1 INTRODUCTION

The Testudines, that is, turtles and tortoises, are one of the most threatened groups of vertebrates (Buhlmann et al., 2009) due to over-exploitation, unregulated trade, habitat loss, and degradation (Van Dijk et al., 2000; Gibbon et al., 2000; Turtle Conservation Fund, 2002). Myanmar, a country recognized as a biodiversity hotspot, is home to some of the rarest tropical flora and fauna, including more than 37 species of Chelonians (sea turtles, tortoises, softshell turtles, and terrapins) (Uetz et al., 2023). The majority of the Testudines (at least eight) (Kutchling et al., 2004; Myanmar Conservation and Development Program MCDP, 2012; Shwe & Grindley, 2012) are found in Indawgyi Lake, Myanmar's largest freshwater lake; and most are classified either as endangered or vulnerable. Despite legal protection under the Wildlife Law (SLORC, 1994), conservation efforts for these species in Myanmar are often insufficient and untimely.

The intrinsic traits of turtles and tortoises, such as their long life expectancy and relatively late sexual maturity—often beyond 7 years—render them particularly susceptible to mortality and various threats (Bu et al., 2023). Turtle and tortoise embryos are also sensitive to microhabitat characteristics. Soil temperature, for instance, plays a crucial role in determining the embryo incubation period, as well as the development and sex of hatchlings (Ernst et al., 1994; Friends of Algonquin Park [FAP], 2005). According to Ernst et al. (1994), soil temperatures ranging from 22°C to 27°C result in the hatching of exclusively male offspring, while temperatures between 30°C and 32°C yield only females; temperatures at 20°C and 28°C can produce offspring of either sex. Soil salinity and moisture content can also affect hatchlings' development and survival rate in turtle nests (Bower et al., 2013; Delmas et al., 2008; Finkler, 2006). Higher soil salinity tends to reduce the survival rates of freshwater turtle hatchlings to around 39% as compared to freshwater substrate areas (Bower et al., 2013). Soils rich in clay or soil organic matter, which retain more water, could potentially suffocate turtle embryos (Costanzo et al., 2001; Delmas et al., 2008; Finkler, 2006). In contrast, soils with low organic matter, such as sandy loam, help increase heat capacity and create favorable conditions for incubating embryos (Costanzo et al., 2001; Hughes & Brooks, 2006; Hughes et al., 2009). These compounded factors elucidate why female turtles prefer specific environmental conditions for their hatchlings (Restrepo et al., 2006; Roosenburg, 1996; Wilson, 1998) and why certain species habitually return to the same nesting sites (Congdon et al., 1983; Jackson & Walker, 1997; Loncke & Obbard, 1977). Consequently, turtle nesting activities modify the conditions of surrounding natural habitats, including the temperature and moisture content of soils (Ackerman, 2009; Bonach et al., 2007; Marchand & Litvaitis, 2004). These alterations can indirectly impact the biotic and abiotic conditions of the habitat, influencing resources shared by other species.

Supporting a diverse array of globally threatened wildlife (MOECAF, 2014), the Indawgyi Lake Wildlife Sanctuary (ILWS) holds several well-known designations, including being a UNESCO Biosphere Reserve (UNESCO, 2019), a Ramsar Site (Ramsar, 2016), an Important Bird Area (IBA, 2004), a site within the East Asian-Australasian Flyway Partnership network (EAAFP, 2014), and an ASEAN Heritage site Park (ASEAN Heritage Parks, 2015). Despite its ecological significance, the diversity and breeding ecology of turtles and tortoises in the area remain less studied compared to that of birds and fishes, with scant research on the breeding ecology of wild softshell turtles in Myanmar (e.g., by Fauna and Flora International, 2023). The ILWS is home to more than 50,000 people from three primary ethnic groups: Shan, Bamar (formerly known as Burmese), and Kachin (Forest Department, 2018; Than & Kraas, 2020; Thazin & Eivin, 2020). Many of these communities rely on agriculture, fishing, subsistence livestock farming, and independent ventures such as gold mining and the collection of non-timber forest products for their livelihoods (Karlsen & Røskaft, 2020; Than et al., 2011). Hence, incorporating local perspectives and gathering local ecological knowledge (LEK) can significantly contribute to conservation efforts in the area (Gadgil et al., 1993; Nabhan & Martinez, 2012; Newton et al., 2008).

This study aims to enhance our understanding of turtle and tortoise conservation efforts in the Indawgyi Lake region. Specifically, the objectives are to (1) assess the species composition of turtles and tortoises, (2) investigate their breeding environments with a focus on disturbances and soil conditions around turtle nests, and (3) summarize the conservation concerns related to Indawgyi turtles and tortoises through semistructural interviews with the local community. For the second objective, the study also aims to test the hypothesis that turtles are selecting nesting sites due to differences in the abiotic conditions of the soils. We anticipate finding more nests in areas characterized by low organic matter, low soil salinity, and minimal disturbance.

2 MATERIALS AND METHODS

2.1 Study area

The study was conducted at Indawgyi Lake within the ILWS (24° 56′ N, 96° 39′ E) in Northern Kachin State, Myanmar (Figure 1). Myanmar is home to three large inland lakes designated as wildlife sanctuaries: Indawgyi Lake (the largest), Inlay Lake (the second largest), and Moeyungyi Wetland (Myanmar's first Ramsar site). Surrounded by mountain ranges, ILWS is positioned at an elevation ranging from 300 to 1300 m above sea level. The lake extends 23.8 km from north to south, with a maximum width of 10 km. Its depth ranges from 15.9 to 22.2 m, covering a catchment area of 850 km² (Davies et al., 2004). Over 30 streams flow into the lake, but the Indaw Stream, a tributary of the Ayeryarwady—Myanmar's largest river—is the lake's only outlet (Than et al., 2011).

The Indawgyi Lake is facing environmental pressures due to overfishing, destructive fishing practices, waste management issues, and sediment and mercury pollution from mining activities in the area. These pressures not only threaten the lake's ecosystem but also endanger its diverse aquatic life, including water birds, turtles, and tortoises. Eight turtle and tortoise species are reported to occur in the Indawgyi Wildlife Sanctuary (Kutchling et al., 2004; Myanmar Conservation and Development Program [MCDP], 2012; Shwe & Grindley, 2012) (Table 1), most of which are classified as vulnerable or endangered (Table 1). Of these species, the Burmese peacock softshell turtle (Nissonia formosa), the Burmese narrow-headed softshell turtle (Chitra vandijki), the Burmese eyed turtle (Morenia ocellata), and the Burmese box turtle (Coura amboinensis lineata) are endemic to Myanmar.

Table 1. Turtle and tortoise species of the Indawgyi Lake Wildlife Sanctuary and their IUCN conservation status from known records.

Family	Scientific name	Common name	IUCN
Bataguridae	Cuora amboinensis lineata	Myammar box turtle	Vulnerable
Bataguridae	Cyclemys dentata	Asian leaf turtle	Near threatened
Geoemydidae	Morenia ocellata	Burmese eyed turtle	Endangered
Testudinidae	Indotestudo elongata	Yellow tortoise	Critically endangered
Testudinidae	Manouria emys	Sian brown tortoise	Critically endangered
Trionychidae	Amyda cartilaginea	Asiatic softshell turtle	Vulnerable
Trionychidae	Nilssonia formosa	Burmese peacock softshell Turtle	Critically endangered
Trionychidae	Chirta vandijki	Burmese narrow-headed Softshell Turtle	Critically endangered

• Abbreviation: IUCN, The International Union for Conservation of Nature.

3 RESULTS

3.1 Diversity of turtles and tortoises

We found seven species of turtles and tortoises in Indawgyi Lake (Supporting Information S1: Table S2). Among these, one species, morphologically belonging to the softshell turtle genus Amyda, was undescribed (10 March 2019) (Figure 2; Table 2; Supporting Information S1: Table S2). Its vernacular name, “Late Par Tate,” describes its carapace pattern, reminiscent of Batik art found in traditional Indonesian fabric. The carapace is oval and yellowish, adorned with serial columns of spiny-like tubercles, and the plastron is white. The head of the turtle is reddish-gold, featuring a prominent tubercle on the neck and black stripes around its eyes. The turtle weighed 11.2 kg, with carapace length (CL) = 49 cm, carapace width (CW) = 42 cm, plastron length (PL) = 34 cm, plastron width (PW) = 26 cm, tail length (TL) = 2.5 cm, and tail width (TW) = 4.5 cm (Figure 2).

Table 2. Turtle and tortoise species sampled, their IUCN, and location information.

Common name (N = 34)	Scientific name	IUCN red list
Burmese softshell turtle (n = 9)	Amyda ornata phayrei	Vulnerable
		Location: Nant Khun Chin Stream, Hepu Village
Late Par Tate (n = 1)	Amyda sp.	No Data
		Location: Indawgyi Lake
Burmese peacock softshell turtle (n = 3)	Nilssonia formosa	Critically Endangered
		Location: Nant Phaung Zin Stream and Lae Pon Lay Waterside
Burmese box turtle (n = 5)	Cuora amboinensis lineata	Vulnerable
		Location: Naung Khun Lake of Nyaung Bin Thar Village, Kyar Stream
Oldham's leaf turtle (n = 14)	Cyclemys oldhamii	Endangered
		Location: Nant Khun Chin Stream, Kyar Stream, Yae Aye Stream and Nant Patae Stream
Myanmar brown-leaf turtle (n = 1)	Cyclemys fusca	Least Concern
		Location: In the forest
Yellow tortoise (n = 1)	Indotestudo elongata	Critically Endangered
		Location: Kyar Stream

• Abbreviation: IUCN, The International Union for Conservation of Nature.

3.2 Breeding sites and effect of environmental factors on turtle nests

Among the identified nesting sites, four damaged nests were found at Shwe Taung (S5), an area frequented by grazing buffaloes and cows. No nests were detected at Hepa Gyi Stream Mouth (S2) or Nant Phaung Zin Stream (S4), but turtles were sampled at these sites. Lae Pon Lay Waterside (S1) and Nant Yin Stream (S3) were identified as potential nesting grounds by local hunters, but no nests or turtles were spotted.

In Shwe Taung (S5), test nests, that is, areas where turtles had dug potential nests (length × width × depth = 14.3 ± 2.8 × 12.8 ± 4.0 × 11.8, 11.8 ± 5.1 cm, n = 3) were on average smaller than the damaged nests (17.8 ± 3.3, 15.3 ± 4.6, 19.3 ± 4.6 cm, n = 4). Higher temperatures (28.3 ± 7.4°C) and shorter distances to water (3.5 ± 1.1 m) were recorded at the test nests in comparison to the damaged nests (22.5 ± 1.0°C, 14 ± 6 m; n = 4).

Disturbance was the main explanatory estimator for the differences in the frequency of nests found (sum of model weight, SW = 1, Table 3, Figure 3), followed by % OM (SW = 0.52), and pH reaction (SW = 0.1) (Supporting Information S1: Figure S1). The sites where buffaloes and cows were grazing (Shwe Taung) exhibited a higher number of nests than other sites, particularly those affected by agriculture and pollution. Additionally, areas with high organic matter and soils with a pH of around 6.0 also demonstrated an increased frequency of nests.

Table 3. Comparison of models representing the effects of environmental variables on the frequency of turtle nests found, where models with ΔAICc ≤ 4 were included.

Fixed effects	AICc	ΔAICc	Model weight	R2
Model 1 (disturbance)
Disturbance + % OM	18.3	0	0.521	0.769
Disturbance	18.9	0.64	0.379
Disturbance – pH reaction	21.6	3.29	0.1

• Note: Predictor variables tested include disturbance, pH reaction, % OM, Ex Mg, % moisture, and % sand.

Soils surrounding the damaged nests at Nant Yin Kha Stream (S3) (median pH 5.0, range 5.0–5.5, n = 3) and Nant Phaung Zin (S4) (pH 5.0, range 5.0–5.2, n = 3) were more acidic than soils at Lay Pon Lay waterside (S1) (pH 6.6, range 6.0─6.7, n = 3), Hepa Gyi Stream (S2) (pH 6.0, range 5.9–6.1, n = 2), and Shwe Taung (S5)) (pH 6.0, range 6.0–6.0, n = 3) (Figure 4). The organic matter content of the soil at Hepa Gyi Stream (S2) (OM 0.4%, range 0.3–0.5, n = 2) was lower than at Lae Pon Lay Waterside (S1) (OM 1.2%, range 0.6–4.2, n = 3), Nant Yin Kha Stream (S3) (OM 3.5%, range 3.1–4.3, n = 3), Nant Phaung Zin Stream (S4) (OM 3.1%, range 2.9–4.4, n = 3), and Shwe Taung (S5) (OM 3.2%, range 1.8–5.0, n = 3) (Figure 4).

3.3 LEK and conservation concerns

A total of 146 turtle and tortoise hunters or egg collectors were interviewed to gather knowledge of the ecology and hunting of turtles and tortoises. Results of the interviews identified the Burmese softshell turtle and Burmese peacock softshell as the most well-known species. The Burmese softshell turtle was commonly observed in multiple habitats, including streams, grasslands, and forests, while the Burmese peacock softshell appeared to be predominantly found in lakes. The Burmese box turtle was reported to be abundant in forested areas (Figure 5). October was identified as the primary month for nesting, with April and July noted as hatching months (Figures S2 and S3). The clutch sizes of turtles and tortoises varied widely, ranging from 2 to 380 eggs (median 40).

A quarter to a third of the respondents located nests and eggs by following footprints and signs of plastron rubbing on the ground. Most did not use specialized tools for hunting, though sticks were occasionally used to confirm the presence of nests. 78% believed that turtles and tortoises were mainly hunted by humans, though dogs (8%) and tigers (4%) were also identified as predators. The years 2009, 2014, and 2015 were noted for low turtle populations, with hunting cited as a major cause (Supporting Information S1: Figure S4). Turtles and tortoises were hunted both for consumption (above 43%) and for sale to middlemen (under 51%), while their eggs were mainly consumed (69%).

Only 1% of respondents reported that the Nature and Wildlife Conservation Division had engaged in conservation activities with them, primarily focusing on fish and birds. A small number of turtle hunters (6%) had been penalized for hunting. 82% felt they have a responsibility to conserve turtles and tortoises. They identified hunting (44%), fishing (28%), and deforestation (13%) as the main activities threatening turtle and tortoise populations. To counteract these threats, they recommended reducing hunting (78%) and increasing conservation extension activities (10%) as strategies for conserving turtles and tortoises (Figure 5).

4 DISCUSSION

Seven turtle and tortoise species (34 individuals) were sampled from Indawgyi Lake. Among these, the Burmese peacock softshell turtle, Burmese softshell turtle, Myanmar brown leaf turtle, and Burmese box turtle are species endemic to Myanmar. To our knowledge, this survey represents the first documented occurrence of the Myanmar brown leaf turtle and Oldham's leaf turtle at Indawgyi Lake. We also discovered an Amyda sp. with the vernacular name “Late Par Tate,” which appears to be a species not previously described. This species is absent from available records (refer to Platt et al., 2014; Vetter & VanDijk, 2006) and was unrecognized among the turtle and tortoise photo specimens checked by Dr. Rao Ding-Qi from the Kunming Institute of Zoology, Chinese Academy of Sciences. DNA samples have been collected from this species for DNA barcoding to ascertain if it is a new species.

As turtles are known for their selectiveness in choosing nesting sites, soil variables were expected to be the determining factor driving nest site selection. Results of our models, however, found disturbance (SW = 1.0) as the main explanatory estimator, followed by organic matter (SW = 0.52). This result is unsurprising given that turtles and tortoises, despite being the most threatened vertebrate group globally, have received little attention, even in the ILWS, a UNESCO Biosphere Reserve (UNESCO, 2019). In this region, most turtle and tortoise habitats are highly disturbed by agricultural, boat, and fishing activities. Consequently, areas where animals grazed became the sites where a greater number of nests were found.

Turtles typically prefer nesting sites with low organic matter, such as sandy loam, which is favorable for embryo incubation (Costanzo et al., 2001; Hughes & Brooks, 2006; Hughes et al., 2009). Therefore, the high organic matter indicated by our model may not accurately reflect turtles' habitat preferences but, instead, could be an artifact of the model. It is likely that sites with elevated organic matter were heavily utilized by various animals, including buffaloes, cows, and turtles, which in turn contributed to the increased organic matter content in the soils. Despite their preference for sandy loam sites, human disturbances may compel turtles to nest in less optimal habitats, such as the Shwe Taung Waterside. Although high organic matter could potentially increase mortality due to suffocation risks, both pH level and soil temperature in these areas remain within suitable ranges for turtle nesting. Moreover, the presence of tall Cogon grass (Imperata cylindrical) at nesting sites could benefit embryo survival rates by offering camouflage from humans and predators.

A high variation in clutch size (min. 2, max. 380) was reported by the local community, which appears to reflect intrinsic differences in breeding characteristics between species rather than extrinsic environmental factors. Local villagers noticed that the Burmese softshell turtles typically lay smaller clutches of 20–25 eggs, characterized by their large, circular shape. In contrast, Burmese peacock softshell turtles are known for larger clutches of 150–200 eggs, comprising smaller, circular eggs. Burmese box turtles and Oldham's leaf turtles tend to lay smaller clutches of 7–8 eggs with a cylindrical shape.

The results of this study may reflect biases arising from responses recalled by the interviewees. Despite this possible limitation, the gathered information can serve as valuable reference points in the absence of alternative data sources. For example, information about the locations where mother turtles lay eggs, along with nesting and hatching periods, can highlight crucial months for the protection of turtle nests, eggs, and hatchlings. Such insights are critical to elucidate the efforts required to help conserve and protect turtles and tortoises, especially considering that hunting (44%) and fishing (28%) have been identified as significant threats to turtle and tortoise populations in the ILWS.

The hunters described various hunting and egg collection practices. Some remove all the eggs from a nest, while others leave a few to ensure the survival of some offspring. During egg-laying, some hunters wait to collect only the eggs, whereas others capture both the mother turtle and her eggs. According to Platt et al. (2017), local fishermen trap turtles from December to April, when the river is at its lowest level during the dry season. However, according to the respondents, hunters, and fishermen set their traps along the shore during the rainy season when turtles swim to the lake's edge to lay their eggs (Supporting Information S1: Figure S2). This practice leads to the capture of numerous turtles, particularly pregnant females, that get ensnared in traps. Platt et al. (2017) suggested reducing or eliminating the use of monofilament gill nets to prevent the accidental capture of both turtles and dolphins.

Turtles and tortoises that were hunted were mostly consumed (above 43%) and sold to local middlemen (less than 51%), with some eventually entering the illegal trade. The Burmese Box turtle is highly sought after, fetching prices between 35,000 and 70,000 Kyats, equivalent to 17–34 USD. Interviewees indicated that a surge in market demand for turtles between 2014 and 2015 led to increased hunting, resulting in the perception of turtles and tortoises as being rare during that period. During interviews, the majority of the villagers showed their willingness to conserve turtles for the benefit of future generations, yet 99% reported a general lack of protection and conservation support from the Nature and Wildlife Conservation Division and NGOs. The observed trends in local hunting practices and perceptions suggest that conservation initiatives could be effectively targeted and managed at the local level.

The Wildlife and Protected Areas Law prohibits hunting without licenses, as well as killing, hunting, and transporting the seasonally protected or completely protected wildlife species. It also forbids processing these species as souvenirs or wearing them as part of traditional costumes (Chapter 11, Sections 35, 36, 27, 38, and 39) (SLORC, 1994). We recommend reinforcing law enforcement for wildlife conservation—imposing penalties for illegal hunting within the forestry, agriculture, and fishing sectors, focusing on the protection of habitats and nesting sites, and working with fishermen to ensure the release of turtles caught in nets during the breeding season. It is vital to expand conservation education and outreach activities (Kalra et al., 2023) for turtles and tortoises among local communities (e.g., via community meetings) and personnel involved in wildlife conservation, emphasizing the ecological and cultural importance of turtles and the legal ramifications of engaging in illegal trade. Training programs on conflict mitigation (e.g., coexistence strategies, fencing, safe handling, and relocating wildlife when necessary), understanding the legal framework and policies, and workshops on wildlife monitoring and data collection to develop target mitigation strategies will be beneficial.

Given that the illegal trade often involves local middlemen, there are opportunities to support alternative livelihoods, reducing their dependence on income from the illegal wildlife trade. Furthermore, increasing efforts by authorities, such as customs officials, to curb illegal cross-border and online trade is crucial (TRAFFIC, 2008). The establishment of dedicated cybercrime units to dismantle illegal wildlife trade activities online (TRAFFIC, 2019), along with bolstering international cooperation (e.g., through CITES) to enforce regulations and share intelligence about illegal trade networks, could enhance efforts against all forms of illegal wildlife trade.

In the vicinity of the ILWS, consideration must be given to regulating buffer zone management practices (Ministry of Water Resources, 2013). This is essential for promoting habitat conservation for aquatic ecosystems, which includes restricting or limiting pesticide use and encouraging sustainable farming. These practices would not only benefit wildlife but also enhance the living conditions of people residing in the area.

5 CONCLUSION

The ILWS stands as a global biodiversity hotspot of great biological significance, and this study highlighted the area's importance as a key Testudines habitat, especially with the discovery of a rare or possibly new species, Amyda sp. Although turtles' breeding and survival are known to be strongly influenced by soil quality, human disturbance plays a significant role in deterring them from nesting in suitable breeding grounds. Local communities, many of whom are fishermen or hunters, possess extensive knowledge of turtle nesting habits. A considerable number have hunted turtles and tortoises for personal consumption or sale through middlemen involved in international illegal trade. The findings of this study suggest that conservation efforts could be more effective with the active participation of local communities, as most hunters are from these communities. Furthermore, allocating sufficient resources and expanding conservation outreach specifically for turtle and tortoise protection could significantly enhance conservation outcomes in the region.

AUTHOR CONTRIBUTIONS

Ei Mon Kyaw and Lisa Ong formulated the research questions, did the data analysis, and wrote the manuscript. Ei Mon Kyaw, Nay Htet Naing, Thet Myat Oo, Kyaw Htet Naing, La Minn KoKo, and Thinn Su Tin contributed to field data collection. Xiao-Yong Chen supervised the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.