Genetic characterization of Bhutanese native chickens based on an analysis of Red Junglefowl (Gallus gallus gallus and Gallus gallus spadecieus), domestic Southeast Asian and commercial chicken lines (Gallus gallus domesticus)

The genetic diversity of Bhutanese chickens needs to be understood in order to develop a suitable conservation strategy for these birds in Bhutan. In this, work, we used microsatellite markers to examine the genetic diversity of Bhutanese chickens. Four Bhutanese chicken varieties (Black plumage, Frizzle, Naked neck and Red Junglefowl-like, corresponding to Yuebjha Narp, Phulom, Khuilay and Seim, respectively), two subspecies of Red Junglefowl (Gallus gallus gallus and Gallus gallus spadecieus), two varieties of Thai native chickens (Pradhu Hang Dam and Chee; Gallus gallus domesticus) representing the Southeast Asian domestic chicken, and two commercial lines (Broiler and Single Comb White Leghorn) were genotyped with 18 microsatellites that included 16 loci recommended by the FAO/ISAG for investigations of genetic variability in chickens. All loci were polymorphic, with the number of alleles ranging from six (MCW0111) to 23 (MCW0183). Substantial genetic variation was observed in all populations, with the Bhutanese native chicken Yuebjha Narp (Black plumage chicken) showing the lowest genetic variability. Despite extensive intrapopulation variation, the genetic differentiation among 10 populations was moderate. A neighbor-joining tree revealed the genetic relationships involved while principal component analysis showed that Bhutanese native chickens should be given priority in conservation efforts because of their genetic distinctiveness. Chee chickens are especially valuable as a reservoir of predomestic diversity, as indicated by their greater genetic variation and their position in the phylogenetic tree.


Introduction
Domestic chickens are believed to have originated in Southeast Asia, with Thai native chickens being regarded as the original domesticated chickens (Hillel et al., 2003). The decrease in the genetic diversity of native chicken populations described in recent genetic studies has raised concern because the unique genotypes and traits of native populations are at risk of being lost, with a consequent threat to a well-established food source (Nassiri et al., 2007). This situation suggests that the management of native chicken genetic resources should be given greater priority.
Bhutanese native chickens are of socio-cultural and economic importance to the livelihood of many rural populations. For instance, these birds are slaughtered to please local deities, feed guests, and supplement the diet (with eggs and meat) of women during pregnancy and after birth (Nidup et al., 2005). For these reasons, native chickens continue to thrive despite the introduction of several breeds and strains of exotic chickens by the Bhutanese government. Native chickens constitute about 95% of the chicken population in Bhutan (Nidup and Tshering, 2007). The phenotypic characteristics (Nidup et al., 2008), blood group polymorphism (Yamamoto et al., 2007) and mitochondrial DNA sequences suggest that Bhutanese native chickens are genetically diverse (Nidup et al., 2005). However, they have not been genotyped using microsatellite markers recommended by the Food and Agriculture Organization and International Society for Animal Genetics (FAO/ISAG). The genetic diversity of native chickens needs to be thoroughly assessed in order to ensure sustainable poultry production.
According to Nidup et al. (2005), there are 13 strains of native chickens in Bhutan. However, the FAO Domestic Animal Diversity Information System lists only 10 strains. Currently, based on their socio-economic importance and popularity among farmers, four common strains may be tentatively considered to represent the general Bhutanese chicken population. Seim (Red Junglefow-like) is a commonly reared breed that is believed to be an immediate descendent of Red Junglefowl (Nidup et al., 2008) while Yuebjha Narp (Black plumage) is considered to have medicinal values. Khuilay (Naked neck) chickens are generally found in warmer regions of the country and this is consistent with the suggestion that these birds are heattolerant (Yunis and Cahaner, 1999). On the other hand, Phulom (Frizzle) chickens are specifically reared by some castes in southern Bhutan. Previous studies have examined genetic variations in Black plumage (Granevitze et al., 2007) and Naked neck (Nassiri et al., 2007;Pirany et al., 2007) chicken populations in several countries. Some Bhutanese chickens resemble to their ancestor (Red Junglefowl) while others resemble commercial lines.
In the last decade, the investigation of population variations has involved multi-allele markers. In particular, microsatellites have become a powerful tool for studying population genetics because of their unique characteristics, such as random distribution over the genome, codominant inheritance, high mutational rate and high reproducibility (Weigend and Romanov, 2001;Hillel et al., 2003). In this study, we used microsatellites to investigate the genetic variation among Bhutanese native chickens and to assess their genetic relatedness to Red Junglefowl, Thai native chickens and commercial lines. This information should provide a basis for developing effective conservation programs.
Blood samples (1.5 mL) were drawn from the ulnar vein into a microtube containing 0.5 M EDTA and DNA was isolated as described in Goodwin et al. (2007). The blood samples from Red Junglefowl were collected at the Department of National Parks, Wildlife and Plant Conservation (DNP), and those from Pradhu Hang Dam and Chee at the Research and Development Network Center for Animal Breeding (Native chicken) and the Department of Livestock Development, respectively; samples from com- 604 Dorji et al.

Microsatellite genotyping
Eighteen microsatellite combinations from the FAO/ISAG list and Nassiri et al. (2007) were used; these microsatellites were also used in the AVIANDIV project (Hillel et al., 2003). Microsatellite loci amplification was done by polymerase chain reaction (PCR) with specific primers and annealing temperatures ( Table 2). The electrophoretic bands of the PCR products were scored using a SYNGENE Gel documentation system (Syngene Inc., UK).

Data and statistical analyses
The alleles were analyzed to determine the mean number of alleles per locus and the observed (H O ) and expected (H E ) heterozygosities. The Chi-square test was used to assess Hardy-Weinberg equilibrium (HWE). The fixation coefficient of an individual within a subpopulation (F IS ) and the fixation coefficient of a subpopulation within the general population (F ST ) were estimated using GE-NEPOP v. 4.0.10 (Raymond and Rousset, 1995;Rousset, 2008). A neighbor-joining method (Saitou and Nei, 1987) in the Numerical Taxonomy System (NTSYSpc) v. 2.10 package was used to construct a phylogenetic tree based on Nei's unbiased genetic distance (Nei, 1978). Principal component analysis -PCA (SAS, 1998) based on individual Dice genetic distances was used to visualize the genetic relationships and detect geographical clines that were not apparent from the phylogenetic tree. Numerical data were expressed as the mean ± SD with p < 0.05 indicating significance.

Microsatellite polymorphism and population diversity
The genetic variability of the microsatellite loci is summarized in Table 2. 255 alleles were detected across 18 loci in ten chicken populations, with a mean number of alleles per locus (MNA ± SD) of 14.17 ± 4.37. Locus MCW0183 was highly polymorphic with 23 alleles while MCW0111 had the lowest polymorphism (6 alleles per locus). Averaged over the 18 loci for each population, the H O and H E ranged from 0.262 (MCW0248) to 0.791 (LEI0094) and 0.669 (MCW0111) to 0.898 (LEI0094), respectively. For all loci, the mean H E was higher than the mean H O (Table 2), which suggested sampling bias or a possible inbreeding mating system. Table 3 summarizes the genetic variation across populations. This variation was greatest for Chee (MNA ± SD, Genetic characterization of Bhutanese chickens 605 The Wright fixation indices for F IS ranged from -0.227 (MCW0034) to 0.857 (MCW0248), F ST ranged from 0.039 (MCW0248) to 0.171 (MCW0104), and F IT ranged from -0.151 (MCW0034) to 0.863 (MCW0248), with means of 0.363 ± 0.247, 0.084 ± 0.037 and 0.414 ± 0.233, respectively (Table 2). A high positive F IS indicated a high degree of observed homozygosity (MCW0248) while there was excessive heterozygosity at locus MCW0034, as indicated by the negative F IS value. Significant deviations from HWE (p < 0.5) were observed across 10 populations at locus MCW0295. High selection pressure resulted in seven loci that deviated from HWE in the Broiler strain (ADL112, ADL0268, ADL0372, MCW0037, MCW0069, MCW0111 and MCW0123). Deviation of HWE was also found in Red Junglefowl (MCW0111), Chee (ADL112) and Khuilay and White Leghorn (MCW0248). Table 4 shows the dissimilarity matrices between pairs of populations. A neighbor-joining tree based on Nei's unbiased genetic distance matrices revealed that Khuilay (Bhutanese naked neck) was most closely related to Pradhu Hang Dam (Thai native black). The other three Bhutanese strains, Seim (Red Junglefowl-like), Yuebjha Narp (Black plumage) and Phulom (Frizzle), were in a separate group with a node connected to Pradhu Hang Dam. These findings indicated that Bhutanese native chickens should be classified as being genetically close to Southeast Asian domestic chickens; they also showed that Bhutanese 606

Genetic relationship and phylogenetic tress
Dorji et al.  native chickens and Thai native chickens (G. g. domesticus) were related to G. g. spadiceus, the red earlobe Red Junglefowl (Figure 1). The relatedness of Khuilay and Pradhu Hang Dam and the separate genetic group formed by the other Bhutanese native chickens were confirmed in the PCA plot. This plot also showed that commercial broilers and the developing line White Leghorn were related to G. g. gallus, the white earlobe Red Junglefowl (Figure 2).

Microsatellite allele diversity and population diversity
The results of this study indicate that the selected loci were reliable and informative because more than four alleles per locus were examined (Nassiri et al., 2007(Nassiri et al., , 2009). Correspondingly, the estimated genetic distances were precise because the standard error was likely to be low (Nassiri et al., 2007). The H E for all loci was > 0.50 and supported the effectiveness of the selected loci.
The MNA per locus calculated for 10 chicken populations was 14.17 ± 0.93 and was similar to the value of 14.00 ± 1.69 recorded in 20 chicken breeds based on 14 markers shared with our study (Rosenberg et al., 2001). In contrast, our value was greater than the 10.00 ± 1.12 reported for 52 chicken populations with 12 shared markers from a set of 22 markers (Hillel et al., 2003), 10.11 ± 0.59 reported for six South African local chicken lines based on nine shared markers (van Marle-Köster et al., 2008) and 10.33 ± 4.33 reported for six Indian chicken populations based on three markers shared with our study (Pirany et al., 2007). Population-specific alleles and/or allele scoring bias (allele dropout, null alleles) could explain these discrepancies in the number of alleles/locus (Nassiri et al., 2007).
Although genetic analyses can reveal the extent of biodiversity in chicken breeds (Nassiri et al., 2007;Semik and Krawczyk, 2011) additional information on specific adaptations, distinct phenotypes, performance level, demography (including effective population size, and geographical distribution), and descriptive databases are required for adequate assessment of each breed when deciding on conservation and breeding programs (Groeneveld et al., 2010).
Genetic characterization of Bhutanese chickens 607 Figure 1 -A phylogenetic tree based on Nei's genetic distance DA 309 (Nei, 1978) for ten chicken populations. The high number of alleles at various loci and the fairly high F IS values may partly reflect the influence of environmental factors and geographical barriers. Although the mean F IS value was high, there was no significant deviation from HWE in native chickens and Junglefowl chickens. On the other hand, eight loci in Broilers and two in White Leghorn deviated from HWE indicating that decades of intensive selection for morphology and production in commercial populations had resulted in genetic subdivision. Some of the loci identified here may be associated with genes that were lost through genetic drift; this could explain why some loci showed strong genetic differentiation while others showed only slight drift. However, the mean F ST value indicated that subpopulation division was moderate, with 8.4% of the total genetic variation being caused by interpopulation differences while 91.6% corresponded to intrapopulation differences.
Comparable population variations were observed for Seim and Khuilay in relation to the original and ancestor fowl populations. Seim chickens are commonly reared by Bhutanese farmers whereas the Khuilay variety has a highly diversified plumage color (soft-red, white, black, partridge and speckled), with possible gene flow from Indian Naked neck populations. The major concern here relates to the Yuebjha Narp variety, which showed low variation. Possible reasons for this reduced variability include lower morphological diversity and finite population sizes (~20-25 individuals per village). As expected, the H E across the loci for the two subspecies of Red Junglefowl was greater than for White Leghorn and higher than that reported by Hillel et al. (2003) and Granevitze et al. (2007).
As shown here, the wild progenitor of domestic chickens contains considerable genetic variation, as also reported for Red Junglefowl in northern India (Mukesh et al., 2011). The wild ancestors of major livestock species are important reservoirs of genetic diversity reservoirs but are either extinct or low in numbers (Hanotte and Jianlin, 2005). Consequently, there is a need for a concerted effort to conserve the putative wild ancestors of present-day chickens, particularly because of the increasing habitat loss and fragmentation, as well as poaching, that threaten the wild varieties with extinction. In contrast, commercial lines have been developed from only a few breeds and therefore have a less varied genetic background, i.e., they have less genetic variation than native and Junglefowl populations. Interestingly, the genetic variation seen here was similar to that reported elsewhere (Pirany et al., 2007).

Genetic relationship and phylogenetic trees
The neighbor-joining (NJ) tree constructed from microsatellite data showed that the two Red Junglefowl subspecies, Gallus gallus gallus and Gallus gallus spadiceus, belonged to different subpopulations. The relatedness of Bhutanese Khuilay (Naked neck) and Thai Pradhu Hang Dam revealed the importance of genetic background in determining heat tolerance. Several reports have demonstrated an association between heat tolerance in Naked neck chickens (Merat, 1986;Yunis and Cahaner 1999;Patra et al., 2002) and the occurrence of heat shock protein 70 (HSP70; Mazzi et al., 2003;Duangdeun, C., 2008, MSc thesis, Khon Kaen University, Thailand). In addition, the highest frequency (~50%) of the HSP70 genotype associated with heat tolerance was found in Pradhu Hang Dam (Tunim et al., 2010). The phylogenetic tree showed that Khuilay and Broiler chickens were sufficiently genetically similar to suggest that Khuilay chickens may be suitable for meat production, with the advantage that they are heat tolerant. The NJ tree and PCA plot confirmed that Bhutanese Seim (Red junglefowl-like), Yuebjha Narp (Black plumage), and Phulom (Frizzle) varieties were distinguishable from another chicken strains; this distinction highlights the importance of genetic diversity among Bhutanese native chickens. The Thai Chee breed may be particularly valuable as a source of genetic variability because it is close to the root of the phylogenetic tree.
The PCA plot provided useful information when the NJ method could not differentiate among closely related chicken populations. This plot showed that three of the Bhutanese native varieties formed a group that fell in a different quadrant from Thai native chickens (G. g. domesticus), Red Junglefowl (G. gallus) and commercial lines (G. g. domesticus). This divergence may reflect environmental adaptation and breeding history (mating system) that affected the genetics of Bhutanese chickens.
In conclusion, this preliminary study of four Bhutanese native chicken varieties based on 18 microsatellite loci clearly demonstrated the genetic diversity of these chickens and reinforced the socio-cultural and economic importance of native chickens in Bhutan. In addition, the PCA analysis showed that Bhutanese native chickens are important contributors to the general poultry gene pool.