Large‐scale genetic surveys for main extant population of wild giant panda (Ailuropoda melanoleuca) reveals an urgent need of human management

Abstract There are only six isolated living giant panda populations, and a comprehensive understanding of their genetic health status is crucial for the conservation of this vulnerable species. Liangshan Mountains is one of the main distribution areas of living giant pandas and is outside the newly established Giant panda national park. In this study, 971 giant panda fecal samples were collected in the heartland of Liangshan Mountains (Mabian Dafengding Nature Reserve: MB; Meigu Dafengding Nature Reserve: MG; and Heizhugou Nature Reserve: HZG). Microsatellite markers and mitochondrial D‐loop sequences were used to estimate population size and genetic diversity. We identified 92 individuals (MB: 27, MG: 22, HZG: 43) from the three reserves. Our results showed that: (1) genetic diversity of three giant panda populations was moderate; (2) several loci deviated significantly from the Hardy–Weinberg equilibrium and almost all these deviated loci showed significant heterozygote deficiencies and inbreeding; (3) three giant panda populations have substantial genetic differentiation with the most differentiation between MB and the two other populations; and (4) a large amount of giant panda feces outside the three reserves were found, implying the existence of protection gap. These results indicated that under stochastic events, the giant panda populations in Liangshan Mountains are at risk of genetic decline or extinction and urgent need of human management. This study revealed that high attention should be paid to the protection of these giant panda populations outside the Giant panda national park, to ensure their survival in their distribution areas.

. The wild giant pandas are subject to different degrees of habitat fragmentation at each of the six mountains, and is further divided into more than 30 small populations (Loucks et al., 2001;Lu et al., 2001;O'Brien et al., 1994;Qing, 2016). Therefore, the giant panda is still at a great risk of extinction (Sichuan Provincial Forestry Department, 2015), particularly being vulnerable to stochastic processes. And thus a comprehensive understanding the population size and genetic health status of giant pandas in these regions is crucial for the protection decision-making and conservation of this vulnerable species.
The Liangshan Mountains is the southernmost distribution of giant pandas and is located in the transition zone between the southwest edge of the Sichuan basin and the Qinghai Tibet Plateau.
The transition zone is within a global biodiversity hotspot, is highly important for the protection of biodiversity in China and is crucial for giant panda protection (Fan et al., 2010). However, the Liangshan  (Table 1). These three reserves are located in the heartland of the Liangshan Mountains, and thus are crucial for the protection of giant pandas in the Liangshan Mountains. However, the accurate number, genetic diversity, gene exchange and stable inheritance of panda populations in these key areas remain unclear.
Understanding these issues will be vital to the protection of giant pandas in the Liangshan Mountains.
Microsatellite markers have become an important genetic markers in the field of molecular biology (Selkoe & Toonen, 2010) and have been widely used in population surveys (Creel et al., 2003;Piggott et al., 2006;Wang et al., 2016), genetic diversity assessments (Du et al., 2016;Li et al., 2010;Shen et al., 2010;Vanhala et al., 1998;Zhang et al., 2007), and genetic management of populations (Shan et al., 2014). The combined application of microsatellite markers, mitochondrial markers and noninvasive genetic sampling has contributed greatly to giant pandas conservation in the past 20 years, allowing giant panda population studies without the risk of capture stress, injury or death. Consequently, we used microsatellite markers and mitochondrial markers (D-loop) to accurately identify population size and assess the genetic traits of giant pandas in Heizhugou, Meigu and Mabian giant panda populations. We aimed to assess the genetic health status of giant pandas and provide reliable data for establishing genetic archives of giant panda populations and developing the genetic management of giant pandas across the Liangshan Mountains. This is the first extensive genetic survey of giant pandas in the Liangshan Mountains.

| Study area and sample collection
Our study area encompassed Heizhugou, Meigu Dafengding and Mabian Nature Reserves of the Liangshan Mountains (Table 1).
The relative position of the study area in China.
Giant pandas fecal samples were collected by ranger staff during their daily monitoring and patrol work in the reserves. The samples were collected from Heizhugou in October 2016 and May 2017, Meigu in October 2017 and May 2018, and Mabian in April and October 2018, respectively. The staff used sterile gloves to collect fresh fecal samples when they detected giant panda activity. Samples were considered fresh based on the color and surface sheen, with dark colored and dull feces being discarded. Each sample was collected in 1-2 copies and stored in a 500 ml sample bottle containing anhydrous ethanol.
Spatial coordinates were recorded from the deposition site (e.g., longitude, latitude, elevation) using GPS units and the distribution of samples was mapped as shown in Figure 2, using ArcGIS 10.6 (Price, 2015).

| DNA extraction and PCR amplification of mitochondrial D-loop
Fecal DNA was extracted using the kit (Biobase Upure DNA stool kit) and nucleic acid purifier (Thermo KingFisher). Fecal samples collected in the field were soaked in anhydrous ethanol and frozen at −20°C. DNA extraction was undertaken according to the manufacturer's instructions. The mitochondrial control region was amplified by PCR for those samples where DNA was successfully extracted, using primers from Zhang et al. (2007).
Amplifications were performed using the following PCR procedure: an initial denaturation step for 5 min at 94°C, followed by 40 cycles of 94°C for 50 s, 55°C annealing for 45 s, 72°C elongation for 50 s and a final elongation for 10 min at 72°C. Finally, samples were stored at 4°C.

| Selection and amplification of microsatellite markers
Our laboratory has screened giant panda DNA for standardized microsatellite loci and obtained 15 loci that can be effectively applied TA B L E 1 Areas and number of giant pandas of each reserve in Liangshan Mountains (Sichuan Provincial Forestry Department, 2015).
Population genetic structure analysis was undertaken using STRUCTURE (Pritchard et al., 2000). The range of possible clusters (K) tested was from 1 to 6, and 10 independent runs were carried out for each. The lengths of Markov Chain Monte Carlo (MCMC) iterations and burn-in were set at 1,000,000 and 100,000, respectively. The true K is selected using the maximal value of the log likelihood [Ln Pr(X/K)] of the posterior probability of the data for a given K (Pritchard et al., 2000). The Fst of giant panda population pair-wise comparisons from the three reserves was calculated and measured by GenALEx 6.5 (Peakall & Smouse, 2012). The gene flow (Nem) among populations was calculated using Nem = (1-Fst)/4Fst (Wright, 1990), where Nem is the effective number of migrations per generation among populations.  Figure S1 for partial electrophoresis).

| RE SULTS
When collecting fecal samples in the field, the freshness of samples was estimated based on intactness, color, odor, and the status of the mucosal outer-layer. We estimated most collected fecal samples to be less than 2 weeks old. In addition, even if the fecal freshness is the same, the integrity of the fecal DNA will be different.
Mitochondrial DNA is multiple copies and is easier to be amplified compared with nuclear genomic DNA. In this study, the quality of fecal DNA samples was preliminary evaluated by the method of whether the mitochondrial control region of giant panda fecal DNA was successfully amplified or not. A total of 686 DNA samples were successfully amplified for the mitochondrial control region (MB: 218, MG: 228, HZG: 240; Figure S2 for partial electrophoresis) were used to amplify and genotype for seven microsatellite loci.

| Genetic diversity based on microsatellite markers
Seven microsatellite markers were successfully amplified from the

| Genetic diversity based on mitochondrial control region sequence
We successfully sequenced the mitochondrial D-loops from 85 of the 92 individuals from the three reserves, with sequencing peaks shown in Figure S3. The number of mitochondrial D-loop sequences (n), haplotype (H), variation sites (s), haplotype diversity (h), and nucleotide diversity (π) of the three populations are summarized alongside other wild and captive populations in Table 5. Compared with other populations, the mitochondrial genetic diversity of giant pandas in these three reserves was significantly lower than in wild giant panda populations from Qinling, Minshan and Qionglai Mountains (Yang, 2013). Mitochondrial genetic diversity of three populations was also lower than captive populations from Wolong, Chengdu and Shaanxi, but higher than Daxiangling and Xiaoxiangling populations (Yang, 2013).

| Geographic isolation and genetic differentiation
According to the distribution map of giant panda fecal samples The software STRUCTURE (Pritchard et al., 2000) was used to analyze the population genetic structure. Our results showed that when K = 2, the value of K peaked and decreased with increasing values of K. As shown in Figure 4, the giant pandas of three reserves were clearly divided into two genetic structural units, Heizhugou Heizhuguo and Meigu population. The Nem of the three wild giant panda populations was relatively low, ranging from 1.32431 to 3.05907 (Table 6). Among them, the Fst between Mabian and Meigu was the largest, resulting from the smallest number of effective migrants exchanged per generation (Nem = 1.32431) ( Table 6).

| Population census
Accurate population census is especially complex and important for giant pandas that they are not readily visible and easily counted in their habitat environments. Population census has served as a basis for judging not only the conservation status of pandas but also the effectiveness of measures designed to protect them and their habitat (Wei et al., 2012). Traditionally, an approach using bamboo bite length has been applied. However, the precision of this approach was always known to be low.
In recent years, the development of molecular biology techniques has provided opportunities for more accurate population census. These techniques are mainly carried out at the DNA level, but the biggest obstacle of DNA analysis for the wild panda is the collection of samples . As opposed to both destructive sampling and nondestructive sampling, the researchers realized that

| Genetic health assessment of populations
Although genetic diversity is merely one of a number of important considerations in species conservation, the protection of species genetic diversity has always been the core of species protection (Frankham, 2005). The evaluation of the genetic diversity within the protected species can allow conservation practitioners to predict the probability of population extinction or survival when under stress and to provide an theoretical basis for the effective conservation of population. As an extranuclear genetic material, mitochondrial DNA is maternal origin, rapidly evolving, and polymorphic, making it a common molecular marker for species evolution and genetic diversity studies. However, due to maternal inheritance, the lack of recombination, different rate of evolution compared with nuclear DNA, and more sensitive to founder effects and small populations, the results of mitochondrial DNA analysis might be partial and inconsistent with that of nuclear DNA (Barton & Hewitt, 1985;Qin et al., 2017). In this study, to improve the reliability and authenticity of the conclusion, both SSR molecular markers and mitochondrial DNA markers were used to assess the genetic diversity of giant panda population.
The study found that the genetic diversity of Mabian Serious genetic imbalance may lead to the loss of genetic diversity and population decline (Kvist et al., 2015). The Hardy-Weinberg equilibrium is often used as an assessment of genetic balance within F I G U R E 4 STRUCTURE analysis results of giant panda populations in Heizhugou, Meigu and Mabian Nature Reserves. Delta K = mean(|L(K)|)/sd(L(K)), the corresponding K value at the peak of Delta K is the optimal K value. STRUCTURE output of two genetic clusters identified (K = 2), represented by the colors red and green. Each individual is represented by a vertical line, and different color length indicate the probabilities of being assigned to different clusters. (1. Heizhugou; 2. Meigu; 3. Mabian; a and b: two individuals in common).
TA B L E 6 F-statistics (F st , below the diagonal) and Nem (above the diagonal) of three giant panda populations. a population (Guo & Thompson, 1992  The genetic and geographical clustering of the three populations suggests that there is a barrier preventing genetic exchange between the two areas. Feng (2015) found that suitable habitats were fragmented in central and northern Mabian Nature Reserve. Unsuitable habitats might be caused by deforestation, road construction and livestock invasion (Feng, 2015;Zhang et al., 2018;Zhao et al., 2017).

| Genetic management recommendations
Liangshan However, captive-bred introductions are difficult and require considerable resources and time (Yang et al., 2018), and therefore it should not be the only strategy for the improvement in genetic health.
Our second recommendation for improving the genetic status of the three populations is to increase connectivity and genetic exchange between the two geographically and genetically distinct panda groups. Although significant genetic differentiation between the two groups exists, no significant difference in behavior and morphology has been found. Similarly, there was no evidence that the Mabian population was subject to different geographical or climatic conditions and thus no unique or local adaptation. Therefore, there should be no genetic, behavioral or morphological impediment to breeding and risk of intraspecific hybridization (Frankham, 2010).
The fecal sample distribution ( Figure 2) and population genetics demonstrated there was limited genetic exchange between Mabian and two other populations. However, there is no topographical barrier between the two groups, and the limiting factor is likely from unsuitable habitat and habitat fragmentation due to disturbance and lack of bamboo vegetation (Feng, 2015;Zhang et al., 2018;Zhao et al., 2017). Consequently, we recommend that suitable habitat and continuity should be rehabilitated and restored. Recent roads should be reforested or add migration channels for pandas and prevented from new construction. Human activities, especially grazing and bamboo shoot collection, should be controlled and minimized.
Existing natural forest (bamboo) should be protected from further damage and the nonbamboo areas should be rehabilitated. As a priority, restoration should focus on creating corridors through the "habitat barrier" to increase panda movement as soon as possible and then expand the area and proportion of suitable habitat. Given that pandas begin moving and they breed, there should be an improvement in the genetic health and population stability of giant panda in Liangshan Mountains.

ACK N OWLED G M ENTS
We are grateful to Prof. Jianghong Ran's comments for our manuscript and the staff of the three reserves for collecting samples. This work was supported by the National Natural Science Foundation of China (32070529) and Chengdu Giant Panda Breeding Research Foundation.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
Haplotype sequences were deposited in the GenBank with the accession number OQ108856-OQ108866.