Preservation of Genetic Diversity of the Asian Native Goats

Using single nucleotide polymorphisms (SNPs), the relative importance of 10 subpopulations of Asian native goats in preserving genetic diversity was investigated. Analysis of prioritizing by removal of subpopulations identified the subpopulations of Mongolia (MGL), Myanmar (MYA), Cambodian plains (CAM_P), India (IND), and Philippine (PHI) as genetically important subpopulations because their removal resulted in a 6.38% reduction of expected heterozygosity. The removal of the remaining five subpopulations resulted in a 1.45% increase. Likewise, analysis using the core set method identified five subpopulations (MGL, MYA, CAM_P, IND, and PHI) as genetically important subpopulations. Among these five subpopulations, the IND was most important because of low molecular coancestry within itself and between the other subpopulations. The subpopulations of Cambodian mountainous (CAM_M) and Vietnam (VIE) were also considered to be important in this analysis. Based on these two investigations, we concluded that MGL, MYA, CAM_P, IND, and PHI are essential, and that CAM_M and VIE are worth preserving.


Introduction
Goats are raised all over the world, and more than 65% of these goats are in developing countries. The production of meat, milk, and hide is very important and supports the lives of people in those countries. Although Asia has rich native goat resources, these animals have been crossed with imported profitable breeds, and face the risk of extinction [1]. Therefore, proposing a specific preservation scheme for native goat genetic resources in Asia is an urgent task.
Recently, the origin, phylogeny, and genetic diversity of Asian native goats were studied using various molecular genetic techniques ( [2], [3], [4]). However, genetically important subpopulations for preserving the genetic variability of Asian native goats have not yet been identified. Prioritizing subpopulations is an essential task because both human and economic resources for the preservation schemes are usually limited, and should be conducted before the establishment of the preservation scheme. In this study, we identified genetically important goat subpopulations among 10 subpopulations from Asia

Animals
Thirty animals were sampled from each subpopulation except for IND, from which 20 animals were sampled. Genomic DNA was extracted from blood according to the standard phenol and chloroform method.

SNP Detection
For detailed information about the process of detection of SNP markers, see Lin et al. [3].

Population Genetic Analysis
Caballero & Toro [5] proposed a method for genetically prioritizing the subpopulations, in which the change (decrease or increase) of expected heterozygosity [6] is examined when one or more subpopulations are assumed to be removed from the whole population. Assuming an equal subpopulation size and letting f ij be the molecular coancestry between subpopulations i and j, average molecular coancestry of the whole population ( ̅ ) can be decomposed into where is the average of within-subpopulation coancestry (f ii ) and D is the average of Nei's minimum distance ( [7]: D ij = (f ii +f jj )/2-f ij ) among all the subpopulations. Equation (1) indicates that the average molecular coancestry of the whole population ( ̅ ) depends on the within-subpopulation component ( ) and averaged distance among the subpopulations (D ). Thus, the total genetic diversity of the whole population (GD T ) expressed as expected heterozygosity is also decomposed as implying that the total genetic diversity of the whole population is partially ascribed to within-diversity components of each subpopulation (GD WS ) and to between-diversity components among all the pairs of subpopulations (GD BS ). In a second investigation for prioritizing the 10 subpopulations, optimal genetic contributions of subpopulations for maximizing the genetic diversity of a hypothetical gamete pool were calculated. This calculation was conducted using the core set method of Eding et al. [8], which was originally proposed for ranking animal breeds regarding conservation of genetic diversity within a livestock species. The similar technique is used in many situations nowadays (e.g. [9]). Suppose g i gametes are sampled from subpopulation i to obtain a hypothetical gamete pool with g T total gametes, so that the genetic contribution of subpopulation i is c i =g i /g T and ∑ = 1. (2) Letting f ij be the molecular coancestry between subpopulations i and j, and F = [f ij ] be the molecular coancestry matrix of all the subpopulations, the expected molecular coancestry of the hypothetical gamete pool is where c is a column vector of c i . Thus, maximization of genetic diversity of the hypothetical gamete pool (1 − ) is equivalent to the problem of finding optimal c (c opt ) such that is minimized under the restriction of Eq. (2). This problem can be solved by applying the Lagrange multiplier method, and the optimal contribution of each subpopulation is obtained as where 1 is a unit vector (for detailed derivation, see Eding et al. [8]).
All the computations were conducted using our original Fortran programs.

Results
Genetic diversity and its proportional changes according to the removal of subpopulations are provided in Table 1. The genetic diversity of the whole population (GD T ) was 0.388, and it was decomposed into 0.331 for the within-diversity component (GD WS ) and 0.057 for the between-diversity component (GD BS ). Removal of MGL, MYA, CAM_P, IND, and PHI was accompanied by the reduction of GD T , indicating that these five subpopulations are genetically more important than the remaining five subpopulations for preserving genetic diversity. The reduction of GD T caused by the simultaneous removal of the previous five subpopulations (MGL, MYA, CAM_P, IND, and PHI) (6.38%) was substantially larger than the sum of individual effects (3.10%). In addition, removal of the remaining five subpopulations (LAO, CAM_M, VIE, BHU, and BAN) increased GD T (1.45%), indicating that all five subpopulations (MGL, MYA, CAM_P, IND, and PHI) are necessary for preserving genetic diversity of Asian native goats. Table 2 lists the average molecular coancestries within and among the 10 subpopulations. Molecular coancestries within each subpopulation were higher than those between the other subpopulations. Specifically, molecular coancestries within CAM_M and PHI were extremely high, indicating that these two subpopulations were affected by the larger amount of genetic drift. MGL and IND have relatively low coancestries between the other subpopulations. Optimal genetic contributions of each subpopulation obtained by the core set method are presented in Table 3. The maximum possible genetic diversity was estimated to be 0.398. The subpopulations with positive contributions were the five subpopulations that were found to be genetically important in the previous analysis (MGL, MYA, CAM_P, IND, and PHI), as well as CAM_M and VIE.

Discussion
Recalling that expected heterozygosity on the biallelic loci reduces at the rate of 1-1/2N e [10], where N e is the effective population size, and that standard effective population sizes of livestock breeds are around 100 [11], even a 0.5% change of heterozygosity resulting from removal of the subpopulation is not negligible. Among the five subpopulations that were found to be genetically important in the first analysis (MGL, MYA, CAM_P, IND, and PHI), the effect of MGL and IND were relatively large, and reductions were observed in both components of GD WS and GD BS . These two subpopulations are geographically separated from the other subpopulations of Southeast Asia and are the oldest subpopulations of domestic goats introduced by nomadic tribes on the Silk Road [12].
Removal of CAM_M substantially reduced GD BS ; however, it greatly increased GD WS and inflated GD T as a consequence (i.e., frequencies of two alleles in each locus became equalized). This is because CAM_M has relatively large genetic distance from the other subpopulations, but also has quite small within-variability (cf. [3]). Steep mountain paths might limit the introgression of new genetic variants. A large reduction of GD BS was also observed when PHI was removed. Geographical effect would be responsible for this reduction, i.e., PHI is an isolated island subpopulation.
In the core set analysis, the contribution of IND was fairly large, indicating that this subpopulation is predominantly important. Molecular coancestry of this subpopulation was the second lowest, and coancestries between the other subpopulations were relatively low ( Table 2). India has rich goat resources; consumption of beef and pork is prohibited by Hinduism and Muslimism so goats are an important protein source [12]. The neighbor-joining tree constructed by Lin et al. [3] formed two clusters (cluster A MGL, BHU, BAN, and IND; cluster B LAO and VIE), and the remaining four subpopulations were independently distributed. MGL and IND were selected as non-zero contributors from cluster A, and VIE was selected from cluster B. This might be why BHU, BAN, and LAO could not contribute to the hypothetical gamete pool (i.e., why these subpopulations were not recognized as genetically important subpopulations).

Conclusion
Based on the analysis in this study, we concluded that the subpopulations MGL, MYA, CAM_P, IND, and PHI are primarily important for preserving the genetic diversity of Asian native goats, and that if the preservation resources remain, CAM_M and VIE are worth preserving. Compared to previous studies that uncovered origin, phylogeny, and genetic diversity of many livestock species in Asia using various molecular genetic techniques, this rare study provides a specific and systematic preservation scheme. In addition to neutral molecular variation, adaptive traits possessed by each subpopulation should also be taken into account in future study, as suggested by Wellmann et al. [13], and deriving the economic values of each subpopulation and designing breeding schemes is essential [14].