Introduction

The genetic diversity of wild animals reflects their evolutionary potential and the quality of their genetic resources, which are also important genetic parameters for evaluating the status of populations of specific species1,2,3. With the development of molecular scatology techniques, the collection and extraction of fecal DNA have been widely used to analyze the genetic diversity of wild animals4,5,6,7. Compared to nuclear DNA, mitochondrial DNA (mtDNA) has multiple copies, simple structure, and high stability8,9. The mtDNA control region (CR) is a non-coding region of the mitochondrial genome rich in A and T bases and has the fastest evolutionary and highest polymorphism rates10,11. Therefore, it has been widely applied in research on genetic diversity, population genetic structure, subspecies, and population differentiation in wild animals12,13,14,15.

Forest musk deer (Moschus berezovskii) is endemic to East Asian forests. Males possess a musk gland that secretes musk, a valuable raw material for traditional medicine and the perfume industry, making this species a valuable medicinal and economic resource16. Owing to illegal poaching, overexploitation, and habitat fragmentation, wild populations of forest musk deer have declined considerably over the past few decades17. They are listed as an endangered species on the International Union for Conservation of Nature Red List18 and is a first-class state-protected wildlife species in China. Only one previous study has investigated the genetic diversity of wild forest musk deer19, whereas another study has combined wild and captive populations20. Located in the Palearctic biogeographic region, the wild musk deer population in Shanxi Province has historically been classified as the Siberian musk deer (Moschus moschiferus)17,21. However, based on the molecular identification by our research team, these populations should be reclassified as forest musk deer (to be reported elsewhere). The current genetic diversity of these populations in Shanxi Province remains poorly understood. Therefore, the present study conducted fecal sampling across all known distribution areas of forest musk deer in Shanxi Province, analyzed their genetic diversity via mtDNA CR, and investigated the correlation between mtDNA CR genetic distances and geographical distances. Based on these findings, conservation strategies for forest musk deer in Shanxi Province were proposed.

Results

Sample collection and species identification

In this study, 243 fecal samples were collected along the sampling lines (Fig. 1) from 5 nature reserves in Shanxi Province. After DNA extraction and polymerase chain reaction (PCR) amplification, 217 fecal samples successfully yielded the target mtDNA CR sequences. Using BLAST searches, sequence assembly, and multiple sequence alignment, 129 forest musk deer mtDNA CR sequences were obtained. The specific sampling numbers and species identification results for the mtDNA CR sequences from each nature reserve are presented in Table 1. The ZT mountain range had the highest number of forest musk deer fecal samples, whereas the TY mountain range had the lowest. Among the individual nature reserves, the number of samples were distributed from highest to lowest as follows: Shanxi Lishan National Nature Reserve (LS) > Yangcheng Mangrove Gibbon National Nature Reserve (MH) > Shanxi Lingkongshan National Nature Reserve (LKS) > Shanxi Luyashan National Nature Reserve (LYS) > Shanxi Pangquangou National Nature Reserve (PQG).

Fig. 1
figure 1

Study area and sampling lines. Shanxi Luyashan National Nature Reserve (LYS), Shanxi Pangquangou National Nature Reserve (PQG), Shanxi Lingkongshan National Nature Reserve (LKS), Shanxi Lishan National Nature Reserve (LS), Yangcheng Mangrove Gibbon National Nature Reserve (MH).

Full size image
Table 1 Number of collected samples, obtained mtDNA sequences, and identified species in Shanxi Province.
Full size table

Analysis of mtDNA CR sequences

The effective length of 129 forest musk deer mtDNA CR sequences was 656 bp. The average nucleotide compositions were T (32.5%), C (14.7%), A (30.0%), and G (22.8%). The average A + T content (62.5%) was greater than the average G + C content (37.5%), indicating a typical A + T bias characteristic of vertebrate mitochondrial genomes. There was 1 deletion, 612 conserved, and 43 variable (polymorphic) sites: 8 single-nucleotide variants and 35 parsimony-informative sites. Among the single-nucleotide variant and parsimony-informative sites, there were 8 and 35 two-variant sites, respectively. No three- or four-nucleotide variant sites were observed.

Haplotype analysis

Twenty haplotypes were identified from 129 mtDNA CR sequences. Using the EstimateS 9.1.0 software, the abundance-based coverage estimator (ACE) value was 22.48, suggesting that the haplotype richness of the target forest musk deer population in the surveyed area was approximately 22 haplotypes. Considering sampling effectiveness, the observed 20 haplotypes accounted for approximately 89% of the estimated total haplotype richness, indicating that most haplotypes in the study population were identified. The haplotype accumulation curve (S. Figure 1) exhibited a rapid initial increase, followed by leveling off, suggesting that the sample size was large enough to identify the majority of mtDNA haplotypes and the sampling was sufficient for this kind of data analysis.

As shown in S. Table 1, 15 of the 20 haplotypes were observed exclusively within a single nature reserve in the whole study area, making them unique haplotypes. The ZT mountain range had the highest number of unique haplotypes, whereas the TY mountain range had the lowest. One haplotype (Hap2) was shared between the LL and TY mountain ranges; however, no shared haplotypes were detected between these two mountain ranges and ZT mountain range. At the nature reserve level, LYS and LKS had only a single haplotype each, with the LYS haplotype (Hap1) being endemic, whereas that of LKS haplotype (Hap2) was shared with that of PQG. The five D-loop sequences from the PQG contained three haplotypes (Hap2, Hap3, and Hap4), with Hap3 and Hap4 being endemic to this reserve. LS had the highest number of haplotypes, with 15 haplotypes out of 20, accounting for 75% of the haplotypes identified. Furthermore, 11 of the 15 haplotypes were endemic to LS, accounting for 55% of the total number of identified haplotypes. Among the 24 D-loop sequences from MH, there were 5 haplotypes, with Hap20 endemic to this site. As shown in S. Table 1, Hap6 was the most common haplotype, observed in 23 of 129 samples (17.83%). Hap5 and Hap14 were relatively common, occurring 16 and 15 times (12.40% and 11.62%), respectively. Five haplotypes (Hap8, Hap9, Hap13, Hap17, and Hap19) were found only once, representing rare haplotypes, and all of these were detected in the LS reserve.

As shown in Table 2, the Hd of the 20 haplotypes was 0.916, Pi was 0.01505, and the average number of nucleotide differences (k) was 9.858. Among the native reserves, the populations in LS had the highest genetic diversity parameters. Contrastingly, LYS and LKS had the lowest genetic diversity. The haplotype polymorphism sites are presented in S. Figure 2.

Table 2 Genetic diversity parameters of forest musk deer in Shanxi Province.
Full size table

An NJ phylogenetic tree was constructed based on the genetic distances of the mtDNA CR haplotypes generated from Shanxi Province (Fig. 2). The 20 haplotypes were divided into 3 major clades, each containing rare haplotypes. NJ tree clustering (Fig. 2) and median-joining network (Fig. 3) yielded consistent results. Haplotypes from the five nature reserves were interspersed, forming the largest dominant clade, Clade I, which included 11 haplotypes. Clades II and III comprised haplotypes from LS and shared haplotypes from LS and MH, containing four and five haplotypes, respectively.

Fig. 2
figure 2

Neighbor-joining phylogenetic tree of wild forest musk deer mtDNA D-loop haplotypes in Shanxi Province. Shanxi Luyashan National Nature Reserve (LYS), Shanxi Pangquangou National Nature Reserve (PQG), Shanxi Lingkongshan National Nature Reserve (LKS), Shanxi Lishan National Nature Reserve (LS), Yangcheng Mangrove Gibbon National Nature Reserve (MH).

Full size image
Fig. 3
figure 3

Median-joining network of wild forest musk deer mtDNA D-loop haplotypes in Shanxi Province. Shanxi Luyashan National Nature Reserve (LYS), Shanxi Pangquangou National Nature Reserve (PQG), Shanxi Lingkongshan National Nature Reserve (LKS), Shanxi Lishan National Nature Reserve (LS), Yangcheng Mangrove Gibbon National Nature Reserve (MH). The black nodes represent missing haplotypes, and the number of short lines on the segments between haplotypes indicates the number of mutational steps. The I-III numbers refer to Clad I-III.

Full size image

The second NJ phylogenetic tree (S. Figure 3) was constructed based on the genetic distances of collected CR haplotype data for four additional populations, including captive and wild forest musk deer from Shaanxi and Sichuan Provinces, and the haplotypes generated from Shanxi Province. This tree included a total of 93 haplotypes, each trimmed to a length of 571 bp. The haplotypes of the Shanxi, Shaanxi and Sichuan populations (both wild and captive) were not completely separated on the phylogenetic tree, with some haplotypes intermingling Additionally, haplotypes of captive populations (from Shaanxi and Sichuan) partially overlapped with those of wild populations. However, the haplotypes of the Shanxi population showed some degree of differentiation from those of the Shaanxi and Sichuan populations. Although haplotypes from different populations were not completely separated on the phylogenetic tree, the majority of haplotypes were geographically distinct, with only 8 out of 93 haplotypes shared between different Provinces.

Population genetic structure

Based on the genetic distances between the haplotypes and the mountain ranges where the nature reserves were located, the five nature reserves were divided into two geographic groups: the Lüliang and Taiyue (LLTY) mountain range group and the Zhongtiao (ZT) mountain range group.

As illustrated in Table 3, the average genetic distance between LS and MH was the largest, whereas that between LYS and LKS was the least. The internal genetic distance within LS was the greatest, whereas LYS and LKS, with only one haplotype each, had the least internal genetic distance of 0. The genetic distance between the LLTY and ZT mountain range groups was greater than that within the LLTY group but smaller than the distance within the ZT group.

Table 3 Mean genetic distances between (under the diagonal) and within (on the diagonal) the wild forest musk deer populations in Shanxi Province.
Full size table

As illustrated in Table 4, there were significant genetic differences (P < 0.05) between the LKS and the other four nature reserves, LS and LYS, and LLTY and the ZT mountain range groups. No significant genetic differences (P > 0.05) were found among the other populations. Gene flow was greatest between the LS and MH populations (12.20) and the least between the LYS and LKS populations (0).

Table 4 The genetic differentiation (Fst, under the diagonal) and gene flow (Nm, above the diagonal) between the populations.
Full size table

AMOVA (Table 5) revealed that much of the genetic variation in the wild forest musk deer populations from the five nature reserves in Shanxi Province was within the populations (88.58%), whereas genetic differentiation among the populations was relatively low (11.42%). The highly significant Fixation Index of 0.1142 indicated that genetic differences among populations were unlikely caused by random factors, indicating that more genetic variation occurred within the reserve populations than among the five reserves in Shanxi Province.

Table 5 Analysis of molecular variance of five wild forest musk deer populations in Shanxi Province.
Full size table

Population historical dynamics

The results of neutrality tests (Tajima’s D and Fu’s Fs) revealed positive values that were not statistically significant (Table 6), indicating that the wild forest musk deer populations in the five nature reserves of Shanxi Province were consistent with that of the neutral selection model and the populations had not expanded. Population dynamics analyses were not conducted for the LYS and LKS nature reserves, each containing only one haplotype. The mismatch distribution curve exhibits a multimodal pattern (Fig. 4) that does not fit the population expansion model. Collectively, these results suggest that wild forest musk deer populations in the five nature reserves of Shanxi Province may not have undergone rapid expansion events.

Table 6 Tajima’s D and Fu’s Fs tests for five wild forest musk deer populations in Shanxi Province.
Full size table
Fig. 4
figure 4

Mismatch distribution curves of the five wild forest musk deer populations in Shanxi Province.

Full size image

Mantel analysis of geographic and genetic distances

A Mantel test was performed using GenAlEx software to analyze the correlation between geographic and genetic distances for wild forest musk deer populations in Shanxi Province (Fig. 5). The results revealed that the correlation between geographic and genetic distances was weak (Rxy = 0.009) and not statistically significant (P = 0.390). Furthermore, regression analysis revealed that geographic distance had an extremely low explanatory power for genetic distance (R2 = 0.0061).

Fig. 5
figure 5

Correlations between genetic and geographic distances for wild forest musk deer in Shanxi Province.

Full size image

Discussion

Our study was based on 129 mtDNA CR sequences from five nature reserves across three mountain ranges in Shanxi Province, which defined 20 haplotypes. The Hd was 0.916, and the Pi was 0.01505. In contrast to other wild forest musk deer populations, the Hd of the wild population in Shanxi Province was greater than that of the wild population in the Baishui River, Sichuan Province (0.7213)19 but slightly lower than that of the wild and captive populations in Qinling, Shaanxi Province (0.931)20. Compared with captive forest musk deer populations, Hd in Shanxi Province was greater than that in the captive population in the Shaanxi Province (0.908)22 but lower than that in the captive population in the Sichuan Province (0.934)23. However, the Pi was lower than that of the aforementioned four forest musk deer populations 19,20,22,23. Compared with those of other musk deer species, the Hd and Pi of the wild forest musk deer in Shanxi Province were greater than those of the wild Alpine musk deer in the Helan Mountain (Hd = 0.716 and Pi = 0.00156)24 but lower than those of the wild populations of the Siberian musk deer in Northern Russia (Hd = 0.970 ± 0.018 and Pi = 0.026 ± 0.005)25, indicating that the wild forest musk deer population in Shanxi Province maintains a relatively high genetic diversity, making it a valuable genetic resource with important conservation value26.

The LLTY mountain range group (consisting of the LYS, PQG, and LKS nature reserves) contained 4 haplotypes, whereas the ZT mountain range group contained 16 haplotypes, and the 2 groups were distinct, indicating that the LLTY and ZT mountain range groups harbor unique genetic resources. In comparison, the LS population from the ZT mountain range group had the highest genetic diversity, with 75% of the total and 55% of unique haplotype numbers found in the whole study area. Five rare haplotypes were found in this population, demonstrating the important role of the LS population in maintaining the overall genetic diversity of forest musk deer in Shanxi Province. Furthermore, four of the five haplotypes found in the MH population were shared with the LS population, while the fifth haplotype was unique to the MH population. This further corroborated the critical role of the musk deer population inhabiting the ZT mountain range in preserving the genetic diversity of forest musk deer in Shanxi.

Genetic structure analysis revealed that the LS and MH populations within the ZT mountain range presented insignificant genetic differentiation (Fst < 0.05, P > 0.05), with a high gene flow of 12.20, indicating frequent genetic exchange between the two, likely because they belong to the same mountain range and are approximately 43.8 km apart. Contrastingly, a relatively high level of genetic differentiation was detected between the LLTY and ZT mountain ranges (Fst = 0.20305, P < 0.05)27, with weaker gene flow (Nm = 1.96245) than that within the ZT mountain range. The AMOVA results revealed that much of the genetic variation in the wild forest musk deer populations in the five nature reserves in Shanxi Province exhibited within-population variation, whereas genetic differentiation among populations was relatively low. Combining the genetic distance, genetic differentiation values, and gene flow between the populations, it was determined that much of the genetic variation originated from the ZT mountain range group and, more precisely, from the LS population, indicating that the LS population accounts for the vast majority of genetic variation in wild forest musk deer populations across Shanxi Province.

Although the average genetic distance between the LS and MH populations is the largest among the studied populations, the genetic structure analysis (Fst < 0.05, P > 0.05) revealed insignificant genetic differentiation. The apparent inconsistency can be attributed to the high genetic diversity within the LS population. The LS population contains 15 haplotypes, accounting for most of the haplotype diversity in the Shanxi region, while the MH population has only 5 haplotypes, 4 of which are shared with the LS population. This shared genetic background reduces inter-population differentiation, as reflected by the low and insignificant Fst value. Moreover, AMOVA analysis confirmed that most of the genetic variation originates from populations rather than between populations, further supporting this conclusion. Therefore, the large genetic distance reflects the contribution of LS population’s internal diversity, rather than true differentiation between LS and MH.

Neutrality tests and mismatch distribution curves are commonly used to analyze the historical dynamics of populations. Specifically, when Tajima’s D and Fu’s Fs values are negative and statistically significant28 and the mismatch distribution curve exhibited a unimodal pattern29, indicating that the target population had experienced population expansion. However, the neutrality test results for the wild forest musk deer populations in Shanxi Province had positive values that were not statistically significant, and the mismatch distribution curves exhibited a multimodal pattern, suggesting that forest musk deer in Shanxi Province may not have experienced rapid expansion.

Geographic isolation has been widely accepted as a key factor driving population genetic differentiation30,31,32. However, the Mantel test analysis of wild forest musk deer populations in Shanxi revealed no significant correlation between geographical and genetic distances, suggesting that geographic distance may not be the primary driver of the observed genetic differentiation. We postulate that Shanxi Province, characterized by rugged mountainous terrain interspersed with basins and as the historic cradle of Chinese agricultural civilization33,34,35, presents a landscape in which intense human economic activities within the basin areas have likely impeded gene flow among forest musk deer populations. This landscape-level anthropogenic influence may be the principal reason for the relatively pronounced genetic differentiation observed among the geographical populations of forest musk deer in Shanxi and most of the genetic variation was within rather than among populations.

The captive breeding of forest musk deer first began in Shaanxi Province in 1958, which was only 66 years ago. The captive populations in Shaanxi and Sichuan Provinces are derived from local wild populations, which explains the observed overlap between the haplotypes of the captive populations and those of the wild populations. Additionally, there have been exchanges of genetic resources between the captive forest musk deer in Sichuan and Shaanxi. Shaanxi Province is located in central China, with Shanxi Province situated to its northeast and Sichuan Province to its southwest. Shaanxi shares a border with both Shanxi and Sichuan, indicating historical gene flow among these populations. These factors account for why the haplotypes of the different populations (both wild and captive) from Shanxi, Shaanxi, and Sichuan were not completely separated in the phylogenetic tree, with some haplotypes intermingling. However, despite this gene flow, the haplotypes of the Shanxi population exhibited a certain degree of differentiation compared to those from Shaanxi and Sichuan. Moreover, the majority of haplotypes were geographically distinct (only 8 out of 93 haplotypes shared between different Provinces), suggesting that the forest musk deer populations in the different Provinces are to a large extent genetically unique. The haplotypes of the wild populations (particularly those in Shanxi) were not fully represented by the captive populations. Therefore, it is recommended to enhance the protection of the wild populations with high genetic diversity and to incorporate more genetic diversity into the genetic management of captive populations to avoid the risks of genetic drift and inbreeding.

Conclusion

Given that the wild forest musk deer populations in Shanxi Province still maintain relatively high genetic diversity, particularly with the ZT mountain range group occupying a prominent position in genetic variation, we propose the following conservation strategy from an ecological genetics’ perspective. Rather than viewing the genetic diversity of the different mountain range system populations in isolation, we recommend considering them as a metapopulation genetic resource. The implementation of strategic interchanges of genetic material among mountain range groups could be an important management approach for improving and enriching the overall genetic diversity of various local populations. This genetic rescue strategy may be crucial for promoting the recovery and long-term evolutionary resilience of wild forest musk deer populations across Shanxi Province.

Materials and methods

Study area

Shanxi Province is in the mountainous area of the southern Loess Plateau and has a temperate continental monsoon climate with four distinct seasons: hot and rainy summers and cold and dry winters. The average annual temperature is 8.5 °C, with January and July being the coldest and hottest months, with average temperatures of approximately -8.5 °C and 23.5 °C, respectively36. The LL, ZT, and TY mountain ranges have elevations ranging from 1,500 to 3,000 m, and precipitation increases with increasing elevations ranging from 400 to 600 mm. Vegetation varies with altitude, presenting as crops, shrublands, mixed coniferous-deciduous and coniferous forests, and subalpine meadows. The dominant tree species were Picea asperata, Larix gmelinii, Pinus tabuliformis, and Quercus mongolica37,38. These three mountain ranges are home to several national first-class key-protected endangered species, including forest musk deer, leopards, and Chinese Grouse39.

The present study was conducted from February to October 2023 in five national nature reserves across three mountain ranges in Shanxi Province: Shanxi Luyashan National Nature Reserve (LYS) and Shanxi Pangquangou National Nature Reserve (PQG) from the LL mountain range; Shanxi Lishan National Nature Reserve (LS) and Yangcheng Mangrove Gibbon National Nature Reserve (MH) from the ZT mountain range; and Shanxi Lingkongshan National Nature Reserve (LKS) from the TY mountain range (S. Table 2 and Fig. 1). These areas were selected because they are known habitats of wild forest musk deer populations, and systematic surveys and fecal sample collections were carried out at these sites.

Sample collection

We reviewed historical grid-based infrared camera data and interviewed local herb collectors, shepherds, and experienced forest rangers to preliminarily determine the potential locations where forest musk deer may appear. Guided by these local informants, we visited areas where forest musk deer had been previously observed and collected suspected forest musk deer feces along survey routes. The forest musk deer fecal pellets were bluntly elliptical, elongated, and cylindrical, approximately 9–10 mm long and 5–6 mm wide40. Owing to the strong territoriality of musk deer, we collected fecal samples at least 300 m apart (considering the territory range of published literature)41,42,43,44 to minimize the likelihood of duplicate sampling. All fecal samples that fit the described morphological characteristics were collected using disposable polyethylene gloves and placed in 25-mL sterile RNase-free centrifuge tubes, with 30 pellets collected per pile. Sample details, including the collection date, locality, sample ID, and freshness of the feces, were recorded. Upon completion of the daily sampling, all samples were stored at -20 °C until further processing.

DNA extraction and polymerase chain reaction amplification

Fecal samples were preprocessed according to the methods described by Piggott45 and Tang46. Approximately eight fecal pellets were processed into cell suspensions, and genomic DNA was extracted from the fecal samples using a blood/cell/tissue genomic DNA extraction kit (Tiangen Biotech Co., Ltd., Beijing, China). The extracted DNA was stored at -20 °C until subsequent analyses.

The PCR system had a total volume of 25 μL, consisting of 12.5 μL of 2 × Es Taq MasterMix Dye (CoWin Biosciences, Cambridge, MA, USA), 4 μL bovine serum albumin (Thermo Fisher Scientific, Waltham, MA, USA), 30–50 ng of DNA template, 1 μL each of forward and reverse primers (10 μM), and the remaining volume was filled with double distilled H2O.

The mitochondrial CR primers used were from Feng47, with the forward primer F: 5′-CAACTAACCTCCCTAAGACTTCAAG-3′ and the reverse primer R: 5′-CCAAATGTATGACAGCACAGTTATG-3′. The PCR program was as follows: 95 °C predenaturation for 1 min; 94 °C predenaturation for 4 min; 94 °C denaturation for 30 s; 56 °C annealing for 45 s; 68 °C extension for 1 min; 35 cycles; final extension at 68 °C for 10 min; and storage at 4 °C.

After PCR, 5 μL of the amplified product was subjected to 1% agarose gel electrophoresis to check for a clear and single target band. The qualified PCR products were sent to Sangon Biotech Co., Ltd., Shanghai, China, for purification and Sanger sequencing in the forward and reverse directions.

Data analysis

Successfully sequenced mtDNA CR sequences were searched for and compared with existing forest musk deer sequences in the National Center for Biotechnology Information database. Samples with more than 98% similarity to known forest musk deer sequences were considered to belong to the target species. The forward and reverse sequences identified as forest musk deer were assembled and trimmed using SeqMan Pro software in DNASTAR Lasergene 7.1 (DNASTAR, Madison, WI, USA), with manual corrections48. All sequences were aligned and analyzed using Molecular Evolutionary Genetic Analysis (MEGA) 11, and nucleotide composition and computed genetic distances were analyzed based on the Kimura 2-parameter model49. DnaSP v5 (DNASTAR) was used to calculate the number of polymorphic sites, haplotype number (h), nucleotide diversity (Pi), haplotype diversity (Hd), and average number of nucleotide differences (K) for each nature reserve and to generate mismatch distribution plots50.

The haplotype sequences were organized and used to construct a neighbor-joining (NJ) phylogenetic tree using MEGA 11 with 1,000 bootstrap replications49. The tree was then beautified via TVBOT software51. A haplotype network was constructed using POPART 1.7 (University of Otago, Dunedin, New Zealand)52. Further analyses, including analysis of molecular variance (AMOVA), pairwise genetic differentiation (Fst) calculations, and Tajima’s D and Fu’s Fs tests, were conducted using Arlequin 3.5 (Swiss Institute of Bioinformatics, Geneva, Switzerland)53. The total number of haplotypes was estimated using the ACE in EstimateS 9.1.0 (Swiss Institute of Bioinformatics)54 and a haplotype accumulation curve was plotted.

The genetic distances between samples were calculated using GenAlEx 6.51 (Swiss Institute of Bioinformatics)55, and the geographic distances between sampling locations were determined based on the coordinates. The Mantel test was performed to explore the correlation between the genetic and geographic distances of forest musk deer populations.

Finally, collecting CR haplotype data for four additional populations in GenBank from previous studies19,20,22,23, including captive and wild forest musk deer from Shaanxi and Sichuan Provinces. Additionally, the CR haplotype sequence of Bos taurus was included as an outgroup to root the phylogenetic tree. By combining these data with the haplotype data from this research, the second phylogenetic tree was constructed. All the CR haplotype sequences above were trimmed to 571 bp to ensure consistent alignment.