Distribution, threats and ecology of black breasted leaf turtle (geoemyda spengleri gmelin, 1789) in tay yen tu nature reserve, bac giang province, viet nam with notes on other sympatric species

MINSTRY OF EDUCATION AND TRAINING MINISTRY OF AGRICULTURE AND RURAL DEVELOPMENT VIETNAM NATIONAL UNIVERSITY OF FORESTRY PHAM VAN THONG DISTRIBUTION, THREATS AND ECOLOGY OF BLACK-BRE

MINSTRY OF EDUCATION

AND TRAINING

MINISTRY OF AGRICULTURE AND RURAL DEVELOPMENT

VIETNAM NATIONAL UNIVERSITY OF FORESTRY

PHAM VAN THONG

DISTRIBUTION, THREATS AND ECOLOGY OF

BLACK-BREASTED LEAF TURTLE (Geoemyda spengleri GMELIN, 1789)

IN TAY YEN TU NATURE RESERVE, BAC GIANG PROVINCE, VIETNAM, WITH NOTES ON OTHER SYMPATRIC SPECIES

MASTER THESIS IN FOREST SCIENCE

Hanoi, October 2018

MINISTRY OF EDUCATION AND TRAINING MINISTRY OF AGRICULTURE AND RURAL

DEVELOPMENT

VIETNAM NATIONAL UNIVERSITY OF FORESTRY

- -

PHAM VAN THONG

DISTRIBUTION, THREATS AND ECOLOGY OF

BLACK-BREASTED LEAF TURTLE (Geoemyda spengleri GMELIN, 1789) IN

TAY YEN TU NATURE RESERVE, BAC GIANG PROVINCE, VIETNAM, WITH NOTES ON OTHER SYMPATRIC SPECIES

Major: Forest Science Code: 8620201

MASTER THESIS IN FOREST SCIENCE

Academic advisor: Dr Luu Quang Vinh

Prof Dr Thomas Ziegler Signature:

Hanoi, October 2018

ABSTRACT

Ranking as number 9th in the top of the world, Vietnam is one the most biodiverse countries in terms of turtle species richness, with 27 species being tortoise and freshwater turtle However, the natural history of all of these species is very little known, but it is thought that the great majority of them are heavily threatened by habitat loss and overhunting Here, this study provides the primarily information from research (based on both interviews local people and field survey through the forest) in the Tay Yen Tu Nature Reserve, Bac Giang Province, Vietnam in order to collect

natural history information (distribution, ecology and threats) on Geoemyda spengleri

There were at least 10% (n=72) of the interviewees recognized 12 species It includes

G spengleri, C mouhotii, C cyclornata, C galbinifrons, I elongata, P steindachneri,

P sinensis, P megacephalum, S quadriocellata, M mutica, C zhoui, C oldhamii G spengleri and C mouhotii were by far the most frequently cited species by

interviewees, and were also dominant in a sample of 40 turtle individuals that were observed on local trade These two species were also observed during field transects

(seven G spengleri and one C mouhotii) It is the first record of the G spengleri population and its co-occurrence in Tay Yen Tu NR Individuals of G spengleri were

observed from 10:47-13:45 hour, in all cases in sunny weather after rains, at 25-300C

temperature range G spengleri was seen in montane bamboo forests at elevation range

from 651m-725m, often inside or nearby rocky caves, and in sites with forest cover

being 70%-90%, at slope angles varying from 5 to 45 degrees While C mouhotii was

recorded at 607m a.s.l during the sunny with forest cover is 50% This study is also

estimated the population size of G spengleri in Tay Yen Tu NP range from 620-1849

individuals

Key words: Tay Yen Tu Nature Reserve, Geoemyda spengleri, first record,

co-occurrence

CONTENTS

ABSTRACT i

CONTENTS ii

ABBREVIATIONS iii

LIST OF FIGURES iv

LIST OF TABLES v

I INTRODUCTION 1

II LITERATURE REVIEW 3

2.1 Chelonian research in the world 3

2.2 Chelonian research in Vietnam 9

2.3 Study species 16

III METHODOLOGY 17

3.1 Study sites 17

3.2 Data collection 19

3.2.1 Interview surveys 19

3.2.2 Field surveys 21

3.2.3 Species distribution and population status 23

3.2.4 Threats 24

3.2.5 Statistical analyses 25

IV RESULTS 26

4.1 Tay Yen Tu NR‟s species composition and status 26

4.2 Field surveys 29

4.2.1 General findings 29

4.2.2 Species ecology 30

4.3 Threats 36

4.3.1 Online trade 36

4.3.2 Threats to G spengleri population and sympatric species in Tay Yen Tu NR 37

V DISCUSSION 39

5.1 Chelonian co-occurrence species status in trade in Tay Yen Tu NR 39

5.2 The wild G spengleri population status 40

5.3 Species ecology 40

5.4 Threats 42

5.5 Recommendations 43

VI CONCLUSIONS 45

ACKNOWLEDGEMENTS 47

REFERENCES 49

ABBREVIATIONS

a.s.l Above sea level

ArcGIS Geographic information system

ATP Asian Turtle Program

BBLT Black-breasted leaf turtle

CR(PE) Critically Endangered (Possibly Extinct)

CR(PEW) Critically Endangered (Possibly Extinct in the Wild)

GPS Global Positioning System

HBWA High-biodiversity wilderness areas

NT Near threatened

TCC Turtle conservation center

TCM Traditional Chinese medicine

TFT Tortoise and freshwater turtle

TFTSG Tortoise and freshwater turtle specialist group

TPA Turtle priority areas

USA United State American

VNUF Vietnam National University of Forestry

LIST OF FIGURES

number Figure 1 Map of turtle diversity countries in the world Source:

Mittermeier et al., 2015

4

Figure 2 Distribution of G spengleri in Vietnam and China and Laos

Source: Turtle Taxonomy Working Group et al.,2017

16

Figure 4 Forest in Tay Yen Tu Nature Reserve 19 Figure 5 Map of Vietnam, showing the position of the interview

points and of the presence sites of turtles at the Tay Yen Tu

NR The field survey tracks are also indicated

29

Figure 6 Bamboo forest (Arundinaria sp.) where is the „favorite‟

habitat for G spengleri in Tay Yen Tu NR

30

Figure 7 Black-breasted leaf turtle (Geoemyda spengleri) in the wild

(male at above, and female at below)

31

Figure 8 A single C mouhotii found in forest by hunting dog 32 Figure 9 G spengleri population structure in Tay Yen Tu NR 33 Figure 10 Predicting potential distribution map of G spengleri in Tay

Yen Tu NR region

35

Figure 11 73 individuals of G.spengleri were offered for sell by

Facebook in Quang Ninh province in October 2017

36

LIST OF TABLES

number Table 1 Top countries in term of tortoise and freshwater turtle taxon

richness Source: Mittermeier et al., 2015

6

Table 2 Number of TFT trade from farm in China Source: Haitao et

al., 2008

8 Table 3 List of TFT species within Vietnam Source: Turtle

Taxonomy Working Group et al., 2017

14

Table 4 Synopsis of the places where turtle individuals were observed

during the interview surveys

28

Table 5 Synopsis of body measurements (mean ± Standard Deviation)

taken from both wild-caught and traded turtle individuals at the study area

33

in the IUCN Red List (IUCN, 2018) For most species, habitat loss and overhunting (for consumption as human food, traditional medicine, religious merit) have been the main threats (Stanford et al., 2018) Indeed, the forest cover in Vietnam has decreased from 43% to 36% from 1945-1990 (De Jong & Hung, 2006) Although the country‟s forest cover has been recently increased due to the national-scale reforestation of Vietnam since 1992 that is assumed to contribute to this recovery, the forest quality still remains poor (Meyfroidt & Lambin, 2009) Indeed, the recovery of forest coverage was due in part to the ban of industrial logging from natural areas and the displacement of wood extraction to neighboring countries, but mainly to extensive programs of homogenous tree plantations (Meyfroidt & Lambin, 2009) These plantation habitats are clearly unsuitable for turtles, whereas the suitable remnant natural forest is still heavily fragmented and isolated (Fox et al., 2004) The second main threat to Vietnamese turtles is trade for domestic consumption, with massive numbers of individuals being exploited in the 1980s-2000s (Van Dijk et al., 2000; Nijman, 2010) also to supply the food and traditional medicine markets of China (Cheung and Dudgeon, 2006; Turtle Conservation Fund,

2002) According to IUCN (2018) assessments, the populations of most Vietnamese turtle species have declined by about 50-90% The Black-breasted leaf turtle

(Geoemyda spengleri), that is the main target of this study, is not an exception

(however, we also present some remarks on the other sympatric turtle species)

The Black-breasted leaf turtle (Geoemyda spengleri) (Figure 1) is among the

smallest Geoemydidae species worldwide It is native to southern China, Northern Vietnam and Laos (Yasukawa and Ota, 2010; Turtle Taxonomy Working Group et al., 2017), and is classified as Endangered by IUCN (2018) due to habitat loss and trade (Gong et al., 2005, 2009) However, the species‟ population size in the wild

and the distribution are poorly known (for instance, G spengleri has recently been

also recorded in Laos (Stuart et al., 2011), with no ecological study published so far (Yasukawa & Ota, 2010) and with behavioral data being recorded only in ex-situ

research (Henze et al., 2004) In Vietnam, the presence of G spengleri has been

documented in Tam Dao National Park, (Bourret, 1934; Nguyen & Ho 1996; Yasukawa et al., 2001), Tuyen Quang, Son La, Yen Bai, Vinh Phuc, Bac Giang, Quang Ninh, Thanh Hoa, Nghe An, and Ha Tinh provinces (Bourret, 1934; Le, 2001; Nguyen et al., 2010) The southernmost boundary of its distribution range lies

in Da Nang and Quang Nam Provinces (Le, 2000; Le & Nguyen, 2003) However, the coastal city of Da Nang may serve as a collection center rather than a field locality (Yasukawa and Ota, 2010) Few conservation actions have been implemented to manage the remnant free-ranging populations of this species Our aim in this study is to present some natural history data on G spengleri in a

protected area of Vietnam (Tay Yen Tu Nature Reserve) in the hope to contribute to

a better understanding of the field ecology and conservation of this threatened turtle species, and also to give some additional notes on the turtle fauna of the study area

In detail the thesis goal aim to provide scientific information on species

distribution, population status, threats, ecology and distribution of G spengleri

within Tay Yen Tu Nature Reserve

To archive the goal there are several objectives that the study needs to reach: (1) provide basic scientific data on the animal ecology and habitat condition to improve captive management of the ex-situ populations; (2) better understand the threats to the species to enhance enforcement; (3) determine the population status in the protected area to provide the scientific information in order to give recommendation the conservation actions

II LITERATURE REVIEW

2.1 Chelonian research in the world

Turtle is unique group of Reptile class with special body design to against the predators and survival for million years Oldest turtle fossil sample has recognized

to live in Upper Triassic from approximate 150 million years ago as same time as Dinosaur dominated (Baur, 1888; Fraas, 2009; Sterli, 2008) After million years, the human start to study about the unique creature- turtle Currently, the turtle taxonomy working group has listed 356 species of tortoise and freshwater turtle

Fig 1 Map of turtle diversity countries in the world Source: Mittermeier et al., 2015

(TFT) belong to 14 families, 13 subfamilies, 94 genera (Turtle Taxonomy Working Group et al., 2017) With consist of 60 polytypic and 122 recognized subspecies, there are totally 478 turtle taxa of modern chelonian in the world (Turtle Taxonomy Working Group et al., 2017)

The turtle working group from IUCN specialist has analyzed the taxon richness and endemism of TFT in the world‟s premier biodiversity conservation priority areas and countries with greatest turtle richness These 16 turtle hotspots

together contain 262 species (83% of total), of which 149 (47%) are endemic to these areas combined, and 134 species (43%) are endemic to just a single priority area At the terminal taxon (subspecies) level, these 16 areas harbor 342 taxa (79%),

of which 210 (48%) are endemic to these combined areas, and 195 (45%) are

endemic to a single priority area These 16 biodiversity hotspots (BHS),

high-biodiversity wilderness areas (HBWA), and additional turtle priority areas (TPA)

account for less than 24 million square kilometers, or 16.0% of planet Earth‟s land

surface, with an estimated 10.4 million square kilometers of original habitat

remaining, or 7.0% of the planet‟s land surface Twenty-one countries are recorded

to harbor 15 or more species of non-marine turtles, with the percentage of endemic

species ranging from 0% to 88% (Mittermeier et al., 2015) According Mittermeier

(2015) the top countries in term of turtle fauna richness is also presented (Figure 1 and Table 1) Among them, USA, Mexico in North America and India in Asia are top three countries in the world in term of turtle fauna richness with 89, 65, and 48 species and sub-species respectively (include marine turtle) while Vietnam is hold number 6th among with big country Australia, and Brazil in term of turtle species richness with 36 turtle species and sub-species (include marine turtle) (Table 1)

However, 69.7% turtle species are listed in IUCN red list 2018 In detail TFTSG Red List 2018 encompasses 356 species, of which 7 are Extinct (EX), 1 Critically Endangered (Possibly Extinct) (CR(PE)), 1 Critically Endangered (Possibly Extinct in the Wild) (CR(PEW)), 69 Critically Endangered (CR), 54 Endangered (EN), 61 Vulnerable (VU), 39 Near Threatened (NT), 93 Least Concern (LC), and 35 Data Deficient (DD) (Rhodin et al., 2018)

Trade of turtles, for both food and pet, represents a substantial business either internationally and within several „mega-biodiverse‟ tropical and subtropical countries, and has consequently risen considerably through the

last decades (Auliya et al., 2016; Böhm et al., 2013; Luiselli et al., 2016) The turtle trade has been especially massive in Asia (Van Dijk et al., 2000; Turtle Conservation Fund, 2002; Cheung and Dudgeon, 2006), thus causing substantial concern among conservation biologists who defined

it as the „Asian turtle crisis‟ (e.g., see Cheung and Dudgeon, 2006; Horne et al., 2011; Ly et al., 2011; Van Dijk et al., 2000) During the

„Asian turtle crisis‟, there were an estimated 13 000 ton of turtles (Table 2), including wild caught as well as farmed individuals, being traded in the Chinese market in the period 1990s-2000s (Cheung and Dudgeon, 2006; Yiming and Dianmo, 1998)

Habitat loss or degradation as well as the extensive consumption and pet trade are among the most significant threats Some estimates put the number of turtles traded annually at ten million (McCormack, 2015) Further (Salzberg, 1998) explains that almost all of these animals are wild caught and considering the low

Table 1 Top countries in term of tortoise and

freshwater turtle taxon richness Source:

Mittermeier et al., 2015

reproduction rate of most species that this exploitation of turtles and tortoises is not sustainable and extinctions of some species in the wild can be expected within the next decade (Gibbons et al., 2000) Since the convertibility of the Chinese currency

in 1989 China‟s turtle imports increased dramatically (Cheung and Dudgeon, 2006) and today China is the largest consumer of turtles (Gibbons, 2000; Gibbons et al., 2000) The country‟s developing economy and the increasing affluence generated a demand for expensive food and medicine made out of turtles Particularly for Asian TFT, which are harvested as food source and for traditional Chinese medicine (TCM), the situation is very acute (Cheung & Dudgeon, 2006) As populations decline and the international pet trade of turtles becomes more popular (Haitao et al., 2008), the amount of animals traded has decreased although the value of each individual turtle has increased (Lau et al., 2000) providing incentives for further exploitation According Haitao (2008) calculated that there were totally about 300 million turtle being kept in farm to serve the China market, this worth $750 million

annually Among them, there were 21 739 individual of G spengleri being keep in

farm to serve the pet trade market (Table 2)

However, the positive news that the conservation turtle and TFT has received

a good attention from community to preserve the unique group of animal As analyzed, the investment on TFT conservation organization worth billion US dollar

in order to preserve the TFT species in-situ and ex-situ (Sigouin et al., 2017)

As ecology study will take time and costly that why it normally was conducted by the developed country As well as, there are more abundant of TFT population

in Western country or USA

is a benefit to conduct the ecology research A well-known research on turtle ecology (movement) was focusing on

Chelodina longicollis (Testudines: Chelidae), in southeast Australia by using radio

telemetry for long time It shows that the species movement was range from 13.8

± 2.8 ha (±SE) home range, 2608 ± 305 m total distance moved, 757 ±

76 m range length (Roe and Georges, 2008) Or long term study of Cyclemys muhlenbergii in USA revealed the species movement and habitat use (Eckler et al.,

1990) Radio tracking is commonly used to monitor the endangered TFT species and to search for their diet in the wild (Jaeger et al., 1997; Mech and Barber, 2002)

A combination between radio tracking and mark recapture has been used for the

common snapping turtle (Chelydra serpentina) in USA to collect data on species

ecology and population status (Obbard and Brooks, 1981) To study the Table 2 Number of TFT trade from farm in China

Source: Haitao et al., 2008

translocation as a conservation tool, the site fidelity and movement of repatriated

gopher tortoise (Gopherus polyphemus) is also conducted in USA to protect them

from threats (Tuberville et al., 2005) There are several studies on the Gopherus polyphemus that give the full understanding on species However, it is costly and

consume such a huge amount of time For example, Castellón et al (2018) has studied the home range, seasonality, burrow use, habitat fidelity and scrub of Gopher Tortoise (Gopherus polyphemus) that spent years on conducting the

research, or nature history and management of the species (Ashton and Ashton, 2008), or nesting behavior of the Gopherus polyphemus (gopher tortoise) (Dziadzio and Smith, 2015) was also consumed huge time The study on Manouria impressa and Indotestudo elongata was also conducted long term in Thailand and Cambodia

to reveal many aspects of the species include diets, habitat use, nesting behavior and

so on (Ihlow et al., 2014; Wanchai et al., 2012) The diet, reproduction and nesting

of several turtle species were also recorded by scholars in China (Ji-Chao et al., 2011; Xiao et al., 2017)

There were studies related to TFT ecology around the world However, publications on taxonomy and distribution study seem to be outweigh of the ecology

2.2 Chelonian research in Vietnam

During 18th - 19th century, there were several publications that documented chelonian biodiversity and described many new species in Vietnam (Bourret, 1941,

1937, 1934; Fan, 1931; Gmelin, 1789) According to Bourret (1934, 1937, 1941) recorded and described 18 species of chelonian in Vietnam and Indochina It

includes 2 marine turtle and 16 TFT species Those are Chelonia myda, Caretta caretta, Platysternum me ga cephalum, C y clemys mouhotii (Cuora mouhotii), Cyclem y s dentata, Cuora amboinen s is, Geoemyd a spengleri, Geoemyda

tcheponensis (Cyclemys oldhamii), D a monia subtr i ju ga (Malayemys subtrijuga),

Hier e mys annandale i (Heoesemys anadalii), Clemm y s mutica (Mauremys mutica), Clemmys bealei quadrio c ellata (Sacalia quadriocelata), Ocadia sinensis (Mauremys sinensis), Testudo e longata (Indotestudo elongata), Testudo impressa (Manouria impressa), T r ionyx cartilaginous (Amyda cartilaginea), Trionyx sinensis

s in en s is (Pelodiscus sinensis), Trionyx steindachneri (Palea steindachneri)

Since then, there were several researches and conservation programs have been established It listed 27 TFT species within Vietnam (Turtle Taxonomy Working Group et al., 2017) (see detail in table 3)

There were many types of material in turtle sector have been published such

as books, papers A photographic guide to the turtle of Thailand, Laos, Cambodia and Vietnam has been release in 2001 to support the conservationists and researcher

in identify (ID) turtle species in region and Vietnam (Stuart et al., 2001) After that,

a new book for Vietnam has been developed by NGOs: Education for Nature of Vietnam (ENV), Asian Turtle Program (ATP) in 2011: a field guide for identify turtle in Vietnam in order to support the law enforcement As well as in the guide of

reptile and amphibian of Vietnam also mention TFT species (Nguyen and Ho, 1996) Even the study in taxonomy new species/subspecies has been split from

original taxa for example Cuora trifasciata was split into Cuora trifasciata for Chinese part of distribution and Cuora cyclornata for Vietnam part (Blanck et al., 2006) or Cuora mouhotii was separated into two subspecies: Cuora mouhotii mouhotii and Cuora mouhotii obsti (Das et al., 2016; PJh struiJk et al., 2016) or Cuora galbinifrons has been separated into 3 full species Cuora galbinifrons, Cuora bourreti and Cuora picturata (Dijk et al., 2014; Stuart and Parham, 2004) There

were several research on species distribution which is interested such as the

integration of Cuora mouhotii mouhotii and Cuora mouhotii obsti and Cuora galbinifrons and Cuora bourreti (Fritz et al., 2002; PJh struiJk et al., 2016) A complete paper about Mauremys annamensis an endemic species to central of

Vietnam has been published by (McCormack et al., 2014)

About turtle trade, there were many studies have been published Vietnam has special geographic location which is close to the most giant turtle consumption country-China being a “cross-bridge” for wildlife trade from Indochina to China (Ngoc and Wyatt, 2013; Van Song, 2008; Yiming and Dianmo, 1998), has been the main turtle supplier to Chinese market (Hendrie, 1998, 2001a) Massive amounts of turtles served the China‟s market as food, ingredients for traditional medicine, farming and pet trade (Hendrie, 2000), whereas Vietnamese people do not traditionally consume turtles for subsistence and rarely use their parts for traditional medicine (Le Dien Duc and Broad, 1995) The turtle business as a whole was worth

USD 750 million annually in China with more than 300 million of turtles being sold yearly (Haitao et al., 2008) As a consequence of decades of massive trade, the wild populations of many Vietnamese turtles have been substantially impoverished (Hendrie, 2001b; Le Dien Duc and Broad, 1995; Le et al., 2004; Ly et al., 2011) with some species being on the brink of extinction (Rhodin et al., 2018) Currently, after decades of overharvesting, the quantity of turtles on trade has declined remarkably in Vietnam presumably due to a massive decline of the wild turtle populations (Van Song, 2008)

In the last five years, the situation has been changing in that increasing numbers of turtles are being imported to Vietnam from China, Thailand and Malaysia to serve the pet market (Thong Pham Van, unpublished observations) Before 2010, keeping turtles as pets was rare within Vietnamese households due to the local culture considering any turtle to be a God, thus discouraging any attempt at keeping turtles at home as a source of bad luck to the owners (Espenshade and Le, 2002; Le Thien Duc, 2003) However, there were a few shops, mostly from the tourist areas such as Tam Dao town, Vinh Phuc Province, where turtles were sold as

pet, with just few species being kept (i.e Geoemyda spengleri, Cuora galbinifrons, Cuora mouhotii) (Le Dien Duc and Broad, 1995) However, after 2010, young

Vietnamese people started to collect turtles for pet, and this trend became a true

“fashion” in 2012-2013 Nowadays, there are dozens of turtle species, including native and allochthonous species of all ages (hatchlings, juveniles, sub-adults and adults) being traded for the pet purpose (Tran et al., 2016) For instance, juvenile

turtles (especially Trachemys scripta) became frequently traded in animal shops in

Hanoi (our unpublished observations)

The genetic study was very limited However, there were a big debate on the taxonomy of a Giant softshell turtle lived in Hoan Kiem lake, old quarter of Hanoi Some people believe it is distinct with individuals being kept in China that separate

into new species under different name: Rafetus leloii, Rafetus vietnamensis, Rafetus swinhoei The genetic finally, solved the problem, all specimens of the species in Vietnam and China was the same species as Rafetus swinhoei (Farkas et al., 2011;

Le et al., 2014; Le and Pritchard, 2009)

Regarding to the turtle ecology, there were very limited publications Even

the species (include highly protection status in IUCN red list such as Rafetus swinhoei, Mauremys annammensis, Cuora ssp.) distribution and population status,

habitat type were poorly known Currently, a holistic study was conducted by

McCormack, Sarah Wahl, Hoang Van Ha on Cuora mouhotii in Cuc Phuong

National Park were considered to be an ecology research (Hoang et al., 2010; Inga

et al., 2010) Besides, the study on turtle in Vietnam only limited in species composition (Duong et al., 2014)

For the last two decades, there were several organizations that have been established in order to protect turtle in Vietnam TCC is a first rescue center specialized in rescue the TFT from confiscation Located in Cuc Phuong NP, Ninh Binh Province, TCC is one of official rescue centers that now captive breeding program for over 600 individuals of 19/27 native species of Vietnam (Asian Turtle

Program, 2014) While ATP has responded for in-situ with the main projects focus

on Critical endangered turtle of Vietnam such as Cuora ssp Rafetus swinhoei, Mauremys anamensis (Asian Turtle Program, n.d.) Another rescue center has been

planning to build at Bach Ma NP, Thua Thien-Hue Province by Turtle Sanctuary and Conservation Center, a new NGO from France Together with awareness program from those NGOs, the national legislation and regulation are also evolved with the born of two strongest law is Decree 32/2006/NĐ-CP and Decree 160/2013/NĐ-CP to protect turtle in Vietnam (detail in table 3)

Family/Species Common name

UCN Red list

2018

Vietnam red book

2007

CITES Appendix

2015

Decree 32/

2006/NĐ-CP

Decree 160/2013/NĐ-

2.3 Study species

This study mostly focuses on Black breasted leaf turtle (G spengleri) (Figure 7) It belongs to the order of the Testudines and the family Geoemydidae and, based

on the taxonomist, is treated under the genus Geoemyda This species is recognized

as small turtle (107 mm CL), being known as terrestrial turtle occurs in Northern Vietnam and Southern China and Laos (Stuart et al., 2011; Yasukawa and Ota, 2010) (Figure 2) Very few data is available regarding their life history and population status Clutch size in captivity range from 1-2 eggs (42-45 x 18 mm in size) with the hatching measurement is about 30 mm CL (Yasukawa and Ota,

2010) Currently, there was none subspecies recognized Synonymy: Testudo spengleri (Gmelin, 1789), Geoemyda spengleri, Testudo serrata Shaw 1802, Testudo tricarinata Bory de Saint-Vincent 1804, Geoemyda spengleri sinensis (Fan,

1931)

Fig 2 Distribution of G spengleri in Vietnam, China and Laos

Yellow dot is field record while the dark red is potential

distribution of G spengleri Source: Turtle Taxonomy Working

Group et al., 2017

III METHODOLOGY

3.1 Study sites

Tay Yen Tu NR was established in 2002 with a total area of 13,002 ha, the Tay Yen Tu NR consists of two parts: Tay Yen Tu and Khe Ro The NR is the largest remnant natural forest patch in the Quang Ninh and Bac Giang provinces, and is mainly located in Luc Nam and Son Dong districts, Bac Giang Province

(coordinates: 21°10'N and 106°43'E; Figure 1)

In this NR, a remnant portion of the natural habitat of the study species still exists (Figure 3) (Forest Protection Department of Bac Giang Province, 2010) The

NR is bordering with three provinces: Quang Ninh, Lang Son, Hai Duong The NR lies on the monsoon climate with characteristic of hot, moisture and rain Average

Fig 3 Map of Tay Yen Tu Nature Reserve

temperature is 230C (range from 15.10C-28.50C) The annual average rainfall is 1483.3mm (with the highest was 317.1mm/month, and lowest 11.4mm/month) An estimation of rainfall duration is 120 days per year that mostly focus on duration from May-August The annual average humidity is range from 79%-82% About watershed, there are seven big streams within the NR which is provide water for Tay Yen Tu region Those streams are: Dong Ri, Dong Bai, Nuoc Nong, Nuoc Vang, Da Ngang, Khe Din, Khe Ro In the Tay Yen Tu section, the mountain peak

is 1068m a.s.l (Mount Yen Tu) while in the Khe Ro section the mountain peak is 886m (Forest Protection Department of Bac Giang Province., 2010) The main forest type of Tay Yen Tu is the lowland evergreen broad-leaved tropical forest (Averyanov et al., 2003)

The NR is known to be home of 728 plant and 285 animal species (Dong, 2016; Forest Protection Department of Bac Giang Province., 2010), including 76 species of amphibians and reptiles (Hecht et al., 2013) There are five types of forest cover in the NR Shrub and grassland are dominated below 100 m a.s.l., followed by small wooden trees mixed with bamboo at the elevation of 100-200m From 200-900m the forest is evergreen broad-leave tropical forest mix with bamboo forest (Figure 4) Above 900m is the broad-leaved forest with the dominant of big wooden tree (Forest Protection Department of Bac Giang Province, 2010)

3.2 Data collection

We decided to carry out field surveys for G spengleri at the study area after

having verified, in the years 2017-2018, that many individuals were offered for sale

on the internet (Facebook, etc.) and were reported to come from Tay Yen Tu NR by the interviewed traders Thus, the research consisted of both interviews with selected people and field surveys in natural habitats that were potentially inhabited

by the study species

3.2.1 Interview surveys

We conducted semi-structured interviews, with identification photos of G spengleri (from hatchling to adult) and other 13 potentially sympatric chelonian species (Platysternon megacephalum, Mauremys sinensis, Mauremys mutica, Cuora galbinifrons, Cuora cyclornata, Cuora mouhotii, Cuora zhoui, Indotestudo

Fig 4 Nature forest in Tay Yen Tu Nature Reserve Photo: Pham Van

Thong

elongata, Sacalia quadriocellata, Palea steindachneri, Pelodiscus sinensis, Rafetus swinhoei, Pelochelys cantorii) with 72 people in 11 communes in June and July,

2018 (Appendix 2) Most of the interviewees were traditional medicine collectors and hunters, who were presumed to be familiar with the species (Table 4) There were 11 interview days from 17th-27th Jun 2018 and 28th July 2018 We also interviewed 38 randomly selected people (neither traders, traditional medicine practitioners nor professional hunters) in order to verify whether non-experienced local people were able to describe the turtle species (Appendix 2) We used the following nine questions to each interviewee: (i) do you know any turtle from your place? (ii) If yes, what is their local name? (iii) What do they look like (carapace color, hardshell or softshell, size of the animal, aquatic or terrestrial)? (iv) What do they eat? (v) When do you see it? (vi) Where do you see it? (vii) How many people buy/sell turtles or what do they do for the turtle if they got them? (viii) Which morphological characteristics do you use to determine the different species (hinged

(for Cuora spp.), serration, size of the head, length of the tail (for Platysternon megacephalum)? (ix) What is the habitat in which the turtle is found? After

completing the interview, to check again their description, we showed them the ID photos of turtle species Approximately, each interview lasted from 5-30 minutes

We also collected information on interviewee‟s profession, gender and address after obtaining his/her explicit consent; however, this information is not shown in this thesis in order to keep their privacy Before starting the interview, the team leader introduced the staff, explaining that these interviews would become the scientific

record for master student projects focusing on turtles We recorded the interview site using a GPS device (see Figure 3 & 5 for the map of the study sites)

During the interview surveys, we also recorded captive turtles held by villagers For each individual, we asked what do they intend to use the animal, how much do they sell/buy it, and we measured the strait max carapace length, max carapace width and shell height by a caliper (30mm) (Appendix 10) and weight by

an Electronic Kitchen Scale, with the accuracy of 0.1g To identify the TFT species,

I used the book of Photographic guide to the turtles of Thailand, Laos, Cambodia and Vietnam (Stuart et al., 2001) The basic information from TFT record will be written in the field record in Appendix 9

3.2.2 Field surveys

We conducted the field surveys of the first section in Dong Ri commune, Son Dong district from 19th-22nd July 2018 with 5 people and two hunting dogs and the second section in a forest in Tuan Mau commune, Son Dong district from 23rd –

27th July 2018 with 6 people and 3 hunting dogs Following Dawson‟s (2016) data, combining with local experience, we selected only forest patches situated at above 400m elevation as field survey sites When a turtle individual was found, we recorded (i) habitat type (mostly bamboo forest with rocky areas), (ii) slope angle, (iii) forest cover, (iv) ground temperature (°C) Slope angle was determined by the Angle Meter version 1.0 software in an Android 8.0 system smartphone Forest cover was estimated by Canopy Cover 1.03 version software, also in an Android 8.0

system smartphone The ground temperature was measured by a thermometer, by placing the tip of the thermometer on the surface in the exact position of the turtle

For locating wild turtles, we used local people (researchers/assistants), and

2-3 hunting dogs in 8 days, all of them being employed to follow transects crossing through the potential turtle habitats (see Figure 5 for the spatial arrangement of the transects) The beginning and ending time of each survey, and the GPS points were recorded at the beginning and at the end of the transect Each person followed the transect by keeping a nearly constant 5 m distance between each surveyor Hunting dogs were used to improve the searching effort These dogs were exercised for their hunting skills for turtles by their owners and were well experienced in searching for turtles inside the forest Hunting dogs normally run to search turtles by themselves without any control from the owner So, we could not include them in our calculations of survey effort If they find any turtle, the dog would bark to inform the owner Hunting dogs are used commonly by some ethnic minorities to hunt for wildlife in protected areas in Vietnam (Espenshade and Le, 2002; Pham et al., 2018) A mean of 45.67 hours survey effort was made per person, with the searching team including 5-6 people and 2-3 dogs on each day Thus, excluding dogs and considering the searches by each person as independent along the transects, the total field effort was 251.2 hours Field surveys were conducted in two areas of Tay Yen Tu NR, with 11,280 m of transects being walked by the researchers The minimum and maximum elevation of our transects were 416m and 795m a.s.l respectively, with the average elevation being 604.5m a.s.l Transects

were walked with all weather conditions, from sunny and dry to cloudy and rainy, including also sunny weather after heavy rains

When a free ranging turtle was encountered, it was photographed and measured as mentioned above We also recorded for each individual (i) the GPS coordinates, (ii) the time (Hanoi Standard Time), (iii) the weather, (iv) the ground temperature (°C) the habitat type, (v) the elevation (m a.s.l.), (vi) slope, and (vii) forest cover We also determined its (a) sex and (b) age class We determined sex by secondary sexual characteristics (tail morphology and head patterns; Yasukawa & Ota, 2010; see Figure 7), and classified the age classes as follows: carapace length

≥80mm = adult, 60-80mm = subadult, 40-50mm = juvenile, <40mm = hatchling

We set free these individuals unharmed after data processing

3.2.3 Species distribution and population status

Basing on data available on internet and the result from interview and field survey, the technology is applied to predict the species distribution in Tay Yen Tu

region and estimate the population size of G spengleri

To estimate number of individual/ha, first step, the size of line transect will

be calculated follow formula (as assume that each person will respond for 5m width

of searching area along line transect): Line transect size = number of researchers *

total line transect length * 5 m Second step based on numbers of individuals of G spengleri found, we can calculate the number of individual/ha following formula: Number of individual/ha=number of G spengleri found/Line transect size

Next step, calculate the size of potential distribution of the species in Tay Yen

Tu NR To calculate the potential distribution of G spengleri in Tay Yen Tu Forest

area, I used ArcGIS 10.2 to classify the potential habitat of species within NR Due

to lack of data and other variable, this study only selects the elevation as variable

We used Digital elevation model (DEM) raster with the resolution of 100*100 pixels Base on local experience, the species is normally distributed in the Tay Yen

Tu NR event at elevation 500m while we found the turtle at 651m So we choose the elevation of 500m to 650m to classify the interval estimate the size of potential distribution of the species in the NR To do this, the DEM raster will classify into two classes from 32m-500m and 500m-1044m (the NR elevation range from 32-1044m), then Reclass/Reclassify to choose the area of elevation above 500m The area size calculated by the number of pixels (represent for area above 500m) multiple by DEM resolution (100*100) The same process applied to elevation above 650m

Then the population size of G spengleri in Tay Yen Tu NR calculated by formula: Population size of G spengleri = size of potential distribution of G spengleri in Tay Yen Tu NR * number of individual/ha

3.2.4 Threats

3.2.4.1 Online trade monitoring

Online trade monitoring Facebook search was the main search tool for

finding the case of buy/sell G spengleri on popular groups such as: Hội Yêu Rùa

Việt Nam, Hội nuôi rùa bảo tồn, bán rùa cảnh, Hội rùa Việt Nam and personal

Facebook who sold turtle online Google search tool was also used to allocate the trade case/confiscation case happened to estimate the volume of trade within Vietnam Pet online websites/forum were also target for monitoring

3.2.4.2 Threats to G spengleri population and sympatric species in Tay Yen Tu

NR Threats were collected during the interview survey and field survey by asking local people about the trend of trade, altitude to turtle, law enforcement, hunting and

so on During field survey, the impact from local people to G spengleri population

and other sympatric TFT species (e.g hunting, road constructions, logging) within the NR is also recorded

3.2.5 Statistical analyses

To calculate turtle relative frequency of observation for each transect, we

used the following formula:

with S = turtle relative frequency of observation, x = number of turtles found in a given transect; a = number of people involved in the time search along a given transect; b = total time spent for a trip along a given transect (follow Timothy

communication)

Observed-versus-expected χ2 test was used to evaluate (i) the frequency differences of occurrence of the various species in the hands of hunters/traders, and (ii) the frequency differences among the various turtle species in terms of the

number of interviewees reporting them to be present at the study area As present following formula:

∑

Oi: observed value; Ei: expected value; Df: degree of freedom

Differences between wild-caught and traded samples of G spengleri in terms

of body size measurements (carapace length, carapace width, shell height and weight) were assessed by Student t-test In the text, the means are presented ± 1 Standard Deviation All statistical analyses were performed by Past 3.0 software, with alpha set at 5% Follow formula:

√ t- student

mA and mB represent the means of measurements

nA, nB represent for the size of observations

S2 is an estimator of the common variance of the two samples

IV RESULTS

4.1 Tay Yen Tu NR’s turtle fauna composition and status

At least 10% of the interviewees identified the presence of seven species, whereas for other seven species the number of people claiming their presence was very low and thus considered unlikely (Appendix 1) The various species differed significantly in terms of frequency of the interviewees reporting them to be present

(χ2= 133.6, df = 11, P < 0.0001), with significantly more interviewees knowing well

G spengleri and Cuora mouhotii Interestingly, no species was confirmed to be

present by more than 50% of the interviewees (Appendix 1), thus indirectly

suggesting that all species should be rare in the study area Concerning G spengleri,

the various interviewees concurred that the species weighed less than 100g, has serration, lives in bamboo forest (they called it „the bamboo turtle‟ (rùa trúc)) and, because of the reddish skin coloration, is also called the ‟fire turtle‟ (Rùa lửa) According to their information, the species inhabit the evergreen forest at high elevation, even on the top of the mountains Hunters also claimed that the species is normally seen when weather change from rain to sunny or vice versa, and that they can be located more easily in rainy season (April to July) when turtles are very active above-ground while searching for the food Based on the interviewees‟

reports, it is reported that G spengleri was the cheapest of all the locally traded

species (Appendix 1)

During these interviews, we observed 40 individuals, belonging to 7 native

species (G spengleri, C mouhotii, C cyclornata, C galbinifrons, I elongata, P steindachneri, and P sinensis) in the hands of the hunters/traders (Table 4 and Appendix 6, 7 and 8) The frequency of occurrence of the various species was

significantly different from equality (χ2

= 142.2, df = 6, P < 0.0001), with G spengleri, C mouhotii being the two dominant species Overall, 57.5 % of the

observed individuals were going to be sold to China, followed by the local consumption as food (17.5%), traditional medicine (15%) and pet (12.5%)

(Appendices 1 & 11) In our observed sample, 6 individuals of three species (one G spengleri, four C mouhotii and one I elongata) were confiscated by rangers of the

NR (Table 4) Body size measurements taken from both wild-caught and traded turtle individuals are reported in the Table 5 All the observed species were listed as

Threatened species by IUCN (2018) Of those, two species (C cyclornata, C galbinifrons) occupied 28.57% (n=7) were fully protected by the strong national

law Decree 160/2013/NĐ-CP While Decree 32/2006/NĐ-CP is also listed two

species in its protection list (C cyclornata, I elongata) Six species (except P sinensis) out of 7 species were listed on appendix II of CITES (2017)

Table 4 Synopsis of the places where turtle individuals were observed during the

interview surveys

Species

Trader's house

Hunter's house

Confiscated by rangers

Other five sympatric species (P megacephalum, S quadriocellata, M mutica,

C zhoui, C oldhamii) were also described by local people, but there was no field

record to confirm the presence of those species

of adult male C mouhotii is also found at 607m a.s.l at the forest path with forest cover is 50% Their relative frequency of observation turtle during the field survey

was 0.037 individual/person/hour

Fig 5 Map of Vietnam, showing the position of the interview points and of

the presence sites of turtles at the Tay Yen Tu NR The field survey tracks

are also indicated

4.2.2 Species ecology

4.2.2.1 Species micro-habitat uses

Out of seven free-ranging G spengleri, only one individual was found by a hunting

dog The turtles were found from 10:47-13:45 hour, in all cases when the weather was sunny following heavy rains Turtles were observed at 25-300C temperature range (Appendix 3) Three turtles were found in rocky caves while four turtles were

found in a forest patch dominated by bamboo (Arundinaria sp.) (Figure 6)

The individuals found outside of rocky caves were not too far from a rocky area (range from 10-50m) when firstly seen (Appendix 4, top and below) All turtle found were not hiding under leaf litter (Appendix 4 top) The forest cover at the

location that G spengleri ranged from 70%-90% (mean =76.4 ± 7.4 %, median =

75%), thus showing that this species inhabits dense forest patches However, there

Fig 6 Bamboo forest (Arundinaria sp.) where is the „favorite‟ micro-habitat

for G spengleri in Tay Yen Tu NR

were no turtles found in wet forest areas with canopy over higher than 90% The slope angles where turtles were found varied considerably, from 5 to 45 degrees, with an average of 24.6 ± 17.5° (median = 25°)

Fig 7 Black-breasted leaf turtle (G spengleri) in the wild

(male at above, and female at below)