Genetic structure in natural populations of Dacrydium elatum (Roxb.) Wall. (Podocarpaceae) in the Central Highlands of Vietnam inferred by Microsatellites

: To provide a reference for the conservation and application of breeding parentallines resource of Dacrydium elatum (Roxb.). Genetic diversity and population structure of eighty individuals from four populations (Kon Tum, Gia Lai, Dak Lak and Lam Dong) in Central Highlands were evaluated using eight SSR markers. Based on the SSR data, 21 alleles were detected by eight SSR with high polymorphism. The genetic diversity of levels within the populations were moderately high (Ho = 0.555, He = 0.429). The average number of shannon information index were 0.618 and genetic differentiation among populations was low (Fst=0.097). The AMOVA revealed high genetic variation within individuals (87%) compared among populations (13%). The UPGMA phenogram showed that the results of molecular clustering largely agreed with the pedigree and geographic origin. Three populations (NL, KCR and BDNB) were clustered together and CYS population was separated. The maximum quantity ΔK was observed for K=2 in population structure analysis, indicating that the entire collection could be divided into two main groups of genes. This study can provide a theoretical basis for genetic resource management and varieties identification of D. elatum resources, and provide reference basis for breeding.


Introduction
Dacrydium elatum (Roxb.) Wall. genus Dacrydium (Podocarpaceae) isconifer species in Vietnam, grows in primary closed evergreen seasonal forest tropical mixed (elevations 700-2000m) they have been extensively used for furniture, handicrafts and construction [1][2][3]. In Vietnam it is found in Ha Giang, Tuyen Quang, Quang Ninh, Lai Chau, Ha Tinh, Quang Binh, Thua Thien Hue, Da Nang, Kon Tum, Gia Lai, Dak Lak, Lam Dong and Kien Giang provinces and other countries like China, Laos, Cambodia, Thailand, Malaysia and Indonesia [1][2]. D. elatum has been classified as an endangered species and put strictly on the conservation list as Vulnerable (VU) [1]. Up to now, many populations of D. elatum are endangered by over exploitation and deforestation and we have serious lack of information on ecological characteristics and genetic variation at population and species levels of D. elatum, especially the negative effects of human activities. This situation urged to undertake a study to address these problem and to provide additional information for the Forest Protection Department in order to stress the need for the conservation and sustainable use of the species of local communities.
Conservation and management of a species require information on the ecological and genetic diversity within and among it's populations [4][5][6]. To develop appropriate conservation strategies, microsatellite markers (simple sequence repeats, SSR) are commonly used to assess the levels of genetic diversity and genetic structure, both at population and species level [4,7]. SSR has the advantages of high polymorphism, codominant inheritance, and widespread presence in the entire genome [8]. It isideal marker to probe genetic diversity and population structure and has a broad application scope [9][10]. Many SSR markers were developed and have been widely used to probe genetic diversity within and among conifer populations [11][12][13]. Previous studies investigated the genetic variation and verified the taxonomic status of the Dacrydium species at the molecular level, such as D. pectinatum [14], D. imbricatus [15], D. elatum [16], however the genetic structure of D. elatum has not been explored in Vietnam. Meanwhile, knowing genetic diversity is important for the development of conservation strategies and the sustainable utilization of thisspecies [6,17]. Therefore, it is a dire need of the day to explore the diversity and population structure of D. elatum and set breeding and conservation strategies.
The main goal of the current study was to determine level of genetic diversity and genetic structure of natural populations of D. elatum in the Central Highlands of Vietnam, and to provide guidelines for the conservation, management, and restoration of the species.

Sample collection
A total of 80 plant samples (leaf) were randomly collected from four populations in the Central Highlands of Vietnam (Table 1; Figure 1). The samples were immediately placed in liquid nitrogen, transferred to the molecular laboratory of Vietnam-Russia Tropical Centre and stored at -80qC until DNA extraction.   Table 2). The amplification products were separated using a Sequi-Gen ® GT DNA electrophoresis system 8% polyacrylamide gels in 1 x TAE buffer and then visualized by GelRed 10.000X. The sizes of the PCR products were detected and analyzed using GelAnalyzer software of with a 20 bp DNA ladder (Invitrogen, USA).

Data analysis
The software GENALEX ver.6.5 [20] was used to calculate the characterization and polymorphism level of 8 SSR loci including the number of alleles per locus (Na), number of effective alleles (Ne), Shannon-Weiner index (I), the observed heterozygosities (Ho), the expected heterozygosities (He) and the genetic differentiation for all populations (FST). Tests of deviation from the Hardy-Weinberg equilibrium per locus in each population were performed at a significance level of 0.05, using GENEPOP v4.6 [21]. The analysis of molecular variance (AMOVA) of data was calculated by Arlequin 3.1 [22]. The genetic association amongst populations was determined by Poptree2 using the UPGMA approach [23]. The population structure was explored by STRUCTUREv.2.3.4 [24]. The targeted population was separated into groups by Structure Harvester [25] based on the ∆K by Evanno et al. [26].

Genetic diversity
Eight SSR primer pairs amplified consistently under standard conditions, clear products and were used to assess the population genetic structure. All primers were submitted to amplification cycles with primer annealing at 55 0 C. All of SSR markers were polymorphic in four D. elatum populations. A total of 21 different alleles wererecorded for D. elatum. Genetic diversity of each population wasshowed in Table 3

Population structure
It can be seen from the data in Table 4 that analysis of molecular variance (AMOVA) was performed based on 279.643 permutation. The AMOVA showed that total variation was highly significant (p<0.001) within individuals (87%), and among populations (13%) ( Table 4). Genetic variation within populations was recorded as 0.097 (0.067-0.166) indicating low genetic differentiation. The highest differentiation value (FST = 0.166) was between the two populations of NL/CYS and low (FST = 0.067) between NL/BDNB (Table 5). clearly at population level (Fig. 2). Two different groups were generated; the 1 st group was contains 3 populations (NL, KCR and BDNB) with a bootstrap value of 100%. This main group divided into two subgroups. Sub-clusters 1 included two population (NL and KCR) were clustered together with a bootstrap value of 62% and sub-clusters 2 only have BDNB population were separated. And the second major was composed of CYS population.  We were performed population structure analysis, the highest ∆K value (49.63), have a clear peak for 80 individuals when K=2 to be the optimum number of genetic clusters and indicated that all the studied plants exhibited admixture from two clusters (Fig. 4). One group (red) was predominant in the two populations (NL and BDNB) with strong ancestry values 86.8% and 93%, respectively and the second group (green) was composed of two populations (CYS and KCR) with strong ancestry values 95.4% and 63.6%, respectively ( Table 6; Fig. 1&4). In particular, 4 D. elatum populations in Central Highlands of Vietnam showed a rich a rich mixture of ancestral genes among individuals.

Discussion
In the current study, all the 8 SSR loci were highly polymorphic in the 80 genotypes with a mean value of 2.031 alleles per locus, the mean number of effective alleles (Ne = 1.823). The genetic diversity level was moderately high with HO = 0.555 and He = 0.429. Ho > He were detected of four populations in the present study, suggesting that these populations are predominantly allogamic, which might be a result of outcrossing and reflecting the population structures. Although, it has previously been studied on the genetic diversity of D. elatum.. Tran et al. [16] had compared the effectiveness between ISSR and SSR markers in assessing genetic diversity of natural populations of D. elatum in Tay Nguyen, Vietnam and show that high level of genetic variation of D. elatum (He = 0.301). This result was consistent with the study of Tran et al. [16]. Several reports have shown that high genetic diversity in other conifer using SSR markers [11,12,[27][28][29][30][31][32] However, low of genetic diversity were found in prior studies [4,33,34]. Our investigations confirm the suggestion that the genetic structure of natural populations of D. elatum was strongly affected by population sizes. Which a lot of individuals remaining in the natural forest from 100 individuals in KCR and NL populations to about 500 in BDNB and CYS populations. All of studied populations were found in the secondary forests at more than 1000 m elevations.
Forests have been greatly fragmented by human activities and formed small forest patches. All populations of this species remain in such small patches. Such populations are likely to be the results of inbreeding and an effect of genetic drift in subsequent generations. Population structure and genetic relationships are important for establishing the appropriate scale and subunits for conservation management [35]. It is affected by mutation, gene flow, natural selection and genetic drift [36][37]. The STRUCTURE analysis suggested two different clusters among 80 studied individuals. Similarly, UPGMA analysis based on Nei's chord distance using POPTREE2 identified two clusters. This indicates that geography has effects on genetic structure of D. elatum. Based on our results, effective management strategies for D. elatum should include both in-situ and ex-situ conservations. Ex-situ conservation of D. elatum could be immediately implemented. It will be a protective and managed site to grown this species and prevent potential genetic erosion of wild D. elatum populations. Establishment of seed orchards from all the populations should secure genetic sources of D. elatum.

Conclusions
The present study shows that the current situation of D. elatum maintains moderately high levels of genetic diversity and low levels of genetic differentiation among populations. The Bayesian analysis confirmed two main genetic groups with UPGMA dendrogram. The high genetic variation within individuals. Therefore, conservators should focus on maintaining the genetic variation within populations. These strategies should also be implemented with ex-situ conservation activities to avoid inbreeding in the future.