turtles climbing slopes of up to 60° in RYER (Platt et al. 2010). However, interviewees did not regard H. depressa as being restricted to steep slopes in the Chin Hills.

Our interviewees reported observing H. depressa consuming shoots and leaves of A. paeoniifolius, shoots of Melocanna baccifera, and fruits of Dillenia pentagyna; the latter 2 items have previously been recovered from the feces of wild H. depressa (Platt et al. 2010). Interviewees also reported occasional captures of H. depressa in pitfall traps set for I. elongata beneath fruiting D. pentagyna. Similarly, hunters in southern Rakhine State often search for H. depressa beneath fruiting D. pentagyna, and we found D. pentagyna fruits and seeds in the feces of both H. depressa and I. elongata (Platt and Khin Myo Myo 2009; Platt et al. 2010). Collectively these observations suggest H. depressa and I. elongata may function as dispersal agents for D. pentagyna, although chelonian saurochory has rarely been investigated in Southeast Asia (Corelett 1998). Interviewees provided Chin names for a number of additional plants (primarily fruits) said to be consumed by H. depressa, but we were unable to equate these vernacular names with either Burmese common names or scientifically recognized taxa. Nonetheless, our interview data from the Chin Hills and central Rakhine State (Platt et al. 2003b), field observations of wild turtles in RYER (Platt et al. 2010), and other reports (Iverson and McCord 1997) strongly suggest the diet of H. depressa consists largely of vegetation and fruit.

A dead female H. depressa found by a hunter in November 2011 reportedly contained 2 intact eggshells. This observation is significant because almost nothing is known regarding reproduction in the wild, and complements reports from other hunters who found clutches of 2–3 eggs when butchering female turtles (Platt et al. 2003b, 2010). Female H. depressa have deposited somewhat larger clutches of 2–6 eggs in captivity (L. Wyrwich and D. Lawson, pers. comm., January 2013). Although the reproductive phenology of wild H. depressa remains unclear, oviductal eggs present in females during June–July (Platt et al. 2003b, 2010), shelled eggs found in November (this study), an incubation period of 110–160 d in captivity (L. Wyrich and D. Lawson, pers. comm., January 2013), and recent hatchlings found aboveground in May (Platt et al. 2013a), lend additional support to our earlier suggestion (Platt et al. 2010) that eggs are deposited late in the wet season (September–November), incubate through the dry season (perhaps with a cool season diapause), and hatch at the onset of the following wet season (May–June). A near-identical reproductive phenology has been reported for I. elongata in western Thailand (van Dijk 1998).

Chin folk taxonomy with regard to H. depressa is confusing, in large part due to the diversity of local dialects, which frequently vary, even between neighboring villages. Among some groups, H. depressa is referred to as lephar, which translates literally as “flattened turtle,” an obvious reference to its depressiform profile. Some interviewees also know Cyclemys fusca as lephar, although this species is usually identified as the “flattened turtle that dwells in water” in contrast to H. depressa, which is described as being wholly terrestrial. In other villages, I. elongata and H. depressa are together known as i'an talee ( =  yellow turtle) owing to the brown-yellow coloration of the carapace in both species. Anyeah ( =  zinbyun turtle) is another local name applied to both H. depressa and I. elongata as these turtles are frequently encountered beneath fruiting zinbyun (D. pentagyna) trees. Although H. depressa and I. elongata are sometimes classified as the same folk taxa, our informants nonetheless recognized morphological differences between the two, particularly the relative thickness of the gular scute and overall body form. Some interviewees considered H. depressa and I. elongata to be females and males, respectively, of the same folk species based on body form; the depressed carapace of H. depressa is thought to “fit” under the plastron of I. elongata during copulation. Underdifferentiation (i.e., a single folk name applied to 2 or more biologically recognized species) is common in folk taxonomies when particular species have minimal significance in local cultures (Berlin et al. 1966). Vernacular names applied to H. depressa in Rakhine State include pyant chee zar leik ( =  turtle that eats rhinoceros feces; Platt et al. 2010) and leik pyin ( =  lazy turtle; Platt et al. 2003b). However, our interviewees were unfamiliar with these names and neither is applied to H. depressa in the southern Chin Hills.

Prior to our recent survey of the southern Chin Hills and KPTWS, RYER was the only protected area in Myanmar known to harbor H. depressa (Platt et al. 2010). Because of its large size (1756 km²), lack of road access, remoteness, intact forests, and minimal human disturbance (Platt et al. 2010), RYER will likely remain the focus for H. depressa conservation. However, the global significance of KPTWS for H. depressa conservation must not be overlooked. KPTWS and RYER are currently the only 2 existing protected areas within the known distribution of H. depressa (Beffasti and Galanti 2011). Despite the large number of people that visit Kyauk Pan Taung, human activity is largely confined to religious sites near the summit (Platt et al. 2013b), well beyond the apparent elevational limit of H. depressa. Consequently, bamboo and evergreen forests on the lower slopes of the mountain appear little impacted by human activities. Most importantly, inhabitants of villages adjacent to the sanctuary are not involved in commercial turtle harvesting, and recognize the importance of biodiversity conservation in promoting the mountain as an ecotourist destination (Platt et al. 2013b).

In conclusion, while our survey has provided further resolution of the geographic distribution of H. depressa in western Myanmar, much remains to be learned. For example, does H. depressa occur on eastern rain-shadow slopes of the Rakhine and Chin hills? To date, H. depressa has only been found on the western, windward side of these mountains. Due to rain shadow effects, habitat differences exist between the eastern and western slopes, with the former supporting a more xeric forest type, which may not be suitable for H. depressa. Habitat considerations aside, do high mountain ridges effectively block the eastward dispersal of H. depressa? Additionally, is H. depressa truly endemic to Myanmar? Heosemys depressa is present at Kyauk Pan Taung in the Kaladan River drainage and low-elevation habitat extends northward into southern Mizoram (India) with no apparent geographic barrier to dispersal as indicated by the occurrence of Amyda cartilaginea in this region (Pawar and Choudhury 2000). Likewise, seemingly suitable H. depressa habitat is found in adjacent regions of Chittagong (Bangladesh). Such questions have obvious conservation implications for this enigmatic chelonian and can only be answered by further fieldwork.