外來龜種對金門地區原生金龜的遺傳入侵

全文

(1)國立臺灣師範大學生命科學系碩士論文. 外來龜種對金門地區原生金龜的遺傳入侵 Genetic introgression from invasive Mauremy species to native M. reevesii in Kinmen. 研究生：李昱 Yu Lee. 指導教授：林思民博士 Si-Min Lin. 中華民國 104 年 6 月.

(2) Table of contents 致謝 ......................................................................................................... 1 中文摘要 ................................................................................................. 3 Abstract ................................................................................................... 5 Introduction ............................................................................................. 7 Materials and Methods .......................................................................... 11 Result ..................................................................................................... 17 Discussion .............................................................................................. 21 References ............................................................................................. 25 Table ...................................................................................................... 32 Figure ..................................................................................................... 36 Appendix................................................................................................ 44.

(3) 致謝完成這篇金龜的研究，要感謝許多人的協助與幫忙。首先感謝林思民老師對我的指導，老師平心靜氣的耐心教導，使我得以完成我的研究。從一個近乎完全不懂研究為何物的人，到今日論文的完成，這是因為拎老師不離不棄的耐心指教阿! 金門金龜的研究先要感謝安東尼和恰恰打下的基礎，也是我研究的開端。也因為有陳添喜老師的幫忙，讓我更熟悉烏龜的調查。再來也要感謝金門縣政府的協助，立偉、光耀與愛瓊姐的支持使得實驗進行的如此順利。金門的調查生活很有趣，因為有金門隊長阿平帶著我東吃西逛，我才能比較快的在金門的烏龜調查中進入狀況；後來在金門的生活裡真威的曾威也是我的好夥伴，我們真是在金門度過了一段很愉快的時光。在草魚實驗室的這幾年，是相當愉快的時光。實驗室的各位不僅是在研究上互相幫忙、討論的夥伴，也是生活中討論各項議題的對象在實驗室內，感謝大學長彥博和野性的阿傑、開直升機的芳神協助，我的分生實驗並不熟練，而每當我有問題的時候你們總是耐心的教導我並幫我解決問題。無聊大師兄展蔚的統計分析和科學怪人的書書在遺傳分析的威能，是我在研究中最大的兩個後盾，感謝你們，讓我在研究中出現的問題都如有神助的被解決掉。實驗室的諸位，聲音存在感強烈的蟲子、勤力健身的閣閣、變胖又變瘦的甩阿、實驗非 1.

(4) 常努力的 A4、提供笑料的阿如、詢問八卦的阿寶，都是我在研究生活中珍貴的好朋友，改天我們在一起回到沙巴去。最後要感謝我的家人與涵瑜，親朋好友、兄弟姊妹對我的研究的支持與鼓勵，妹妹現在也都走在研究的道路上。而涵瑜的陪伴也讓我在研究最後的生活豐富而又有趣。老媽老爸盡心盡力的為我付出提供了我穩定的生活。他們的行動總是打從心底支持我。不僅讓我在研究所求學時沒有後顧之憂，也從未要求回報。我深深的感受到父母對我的愛，我愛你們。. 2.

(5) 中文摘要金門是金龜（Mauremys reevesii）在台灣唯一擁有穩定族群的棲地，金龜也是島上除了中華鱉之外最優勢的原生龜種。然而隨著寵物貿易的流通，也可能對當地的龜鱉類群聚產生相當程度的改變。我們在金門島上 2011 至 2013 年進行金門的淡水龜普查中，不僅發現了過去在當地較為罕見的斑龜（M. sinensis），還觀察到疑似金龜與斑龜、金龜與柴棺龜雜交的個體。這些疑似雜交的表形顯示在生殖隔離不完全的情形下，不同種的淡水龜可能發生雜交的現象。若雜交的後代仍具有生殖能力，這些近緣龜種可能會對金龜產生遺傳入侵（genetic introgression）的負面效應。為了瞭解斑龜與柴棺龜是否會與原生的金龜雜交，並評估遺傳滲入的嚴重程度，找出影響池塘雜交程度的因子，我們採用淡水龜的表形、粒線體 cytochrome b 基因序列、與 13 組微衛星基因座進行遺傳滲入的檢測，並用池塘的環境資料：距主要道路距離（DM）、距次要道路距離（DS）、植被覆蓋程度（VL）、池塘水面上植物量（VW）、池塘邊緣形式（PS）、池塘面積大小（PA）、池塘深度（PD）以及池塘內斑龜的數量等因子，利用廣義線性迴歸分析（general linear regression）檢視上述各項棲地因子是否對金龜的雜交程度造成影響。結果顯示金門的金龜確實會與斑龜和柴棺龜雜交。從粒線體基因與表形的比較結果來看，表形與基因型不一致的金龜共 3.

(6) 有 10 隻，其中有 6 隻金龜帶有斑龜的基因型；有 4 隻帶有柴棺龜的基因型。利用微衛星基因座，並使用 STRUCTURE 分群判別，在全部使用純系龜隻的情況下，三種淡水龜可以清晰的分群，最佳分群數為三。但是當加入金門當地的金龜之後，最佳分群數為二。表形初步判定為雜交的 19 隻淡水龜，有 18 隻確認為斑龜或柴棺龜和金龜的雜交龜。而表形看似金龜的 241 隻個體當中，有 28 隻遭到斑龜或柴棺龜的遺傳入侵。總和來說，原生金龜有高達 11.6%的比例與近緣龜種雜交。採樣的 44 個樣點裡有 22 個樣點採集到金龜，而其中有 16 個水域（72.7%）有雜交的現象產生。影響池塘雜交的程度則與共域的外來斑龜數量與池塘的人工化程度有關，斑龜的數量與池塘雜交的程度呈現顯著的正相關（P=0.0002），天然的池塘受到遺傳入侵的程度也顯著地低於人工池塘（P=0.037）。研究顯示這三種已經分化的淡水龜若受到人類行為的干擾（例如異種共同圈養，或是人為進行的釋放），很可能會提升種間雜交的機會，對物種的保育有不利的影響。. 關鍵詞：金龜，金門，停戰區，遺傳入侵，野生動物貿易. 4.

(7) Abstract Kinmen Archipelago, comprising Greater and Lesser Kinmen islands, is located on the coastline of China and sustaining a stable population of Chinese pond turtle (Mauremys reevesii). Since this turtle is widely consumed for food and medical purposes, Kinmen has become one of the most important natural habitats for the conservation of this species in Taiwan. However, community of turtles in Kinmen Islands might have changed in recent years because of human activities. Chinese stripe-necked turtle (M.sinensis), which was extremely rare in literature, was found on the islands; while genetic introgression was suspected from M. sinensis and the even rarer Asian brown pond turtle (M. mutica). In this study, we aim to evaluate the level and magnitude of genetic introgression among closely related turtles by using mitochondrial cytochrome b sequences and 13 microsatellite loci. We further used 8 environmental and biological factors of the ponds to conduct general linear regression in order to figure out the critical factors which influence the hybrid rates of the turtles. Our results indicated strong evidences for the hybridization between M. reevesii and M. sinensis, and occasionally between M. reevesii and M. mutica. Six turtles showed inconsistence between their M. reevesii morphology but M. sinensis mitochondrial haplotypes; while 4 others showed M.reevesii morphology but M. multica haplotypes. Microsatellite genotyping further revealed 18 among 19 turtles with intermediate form and 28 among 241 M.reevesii individuals confirmed as hybrids between M.reevesii and M. sinensis or M . mutica. Among the 44 water bodies we sampled, 22 localities have M. reevesii 5.

(8) records and 16 (72.7%) have hybrids detected. The degree of introgression in a pond is positively correlated to the abundance of M. sinensis (P = 0.0223) and the artificial modification of the pond (P = 0.0002). Our research shows that human mediated dispersal of Mauremys and destruction of natural habitats may increase the possibility of interspecies hybridization, and cause negative impact to the endangered species. Keywords: Mauremys reevesii, Kinmen, demilitarized zone, genetic introgression, wildlife trade. 6.

(9) Introduction The invasion of the alien species is believed to cause negative impact to native ecosystem. The interaction between introduced species and native species may influence local community by competition, predation, introduction of novel pathogen, and genetic introgression (Lockwood et al. 2013). The genetic introgression between native species and closed relative invasive species may result in the loss of local genetic variation. Mallard ducks (Anas platyrhynchos) is the classic example of human-mediated dispersal causing hybridization between invasive and local species. The introduction of the domestic Mallard ducks result in genetic introgression to several closely relative species which used to be endemic to local area such, as New Zealand Grey Ducks (Anas superciliosa), Eastern Spot-billed Ducks (Anas zonorhyncha), and American black Ducks (Anas rubripes) (Gillespie 1985, Kulikova et al. 2004, Mank et al. 2004). Another case of human-mediated hybridization is the introduction of green iguana (Iguana iguana) in Lesser Antilles, which has caused genetic introgression to the Lesser Antilles iguana (Iguana delicatissima) and made profound implications to the conservation of this highly endangered iguana (Knapp et al. 2014). Hybridization among species seems to be a process more common than previous expectation during Testudines speciation (Stuart and Parham 2007). Ancient alleles of extinct species have also been found in living descent species (Poulakakis et al. 2008). With the ability to hybridize among distantly related lineages (Spinks et al. 2004, Buskirk et al. 2005), Testudines tend to have more serious problem facing human-mediated 7.

(10) hybridization issue. Although most cases of hybridization in Geoemydidae turtles were reported from human-made environment such as turtle farms and restricted artificial lake, some records have been reported from the wild (Otani 1995, Shi et al. 2005, Fritz et al. 2008, Fong and Chen 2010, Fujii et al. 2014). Nevertheless, literature discussing hybridization among Testudines are mostly case studies. This phenomenon has not been thoroughly discussed neither by multiple nuclear markers, nor in the aspect of conservation biology. Reeves’ Turtle (or Chinese Three-keeled Pond Turtle, Mauremys reevesii) is one of the native turtles in Taiwan. It is an aquatic species with a maximum size to 300 mm in carapace length. Mauremys reevesii is widely distributed in East Asia including central and eastern China, southern Japan, and south part of the Korean. The turtle lives in freshwater habitats with still or slowly moving water in lowland areas (Lovich et al. 2011). Although it is widely distributed in East Asia and the native distribution has been extended by human-aided translocations (Suzuki et al. 2011) , the commercial use for food supplement and medical demand of China have resulted in overhunting and habitat destruction in the wild population (Haitao et al. 2008). Most of the M. reevesii populations have experienced major declines because of habitat destruction and commercial over-exploitation. Nowadays, this species has been protected in Taiwan (Chen and Lue 2010) and listed as endangered in the IUCN Red List of Threatened Species and AppendixⅢof CITES.. 8.

(11) Owing to the reduction and degradation of freshwater wetlands, natural population of M. reevesii in Taiwan has faced dramatic population decline. Kinmen Archipelago became the only place to sustain a stable population of the turtle. These islands, located off southeast China coastline, are used to be the military front line between China and Taiwan. Though the total area of these islands is only 150 square kilometer, military in Kinmen has produced hundreds of pond and lakes since the requirement of water is so important for the supplement of the army during the Cold War (Chang 2005). As the tension situation between China and Taiwan mitigated, Kinmen become demilitarized zone and the natural environment maintained a lower degree of development (You et al. 2013; Lin et al. 2015). These water regions turned to be habitats for species rely on water region such as turtles and Otters (Lin et al. 2015). These water ponds made for human need and the located of demilitarized zone provide an excellent habitat for M. reevesii. An early investigation on amphibians and reptiles of Kinmen showed that there was only one turtle, M. reevesii, in Kinmen islands. However, the investigation of local turtles in Kinmen we conducted in 2011 shows the community of turtles in Kinmen Islands might have changed in recent years because of human activities. The investigation shows that the turtles in Kinmen includes Trachemys scripta elegans, and Chinese stripe-necked turtle (M.sinensis), which was extremely rare in literature, was found on the island. We had also found some Mauremys turtles which represented intermediate form between different species (Fig. 1). We suspect that theses turtle might possibly be hybrids between M. reevesii and closely relative 9.

(12) Mauremys turtles. Since Kinmen is such an important habitat for M. reevesii, to protect the turtles in Kinmen is necessary. If closely relative species may cause hybridization problem to M. reevesii, it would very important to figure out the how serious the problem is for conservation. In this study, we aim to study the relationship between the predict hybrid turtles and evaluate the degree of hybridization of the ponds in Kinmen by using mitochondrial and nuclear genetic markers. We aim to answer the following question below: (1) To detect probable hybrids by using molecular tools. (2) To evaluate the level and magnitude of hybridization. (3) To figure out the habitat with high hybrid rates. 10.

(13) Materials and Methods Turtle sampling We conducted our sampling on Kinmen Archipelago (24.386° to 24.529°N, 118.210° to 118.473°E) during 2011 to 2013, including 35 localities from Major Kinmen and 7 from Lesser Kinmen islands (Fig. 2, Table 1). Habitat of sampling regions included ponds, marshes, reservoirs, and rivers. Turtle were collected with funnel traps, using sardine cans as bait. Empty plastic bottles were put inside each trap to prevent the trapped turtles from drowning. Three to five traps were placed in each water region for 15 days, checked every day and renewed for fresh baits. No turtle was drowned or injured inside the traps during sampling. Tissues were collected from webbed toes and placed in 95% ethanol. Weight, carapace length, plastron length and sex (if available) were recorded from each turtle. Turtles which are recently introduced to Kinmen, such as Mauremys sinensis, Mauremys mutica, Trachemys scripta, and suspected hybrids were removed from the wild and kept in long-term captivity. Others were released to the same site after receiving individual marks from their marginal scutes (Cagle 1939). We also obtained pure-lined turtles and turtles with confirmed hybrid status F1 to help us recognized the phenomenon of hybridization. We collected 12 M. sinensis tissue from Taiwan, 11 M. reevesii tissue from China and 9 M. mutica tissue from Taiwan as pure line. We also bring in 2 samples from turtle farm which was confirmed as F1 hybrid between M. reevesii and M. sinensis.. 11.

(14) DNA extraction Genomic DNA was extracted by using EasyPure Genomic DNA spin kit (Bioman) according to manufacture indication. DNA was stored in elution buffer at -20 ∘C until polymerase chain reaction (PCR). We use one mitochondrial and thirteen microsatellite markers to evaluate the status of hybridization. (Table 2). Mitochondrial DNA sequencing Mitochondrial DNA cytochrome b was amplified by using the primers mt-A (forward 5’-CAACATCTCAGCATGATGAAACTTCG-3’) and H15909 (reverse 5’-CAGTTTTTGGTTTACAAGACCAATG-3’), with annealing temperature 50∘C (Barth et al. 2002). The PCR were set in a volume of 20μl containing 10μl 2X GoTaq® Green Master Mix Buffer, 0.1-0.5 ng of template DNA, 0.25μM of front and reverse primer and 8μl of ddH2O. The PCR condition were set as below: 1. Denature at 95 ∘C for 3 min. 2. Cyclic run 35 cycles at 95 ∘C for 30 seconds, at annealing temperature for 40 seconds(Each Tm are list in Table 2) , extension at 72 ∘C for 40 seconds 3. Final cycle at 72∘C for 10 min After PCR, products were electrophoresed in a MegaBASE with 1.25% agarose gel in 1X TBE buffer to ensure the length of the product were the correct fragment. DNA genotyping were carried out by Genomics BioSci & Tech Corp. (Taipei, Taiwan). 12.

(15) Automatic PCR products sequencing were carried out on ABI 3730 auto-sequencing machine by Genomics BioSci & Tech Corp. (Taipei,Taiwan). Sequence for cytochrome b was proofread and assembled by Sequencher v4.9 software (GeneCode, Boston, MA, USA) with the default settings.. Microsatellite genotyping Genotyping of microsatellite were checked by Peak Scanner Software v1.0 to confirm the peak of each marker. We use MicroChecker (Van Oosterhout et al. 2004) to compute the expected frequency of heterozygous and observed frequency of heterozygous in each locus in the pure population of M. reevesii and M. sinensis, to check if the present of null allele as the result of Hardy-Weinberg inequilibrium.. Data analysis Haplotype reconstruction The Mauremys turtles collected in Kinmen were compared with pure-lined individuals collected in Taiwan and China. We used mitochondrial cytochrome b gene as marker to exam if the maternal genotype fit with turtle’s morphology as a preliminary data for possibly hybrid specimens. We used cytochrome b sequences downloaded from GeneBank include nine Mauremys and seven Cuora spp. as outgroup to construct the phylogenic tree among these closely related species. Total of 286 individuals with 813bp were sequenced and aligned. We use MEGA 6 (Tamura et al. 2013) to align the sequences, use DnaSP v5 (Librado and 13.

(16) Rozas 2009) to generate the individuals data to haplotype data. The best model was selected by MEGA 6 and used to construct the ML haplotype tree. The best substitution model selected by MEGA 6 was HKY+G (BIC =4091.506), clade support was examined by non-parametric bootstrap analyses (1,000 replicates) summarized with 50% majority rule consensus trees. After analysis, we use FigTree 1.4.0 to modify and exported the ML tree.. Hybridization status assess We used 13 sets of microsatellite loci data to assess the status of hybridization within each individual. We used a discriminant analysis of principal component (DAPC) (Jombart 2013) implemented in the adegenet package available for R version 2.15.2 to investigate the population genetic structure between populations within species. We predefined all the individuals into 7 population: M. reevesii from China (MS), M. sinensis from Taiwan (MR), M. mutica from Taiwan (MM), hybrids between M.reevesii and M.sinensis from Taiwan’s turtle farm (TW), M. reevesii from Kinmen (KMR), M. sinensis from Kinmen (KMS) and intermediate form turtles from Kinmen (HY). We ran DAPC by performing a principal component analysis (PCA) equal to 50 which retaining 80% of the variation in the original data. A discriminant analysis (DA) was then applied with 6 discriminant functions that carried most information. Relationships between clusters from DAPC were visualized using scatterplot. We used STRUCTURE 2.3.3 (Hubisz et al. 2009) to assign each individual to species by genotype of 13 microsatellites for 313 individuals 14.

(17) with no priori assumption to check the introgression within turtles. We set the number of clusters from one to five (K = 1-5) and made 4 independent runs by 200,000 Markov Chain Monte Carlo iterations and burn-in period of 100,000 steps. The best number of cluster was indicated by the higher log likelihood value and was chosen by STRUCTURE Harvester (Earl 2012). The most likely K with 4 runs was chosen for individual assignment to species and identification of putative hybrids.. Genetic introgression assess We regard each water region we sampled as a population to calculate the genetic introgression between M. sinensis to M. reevesii and M. mutica to M. reevesii at the level of different population. The program BayesAss was used to test contemporary gene flow between the different populations. We use M. sinensis from Taiwan and Kinmen as the population of M. sinensis (N=38); M. mutica from Taiwan as the population of M. mutica. With 13 locus microsatellite genotype data and Bayesian framework, we estimate recent migration rate (within two generations) from M. sinensis and M.mutica to each pond (Wilson and Rannala 2003). We removed the pond which turtle sampled less than 5 individuals. The program was run for 200,000 MCMC steps, with 100,000 steps as the burn-in and a sampling frequency of 2,000. The δ values for migration rate, and inbreeding values were set to 0.10. We use the migration rate between M. sinensis from Taiwan and M. reevesii from China as the standard of no genetic introgression to evaluate the magnitude of genetic introgression between different water regions in Kinmen. The result of BayesAss showed in Table 15.

(18) 1 and use as the genetic introgression degree from M. sinensis and M. mutica to M. reevesii in each pond. We used JUMP 7 to do statistical analysis with eight habitat factors to build regression models with degree of introgression to M. reevesii. Use forward stepwise to choose the potential influence variable to the genetic introgression degree. These factors were clearly defined in Table 3 and were used to predict the presence and abundance of M. reevesii. Three among these factors, shortest distance to main roads (DM), shortest distance to secondary roads (DS) and the amount of M. sinensis in each pond (MS) were treated as contiguous variables. The others, including level of vegetation coverage on the land (VL), level of vegetation coverage in the water (VW), substrate type around of the pond (PS), area of the pond (PA), depth of the pond (PD), were categorized into ordinal parameters. Habitat characters of each sample site were recorded by practiced researchers at the start of capture occasion.. 16.

(19) Result Turtle sampling The total number of turtles sampled in Kinmen Islands is 288. Based on morphology, they were preliminarily identified as 241 M. reevesii, 28 M. sinenesis and 19 turtles with intermediate form. The name and sample size of each pond was listed in Table 1.. Hybrid identification by mtDNA sequences The maximum likelihood (ML) tree of mitochondrial cytochrome b sequences (813 bp from 286 individuals) has showed a preliminary result of hybridization among Mauremys spp. (Fig. 3). Phylogeny among the species showed that all the pure-lined M. reevesii, M. sinensis and M. mutica formed monophyletic clades with high statistic supports, and all the turtles from Kinmen were designated into these three clades. However, one turtle with M. sinensis morphology fell into the clade of M. reevesii (haplotype 1), revealing the maternal heredity from the latter. Six turtles with M. reevesii morphology fell into the clade of M. sinensis (haplotypes 5, 7, 9, 13, 14), while four of M. reevesii morphology designated into the clade of M. mutica (haplotypes 4, 6 and 8). The turtles with intermediate morphology fell in to either M. reevesii or M. sinensis clades. The inconsistency between morphology and mitochondrial DNA sequences indicated the occurrence of hybridization.. Microsatellite analysis of pure-lined turtles by STRUCTURE The best cluster K was calculated by STRUCTURE harvester with 17.

(20) K=3 (ΔK=769.803) when only pure-lined individuals were used in STRUCTURE analysis (Appendix 2a). Mauremys reevesii (N = 9) belongs to cluster 1 (blue) with probabilities ranging from 0.981 to 0.994 (mean: 0.989, sd: 0.0045). Mauremys sinensis (N = 12) belongs to cluster 2 (green) with probabilities ranging from 0.96 to 0.965 (mean: 0.986, sd: 0.012). Mauremys mutica (N = 9) belongs to cluster 3 (yellow) with probabilities ranging from 0.965 to 0.994 (mean: 0.986, sd: 0.009). Analysis from STRUCTURE revealed a clear identification among the three species by using genetic data.. Microsatellite analysis of Kinmen’s turtle by STRUCTURE Out of our expect, the best cluster calculated by STRUCTURE HARVESTER was K= 2 (ΔK=8496.86) instead of K =3 (ΔK = 35.04) when all individuals were included in the analysis (Fig. 4, Appendix 2b). According to the two-clusters identification (K=2), STRUCTURE clearly identifies Mauremys reevesii from Mauremys sinensis and Mauremys mutica while M. sinensis and M. mutica could not be clearly identified (Fig. 5). Mauremys reevesii (N = 11) has probability belonging to cluster 1 (blue) ranging from 0.885 to 0.994 (mean: 0.963, sd: 0.037). The two confirmed F1 hybrids between M. sinensis and M. reevesii from the turtle farm have probability belonging to cluster 1with 0.598 and 0.415 respectively. Pure-lined M. sinensis (N = 12) has probability belonging to cluster 2 (green) ranging from 0.911 to 0.995 (mean: 0.98, sd: 0.026). Pure-lined M. mutica (N = 9) has probability belonging to cluster 2 (green) with one individual is 0.603 while remain ranging from 0.883 to 0.992, the total 18.

(21) mean =0.98 and sd= 0.122. This result indicated that M.sinensis and M. mutica could not be clearly distinguished by using markers we used in this study. Although three pure species can be clearly divided to three cluster themselves, the possibly homoplasy in microsatellite between M. sinensis and M. mutica might result to the difficulty of separating between these two species. Nevertheless, the result represented high resolution to detect the introgression from introduced the two invasive Mauremys toward M. reevesii. There are 242 M. reevesii individuals in Kinmesn, 31 individuals are considered as hybrids with the degree assigned to cluster 1(blue) lower than 90%; 19 turtles with intermediated form has one individual assign to cluster 1 (99.3%), the remainder with 15 individuals would consider as hybrids ranging from 0.129 to 0.687 to cluster 1. All 27 individuals of M. sinensis collected from Kinmen belongs to cluster 2 (green) ranging from 0.944 to 0.995 (mean: 0.99, sd: 0.014).. DAPC result of microsatellite analysis The results of DAPC were conducted in Fig.6. We predefined all the turtles into 7 different category: 3 pure Mauremys turtles (represent three species abridge as MR, MS and MM), M. reevesii collected from Kinmen (KMR), M. sinensis collected from kinmen (KMS), intermediate forms (HY), and the confirmed F1 hybrid from turtle farms (TW). The three pure species were clearly divided into three distantly separated clusters, while F1 hybrids of M. reevesii and M. sinensis (TW) position in the middle between the two. The intermediated formed turtle (HY) and some of the M. reevesii from Kinmen (KMS) were also positioned between M. reevesii and M. 19.

(22) sinensis pure species (MR and MS). This result suggested that the hybridization event mainly occurred between these two species, while the genetic introgression from M. mutica is less detected from population in Kinmen.. Genetic introgression assess Genetic introgression from invasive to native species was calculated and listed in Table 1. Significant correlations between genetic introgression from M. sinensis and habitat characters appeared in 2 pairs of habitat characters, substrate type around of the pond (P = 0.0002) and the amount of M. sinensis in each pond (P = 0.0223) in step wise variable selection (Table 3). We use the residuals of genetic introgression degree against to the amount of M. sinensis in each pond, then use Wilcoxon analysis to know whether artificial and natural side type have different genetic introgression rate (Fig. 7). The genetic introgression degree significantly higher for the artificial habitat than natural habitat (Wilcoxon signed rank test, two-tailed, Z = 2.08; p < 0.05). This result indicated that the degree of introgression in a pond is positively correlated to the amount of M. sinensis and the artificial degree of the pond.. 20.

(23) Discussion Our results indicated that the hybridization among turtles in Kinmen Islands was caused by the genetic introgression from two closely relative species, M. sinensis and M. mutica, to native M. reevesii. Among the 288 turtles collected from Kinmen, 55 turtles had been confirmed to be hybrids through the inconsistence between cyt b haplotype and morphology, or the genotypic analyses from microsatellite markers. Six turtles showed inconsistency between their M. reevesii morphology but M. sinensis mitochondrial haplotypes; while 4 others showed M. reevesii morphology but M. mutica haplotypes. Microsatellite genotyping further revealed that 15 among 19 turtles with intermediate form and 31 among 241 M. reevesii individuals were confirmed as hybrids between M. reevesii and M. sinensis or M. mutica. It is quite interest that all of the turtles with M. sinensis morphology are genetically recognized as pure M. sinensis, yet 31 out of 241 turtles with M. reevesii morphology are identified as genetic hybrids. This results indicated the fact that genetic introgression to M. reevesii is hard to be detected solely by morphological data. Furthermore, M. sinensis morphology are temp to be influence by hybridization easily.. Genetic introgression caused by M. mutica The hybridization event produced by M. mutica is somehow out of our expectation, since we have never sampled any individual of this species from Kinmen. However, 3 individuals showed intermediate form between M. miltica and M. reevesii; while 4 individuals showed the cyt b from M. mutica. The ponds which have individuals hybrid with M. mutica are 21.

(24) Long-Ling, Xi-Cun, Shuang-Li-Hu and Shuang-Li wet land, which locate at East, Southeast, Northwest and Northwest of Kinmen respectively. We cannot find any habitat factors that would influence the degree of M. mutica hybrids. Possible reason for the result is the ponds highly introgression by M. mutica didn’t have enough individuals samples such as Xi-Cun (N=2) and Long-Tang (N=3) were not enough as population for introgression test and were not included in the analysis. With lower estimation of genetic introgression from M. multica, this problem should be carefully monitored in the near future.. Genetic introgression caused by M. sinensis The result of DAPC showed that most of the hybridization happened between M. sinensis and M. reevesii because all the turtles in Kinmen distribute between these two species (Fig. 6). Mauremys sinensis has been reported to hybrid with M. reevesii not only in captive pet market (HE et al. 2011, Xia et al. 2013) but also in wild survey in Taiwan (Fong and Chen 2010). Based on our results, 16 out of 22 ponds sampled in Kinmen were genetically introgressed caused by M. sinensis or M. mutica (Fig. 8). Ten of these ponds have M. sinensis records and another one pond has the intermediate form turtle carry M. sinensis cyt b sequence. In the analysis to the magnitude of genetic introgression to related habitat factor showed the significant correlation between the magnitude of genetic introgression and the artificial level or the amount of M. sinensis in the pond. The amount of M. sinensis positively correlated to the introgression degree is considerably make sense and confirm to the suspect 22.

(25) since hybridization should occur through the emerge of two species. On the other hand, the water regions with high level of artificial degree have similar characters with deep water and steep banks which is unfriendly for turtles. The water regions with higher level of genetic introgression from M. sinenesis include Qiong-An and Shui-Tao, while these places are artificially modified by concrete water bank with an over 2 meter height and are difficult for turtle to climb out. In contrast of the other drill, Guan-Lu-Bian , which has a natural slope on the drill side not only didn’t sampled so many M. sinensis individuals (N=1) but also rare in the detect of genetic introgression from M. sinensis. One of the possibly explanation is the water region with higher degree of artificial modification happened to be the habitat favored by M. sinensis, in which would cause higher degree of genetic introgression to M. reevesii. The second explanation is this result implies that the isolation of the water region may strongly influence the degree of introgression from alien turtles to M. reevesii. The water region with high border which confined the movement of turtles may force the introduced turtles from migrating to meet another individual result in an unnatural condition which M. sinensis has no choice but hybrid with M. reevesii while facing dramatically bias of species amount. This phenomenon has been observed in the pond with highest introgression degree, Ming-Tan. Ming-Tan locate in the middle of Kinmen Island and is surrounding by a closed environment. The pond is at the edge of military facilitate, west and south side of the pond is roads 4 meters higher than the pond while north side is the Tai Wu mountain which is a mountain consist of granite. The turtle of the ponds seems to be highly introgression by M. 23.

(26) sinensis since half of the turtles sampled inside are hybrids.. Conservation suggestion for M. reevesii in Kinmen Among the 44 water bodies we sampled, 22 localities have M. reevesii records and 16 (72.7%) have hybrids being detected. The introduction of M. sinensis and M. mutica supposed to be come from pet market from Taiwan since they have never been found in the former investigation. Kinmen is one of the most important habitats in the world for the conservation of M. reevesii with lots of natural habitat and stable wild population with sufficient amount. Long term investigation for the population dynamic of turtles in Kinmen is necessary. The construction of the future water region should also avoid the side with highly artificial level which not only provide a better habitat for M. reevesii but also reduce the risk for genetic introgression from M. sinensis to M. reevesii.. 24.

(27) References Blacket, M. J., C. Robin, R. T. Good, S. F. Lee and A. D. Miller (2012). "Universal primers for fluorescent labelling of PCR fragments--an efficient and cost-effective approach to genotyping by fluorescence." Molecular Ecology Resource 12(3): 456-463. Barth, D., D. Bernhard, D. Guicking, M. Stock and U. Fritz (2002). "Is Chinemys megalocephala Fang, 1934 a valid species? New insights based on mitochondrial DNA sequence data." SALAMANDRA-BONN- 38(4): 233-244. Buskirk, J. R., J. F. Parham and C. R. Feldman (2005). "On the hybridisation between two distantly related Asian turtles (Testudines: Sacalia x Mauremys)." SALAMANDRA-BONN- 41(1/2): 21. Cagle, F. R. (1939). "A system of marking turtles for future identification." Copeia 1939(3): 170-173. Chang, C.-M. (2005). "The analysis on the water resource development strategy in Kinmen-A viewpoint of public value." National Sun Yat-sen University Thesis, Kaohsiung, Taiwan. Chen, T.-H. and K.-Y. Lue (2010). "Population status and distribution of freshwater turtles in Taiwan." Oryx 44(02): 261-266. Earl, D. A. (2012). "STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method." Conservation Genetics Resources 4(2): 359-361. Fong, J. J. and T.-H. Chen (2010). "DNA evidence for the hybridization of wild turtles in Taiwan: possible genetic pollution from trade animals." Conservation Genetics 11(5): 2061-2066. 25.

(28) Fritz, U., D. Ayaz, J. Buschbom, H. Kami, L. Mazanaeva, A. Aloufi, M. Auer, L. Rifai, T. Šilić and A. Hundsdörfer (2008). "Go east: phylogeographies of Mauremys caspica and M. rivulata–discordance of morphology, mitochondrial and nuclear genomic markers and rare hybridization." Journal of Evolutionary Biology 21(2): 527-540. Fujii, R., H. Ota and M. Toda (2014). "Genetic and Morphological Assessments of Hybridization Between Two Non-Native Geoemydid Turtles, Mauremys reevesii and Mauremys mutica, in Northcentral Japan." Chelonian Conservation and Biology 13(2): 191-201. Garrick, R. C., E. Benavides, M. A. Russello, J. P. Gibbs, N. Poulakakis, K. B. Dion, C. Hyseni, B. Kajdacsi, L. Marquez, S. Bahan, C. Ciofi, W. Tapia and A. Caccone (2012). "Genetic rediscovery of an 'extinct' Galapagos giant tortoise species." Current Biology 22(1): R10-11. Gillespie, G. D. (1985). "Hybridization, introgression, and morphometric differentiation between Mallard (Anas platyrhynchos) and Grey Duck (Anas superciliosa) in Otago, New Zealand." The Auk 102(3) : 459-469. Guillon, J.-M., L. Guéry, V. Hulin and M. Girondot (2012). "A large phylogeny of turtles (Testudines) using molecular data." Contributions to Zoology 81(3): 147-158. Haitao, S., J. F. Parham, F. Zhiyong, H. Meiling and Y. Feng (2008). "Evidence for the massive scale of turtle farming in China." Oryx 42(01): 147-150. Haskell, A. and M. Pokras (1994). "Nonlethal blood and muscle tissue collection from redbelly turtles for genetic studies." Herpetological 26.

(29) Review 25(1): 11-11. Hasselman, D. J., E. E. Argo, M. C. McBride, P. Bentzen, T. F. Schultz, A. A. Perez-Umphrey and E. P. Palkovacs (2014). "Human disturbance causes the formation of a hybrid swarm between two naturally sympatric fish species." Molecular Ecology 23(5): 1137-1152. He, G., L. He, C.-l. Fang, Y.-F. Xu, D.-Q. Wang and C.-P. Fei (2011). "Morphologic characters of hybrid (Chinemys reevesii♀× Ocadia sinensis♂) and comparison with their parents." Journal of Nanchang University (Natural Science) 6: 012. Hubisz, M. J., D. Falush, M. Stephens and J. K. Pritchard (2009). "Inferring weak population structure with the assistance of sample group information." Molecular Ecology Resource 9(5): 1322-1332. Jiang, Y., L. W. Nie, Z. F. Huang, W. X. Jing, L. Wang, L. Liu and X. T. Dai (2011). "Comparison of complete mitochondrial DNA control regions among five Asian freshwater turtle species and their phylogenetic relationships." Genetic Molecular Resource 10(3): 1545-1557. Jombart, T. (2013). A tutorial for Discriminant Analysis of Principal Components (DAPC) using adegenet 1.4-0. From http://adegenet.r-forge.r-project.org/files/tutorial-dapc.pdf Knapp, C., M. Breuil, C. Rodrigues and J. Iverson (2014). Lesser Antillean Iguana, Iguana delicatissima: Conservation Action Plan, 2014—2016, by: IUCN, Gland, Switzerland. Kulikova, I. V., Y. N. Zhuravlev, K. G. McCracken and D. Haukos (2004). "Asymmetric hybridization and sex-biased gene flow between Eastern Spot-billed Ducks (Anas zonorhyncha) and Mallards (A. platyrhynchos) 27.

(30) in the Russian Far East." The Auk 121(3): 930-949. Librado, P. and J. Rozas (2009). "DnaSP v5: a software for comprehensive analysis of DNA polymorphism data." Bioinformatics 25(11): 1451-1452. Lin, S.-M., Y. Lee, T.-H. Chen and J.-W. Lin (2015). "Habitat Preference and Management of a Chinese Pond Turtle Population Protected by the Demilitarized Kinmen Islands." Journal of Herpetology. (In-Press) Lockwood, J. L., M. F. Hoopes and M. P. Marchetti (2013). Invasion ecology, Blackwell, Oxford, UK. Liu L., L.-W. Nie, X.-J. Bu, X.-Q. Xia, Z.-F. Huang and W.-X. Jing (2012). "Isolation and characterization of ten novel polymorphic microsatellites loci in the Chinese pond turtle (Mauremys reevesii) and cross-species amplification in other Cryptodira species." Anhui Normal University Master Thesis, Anhui, China. Mank, J. E., J. E. Carlson and M. C. Brittingham (2004). "A century of hybridization: decreasing genetic distance between American black ducks and mallards." Conservation Genetics 5(3): 395-403. Otani, T. (1995). "A possible hybrid between Geoemyda japonica and Cuora flavomarginata obtained in captivity." Akamata 11: 22-24. Parham, J. F., T. J. Papenfuss, P. P. Dijk, B. S. Wilson, C. Marte, L. R. Schettino and W. Brian Simison (2013). "Genetic introgression and hybridization in Antillean freshwater turtles (Trachemys) revealed by coalescent analyses of mitochondrial and cloned nuclear markers." Molecular Phylogenetics and Evolution 67(1): 176-187. Poulakakis, N., S. Glaberman, M. Russello, L. B. Beheregaray, C. Ciofi, J. 28.

(31) R. Powell and A. Caccone (2008). "Historical DNA analysis reveals living descendants of an extinct species of Galapagos tortoise." Proceedings of the National Academy of Sciences of the United States of America 105(40): 15464-15469. Rhodin, A., J. Parham, P. Van Dijk and J. Iverson (2009). "Turtles of the world: annotated checklist of taxonomy and synonymy, 2009 update, with conservation status summary." Conservation biology of freshwater turtles and tortoises: A compilation project of the IUCN/SSC Tortoise and Freshwater Turtle Specialist Group. Chelonian Research Monographs (5): 000.39-000.84. Rhymer, J. M. and D. Simberloff (1996). "Extinction by hybridization and introgression." Annual Review of Ecology and Systematics: 83-109. Shi, H., J. F. Parham, W. B. Simison, J. Wang, S. Gong and B. Fu (2005). "A report on the hybridization between two species of threatened Asian box turtles (Testudines: Cuora) in the wild on Hainan Island (China) with comments on the origin of'serrata'-like turtles." Amphibia-Reptilia 26(3): 377. Spinks, P. Q., H. B. Shaffer, J. B. Iverson and W. P. McCord (2004). "Phylogenetic hypotheses for the turtle family Geoemydidae." Molecular Phylogenetics and Evolution 32(1): 164-182. Stuart, B. L. and J. F. Parham (2007). "Recent hybrid origin of three rare Chinese turtles." Conservation Genetics 8(1): 169-175. Suzuki, D., H. Ota, H.-S. Oh and T. Hikida (2011). "Origin of Japanese Populations of Reeves' Pond Turtle, Mauremys reevesii (Reptilia: Geoemydidae), as Inferred by a Molecular Approach." Chelonian 29.

(32) Conservation and Biology 10(2): 237-249. Tamura, K., G. Stecher, D. Peterson, A. Filipski and S. Kumar (2013). "MEGA6: Molecular Evolutionary Genetics Analysis version 6.0." Mol Biol Evol 30(12): 2725-2729. Van Oosterhout, C., W. F. Hutchinson, D. P. Wills and P. Shipley (2004). "MICRO‐CHECKER: software for identifying and correcting genotyping errors in microsatellite data." Molecular Ecology Notes 4(3): 535-538. Wilson, G. A. and B. Rannala (2003). "Bayesian inference of recent migration rates using multilocus genotypes." Genetics 163(3): 1177-1191. Wink, M., D. Guicking and U. Fritz (2000). "Molecular evidence for hybrid origin of Mauremys iversoni (Pritchard et McCord, 1991) and Mauremys pritchardi (McCord, 1997) (Reptilia: Testudines: Bataguridae)." Zoologische Abhandlungen-Staatliches Museum Fur Tierkunde in Dresden 51(1): 41-49. Xia, X., L. Wang, L. Nie, Z. Huang, Y. Jiang, W. Jing and L. Liu (2013). "Interspecific hybridization between Mauremys reevesii and Mauremys sinensis: Evidence from morphology and DNA sequence data." African Journal of Biotechnology 10(35): 6716-6724. Ye, R., R. Zheng, L. Wang and W. Du (2008). "Polymorphic microsatellite loci in the Chinese pond turtle (Chinemys reevesii)." Conservation Genetics 10(4): 1045-1048. You, C.-W., Y.-P. Lin, Y.-H. Lai, Y.-L. Chen, Y. Tang, S.-P. Chou, H.-Y. Chang, R. T. Zappalorti and S.-M. Lin (2013). "Return of the pythons: 30.

(33) first formal records, with a special note on recovery of the Burmese python in the demilitarized Kinmen islands." Zoological Studies 52(1): 8 Zhang, Y., J. Y. Zhang and R. Q. Zheng (2010). "Isolation and characterization of 14 polymorphic microsatellite loci in the Asian yellow pond turtle, Mauremys mutica (Cantor, 1842)." Aquaculture Research 41(9): e353-e356.. 31.

(34) Table 1: List of ponds sampled in Kinmen islands number. name. Turtle sampled (M.reevesii/ M.sinensis/intermediate). Migration rate (M. sienesis/M. mutica). 1 2 3 4. Shan-Xi Sha-Mei Guang-Qian-Xi Kai-Hu. 2/0/0 1/0/0 9/1/0 0/0/1. None None 0.0111/0.0112 None. 5 6 7 8 9. Gui-Wu Long-Ling Yang-Ming-Hu Xi-Pu Xi-Cun. 52/2/4 18/1/0 8/0/1 6/0/0 2/0/0. 0.0204/0.0055 0.0082/0.0083 0.0346/0.0112 0.0127/0.0129 None. 10 11 12 13. Liao-Luo Tai-Hu Ming-Tan Long-Tang. 12/0/0 9/2/0 11/1/10 3/0/0. 0.0099/0.0099 0.0113/0.0113 0.088/0.0085 None. 14 15 16 17 18 19 20 21 22. Qiong-An Guan-Lu-Bian Tong-Sha Ou-Cuo Gu-Gang Shui-Tao Ling-Shui-Hu Shuang-Li-Shidi Shuang-Li-Hu. 21/15/3 37/1/0 4/0/0 20/0/0 1/0/0 7/0/0 3/0/0 5/0/0 11/4/0. 0.0157/0.0097 0.0057/0.0056 0.0138/0.014 0.008/0.0081 None 0.0167/0.0126 None 0.0139/0.024 0.0112/0.0103. Note: Migration rate is performed by Bayesass with ponds sampled over 5 individuals.. 32.

(35) Table 2: Locus name, repeat number, primer sequences, range of allele size, annealing temperature, Accession, and reference for the 13 microsatellite loci used in this study. Locus. Repeat motif. Primer sequences. M01. (GT)10. M02. (CA)13. M03. (TG)11. M04. (TG)11. M05. (TG)15(AG)8. M07. (TG)9(AG)18. M08. (TG)13. Cree22. (CA)8. Cree37. (CT)7G(AC)9. Cree46. (AC)11. Cree107. (TG)19. Mum1. (CA)12. Mum13. (GT)20. F: TCCTATACCCAGTGGGACATG R: AAGTTTCACCCATCCATCAGC F: ACTGTGCCGGGTCGTGATGGG R: GTGGTGTCTGAGTGTTCTTGC F: GTGGTGGCACAGAGGTAGTTG R: CTCACATTTTCAGTTTGGTTA F: GAAAGGCTGCTGCTCACCACG R: ACTGACCCAACCCTCCCTGCT F: AGCAAGGCCAGGTAAAGG R: CATCTGCGCTGAGGGTTA F: GAAGAGCCGACCGTCAACCT R: CCCGTAGCCTTTCAAACCAC F: AGCTCCTCCAGGAACTAAAA R: AAACCAAAGTCTTTCCAACC F: GTCCGTGGGTCACATACT R: AGAGACGCCATTCCTTTA F: GCTGGTTGTGTCTCACTTGA R: CCCTGCCTTTGCTTATTC F: ACATACAACTTACACAAGC R: GACAAAATGCAGACTACA F: AGCAACAGCAAAATGAAG R: AGAGGGATAAGGCAAAGA F: CCCCACTTTTTACCAGCC R: CCAAATGTTGCCACAATCTA F: ACTGAAGGGAGGATTGG R: TACTGACTGGTATTATTTCTATTG 33. Range of allele size 165-185. Annealing temperature 53.9. Accession No. JF712880. Liu et al. 2012. 255-265. 58.0. JF712881. Liu et al. 2012. 216-290. 57.5. JF712882. Liu et al. 2012. 194-228. 57.5. JF712883. Liu et al. 2012. 302-346. 60.0. JF712884. Liu et al. 2012. 214-298. 57.5. JF712886. Liu et al. 2012. 239-303. 57.5. JF712887. Liu et al. 2012. 388–424. 52.5. EU825727. Ye et al. 2008. 414–448. 52.0. EU825728. Ye et al. 2008. 296–314. 55.0. EU825730. Ye et al. 2008. 377–393. 50.0. EU825732. Ye et al. 2008. 261–287. 59.0. GU062888. Zhang et al. 2010. 203–243. 59.0. GU062900. Zhang et al. 2010. Reference.

(36) Table 3: The habitat factors which influent the genetic introgression degree of M. reevesii Abbreviation. Description. Data type. p value. Selected for logistic regression. Coefficient. DM DS. Shortest distance to main road Shortest distance to secondary road. Continuous Continuous. 0.1399 0.1595. No No. No No. MS VL VW PS PA PD. The amount of M. sinensis in each pond Vegetation coverage on land1 Vegetation coverage in water2 Substrate type around of the pond3 Area of the ponds4 Depth of the ponds5. Continuous Ordinal Ordinal Ordinal Ordinal Ordinal. 0.0002 0.2282 0.9101 0.0023 0.8582 0.8682. Yes No No Yes No No. 0.0048 No No -0.019 No No. Note: Selection of variable is performed by stepwise with enter probability 0.1 and leave probability 0.1. 1 Defined by the height of vegetation within 10 meters around the shore. Low: vegetation height under 50 cm; high: vegetation height above 50 cm. 2 Defined by the ratio of vegetation coverage with respect to the pond area. Low: <10%; medium: 10%–50%; High: >50%. 3 Defined by the level of artificial modification: natural substrate, artificial substrate, and partially modified substrate. 4 Small: <1000 m2; medium: 1500–10,000 m2; large: >10,000 m2. 5. Shallow: <0.6 m; medium: 0.6–1.5 m; deep: >1.5 m.. 34.

(37) Table 4: Inconsistency between morphology and mitochondrial cytochrome b sequences Cytochrome b Morphology M. reevesii M. sinensis M. mutica Intermediate form. M. reevesii. M. sinensis. M. mutica. U/N. 231 1 0. 6 27 0. 4 0 0. 0 0 0. 6. 13. 0. 0. Table 5: Number of hybrids identified by microsatellite markers. Microsatellite Morphology M. reevesii M. sinensis M. mutica Intermediate form. M. reevesii. Hybrids. M. sinensis. 213. 28. 0. 0 0 1. 1 0 15. 27 0 3. 35.

(38) D. A. E B. F. C G. Figure 1. Turtles in Kinmen with intermediate form Some of the turtles sampled in Kinmen Islands displayed the morphology with intermediate form between several Mauremys species. A, B, and C represent Mauremys reevesii, M. sinensis, and M. mutica, respectively. D, E, F and G are turtles with intermediate forms. 36.

(39) Figure 2. Sample localities of Turtles in Kinmen Turtles sampled in Kinmen Islands with a total of 43 water regions. The closed circles represent ponds with turtle records, while open circles are ponds without turtles. Numbers beside the ponds are the serial numbers of the ponds (Table 1).. 37.

(40) Figure 2. Maximum likelihood (ML) tree Phylogeny of turtles in Kinmen based on cytochrome b sequences. Asterisks (*) indicate the inconsistency between morphology and mtDNA grouping.. 38.

(41) Mauremys reevesii from China. Mauremys sinensis from Taiwan. Mauremys mutica from Taiwan. Figure 4. Genetic clustering of the three pure species using STRUCTURE The cluster analysis (K=3) conducted by STRUCTURE 2.3.3 of Mauremys reevesii, M. sinensis and M. mutica.. 39.

(42) Figure 5. The STRUCTURE 2.3.3 result of Kinmen's turtles. The result of all turtles used in this study with two clusters (K=2, Fig. 5a) and three clusters (K=3, Fig. 5b).. 40.

(43) Figure 6. The DAPC analysis of Kinmesn’s turtles. 7 clusters represent 3 pure Mauremys turtle (abridge as MR, MS and MM), hybrid F1 turtles from Taiwan’s turtle farm(TW), M. reevesii from Kinmen (KMR), M. sinensis from Kinmen (KMS), and intermediated form turtles (HY). The intermediate forms turtle (HY) and some of the M. reevesii from Kinmen (KMS) are positioned in the middle of M. reevesii and M. sinensis pure species (MR and MS), suggesting that genetic introgression mainly occurred from M. sinensis. Genetic introgression from M. mutica is less obvious in Kinmen’s population. 41.

(44) Figure 7. Wilcoxon signed rank test of the residuals of genetic introgression degree against to the amount invasion Mauremys turtles between artificial and natural side type. The genetic introgression degree significantly higher for the artificial habitat than natural habitat (Wilcoxon signed rank test, two-tailed, Z = 2.08; p < 0.05). 42.

(45) Figure 8. Genetic introgression to M. reevesii in different ponds in Kinmen Different ratio of genetic introgression in each pond. The red part represent number of M. reevesii sampled in each pond, green part represent numbers of M. sinensis while black part represent the numbers of hybrid turtles. The size of circles represent amount of turtles within each pond.. 43.

(46) Appendix 1. Mitochondrial DNA cytochrome b haplotype data of turtles from Kinmen, Taiwan and China Morphology Haplotype. M. reevesii. M. sinensis. M. mutica. Intermediate form. Hap_1. 199. 1. 0. 5. Hap_2. 2. 0. 0. 0. Hap_3. 1. 0. 0. 0. Hap_10. 8. 0. 0. 0. Hap_11. 1. 0. 0. 0. Hap_12. 1. 0. 0. 0. Hap_15. 3. 0. 0. 0. Hap_16. 1. 0. 0. 0. Hap_5. 1. 1. 0. 7. Hap_7. 1. 5. 0. 0. Hap_9. 3. 17. 0. 1. Hap_13. 1. 6. 0. 1. Hap_14. 1. 0. 0. 0. Hap_17. 0. 2. 0. 0. Hap_23. 0. 1. 0. 0. Hap_24. 0. 1. 0. 0. M. mutica Hap_4 Hap_6 Hap_8 Hap_18. 2 1 1 0. 0 0 0 0. 1 0 2 1. 0 0 0 0. Hap_19 Hap_20 Hap_21 Hap_22. 0 0 0 0. 0 0 0 0. 1 1 1 1. 0 0 0 0. M. reevesii. M.sinensis. 44.

(47) Appendix 2. The STRUCTURE cluster (K) calculated by STRUCTURE HARVESTER a K. Mean LnP(K). Stdev LnP(K). Ln’(K). |Ln''(K)|. Delta K. 1. -133.55. 1.202082. --------. --------. --------. 2. -1019.25. 1.767767. 114.3. 6.3. 3.563818. 3 4 5. -898.65. 0.212132. 120.6. 163.3. 769.803582. -941.35. 16.051324. -42.7. 16.6. 1.034183. -1000.65. 103.591143. -59.3. --------. --------. K. Mean LnP(K). Stdev LnP(K). Ln’(K). |Ln''(K)|. Delta K. 1. -14081.9. 0. --------. --------. --------. 2. -12592.35. 0.212132. 1489.55. 1229.7. 5796.861. 3 4. -12332.5. 2.545584. 259.85. 89.2. 35.041. -12161.85. 17.324116. 170.65. --------. --------. b. 45.

(48)