Integrating whole-genome resequencing data reveals adaptive selection signatures under extreme environments of goat populations in China

Abstract

With the increasing frequency of extreme weather events globally, livestock are facing ever-increasing demands for environmental adaptability. Some goat populations have demonstrated remarkable adaptability under long-term extreme environmental stress, and their genomes retain characteristic genetic imprints formed by adaptive evolution. This study integrated whole-genome resequencing data from 496 individuals from 29 endemic goat populations in China, along with data from 11 environmental factors. Through signal selection analysis and genome-environment association analysis, we identified 262 candidate genes associated with environmental adaptation. Subsequently, we screened Chinese goat populations in extreme environments based on environmental data. We used the FST and θπ methods to scan the genomes of these goat populations. In four extreme environments (high elevation, hot, and arid), 95, 74, and 162 candidate genes were identified, respectively. Based on research related to environmental adaptation, we found that genes such as FGF5, CAMK4, COL4A3BP, and CACNG4 may play important roles in goat adaptation to extreme environments. This study not only deepens our understanding of the evolutionary mechanisms of environmental adaptation in Chinese goats, but also provides important theoretical basis for goat stress resistance breeding and germplasm resource conservation.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12864-026-13176-x.

Background

With the intensification of global climate change, the frequency and intensity of extreme climate events (such as heat waves, cold waves and abnormal precipitation) have become more frequent and intense [1], posing a severe challenge to the livestock production system [2]. China has a vast territory and complex climate types, and livestock in different ecological regions face different and constantly changing environmental pressures [3]. As an important livestock resource, the production performance and survival ability of Chinese goats are directly affected by climate change. Goats were domesticated in the Fertile Crescent about 10,000 years ago [4]. After being gradually introduced to China, they are widely distributed in the Qinghai-Tibet Plateau, Loess Plateau, southwestern mountainous areas and southeastern coastal areas. Under the long-term effects of natural and artificial selection, Chinese goats have developed rich genetic diversity, making them a valuable resource for studying the mechanism of animal environmental adaptation.

Extreme environmental stress is considered to be one of the important driving forces shaping the structure and function of animal genomes [5]. For Chinese goats, their long-term distribution in various ecological environments such as high elevation, extreme cold, aridity and high humidity has evolved them to develop unique genetic adaptation characteristics in the process of coping with complex climatic conditions [6]. Under cold stimulation, the energy metabolism in goats is significantly reprogrammed, manifested as increased plasma glucose levels and decreased levels of non-esterified fatty acids and triglycerides, accompanied by increased liver glycogen content and decreased liver lipid deposition [7]. These metabolic changes help maintain the body’s energy supply and body temperature homeostasis, thereby enhancing their ability to survive in cold environments. For high-elevation populations such as Tibetan goats distributed on the Qinghai-Tibet Plateau, they have formed specific genetic regulatory patterns in terms of hypoxic adaptation, energy metabolism regulation and cold stress resistance [8]. These studies provide important clues to reveal the adaptive evolution of Chinese goats in extreme environments, but they mostly rely on specific environmental interventions or single-population studies, which are not only time-consuming and costly, but also have certain limitations in their universality across different ecological regions. Therefore, it is necessary to combine multi-population comparison and genome-scale analysis strategies to systematically explore the adaptive genetic basis of Chinese goats in diverse ecological environments, so as to more comprehensively reveal the molecular mechanisms underlying their responses to complex climate change.

In recent years, with the development of high-throughput sequencing technology and population genomics, key genes and regulatory regions subjected to natural selection at the genome-wide level have been identified [9]. However, in commercial breeding, performance-oriented artificial selection often weakens or even masks genetic signals related to environmental adaptation. In contrast, indigenous goat breeds having been under relatively natural selection pressure for a long time, are more likely to reflect their adaptation process to specific ecological environments. Therefore, they have important research value for dissecting the mechanisms of adaptive evolution of animals under the background of climate change [10]. The continuous improvement of the global high-resolution climate databases provides critical data support for the systematic analysis of the environmental adaptation mechanism of Chinese goats. By integrating population geographic distribution information with multidimensional climate variables, genome-environment association (GEA) analysis can be performed to systematically reveal the interplay between genetic variation and environmental factors [11]. This analytical framework has been widely applied in a variety of livestock and poultry species and has successfully identified adaptive genes related to key ecological factors such as temperature, precipitation and elevation [12–14]. In Chinese goats, such methods can not only help to explore their genetic basis underlying cold tolerance, heat resistance, drought resistance and hypoxia adaptation, but also identify local breeds that have long inhabited extreme environments for a long time by combining long-term climate data [15], and analyze the selection signals and genetic characteristics formed elucidating the evolutionary trajectory of environmental adaptation of Chinese goats and its molecular regulatory mechanism.

To reveal the adaptation mechanisms of Chinese goats in different extreme environments, this study integrated whole-genome resequencing data of 496 individuals from 29 endemic goat populations in China and data from 11 environmental variables. By studying their population genetic structure and environmental clustering characteristics, Chinese goat populations in extreme environments, such as high elevation, heat, cold and drought, were defined and screened. Using selection signal analysis and GEA, we identified genomic regions and candidate genes associated with adaptation to extreme environments, elucidating their biological pathways for coping with challenging environments. This study analyzed the adaptation mechanisms of goats to extreme environments during long-term evolution, providing a scientific basis for the protection and rational utilization of China’s goat genetic resources and providing molecular markers for the breeding of environmentally adaptable goat breeds.

Materials and methods

Genomic data preparation and preprocessing

This study collected high-depth whole-genome resequencing data from 29 goat populations in my country. Data from 28 of these populations were downloaded from public databases, while the data from the Lüliang Black Goat population came from our research group’s previous independent sequencing data. (Table S1). Raw sequencing reads were quality controlled using fastp (v0.23.4) [16] to obtain high-quality, clean data for subsequent analysis. Filtering criteria included removing reads with > 1% base ambiguity, abnormal GC content, low-quality bases, and adapter contamination. Subsequently, clean reads were aligned to the reference genome (Capra hircus, https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/001/704/415/GCF_001704415.1_ARS1/GCF_001704415.1_ARS1_genomic.gff.gz) using BWA (v0.7.17) [17] Sequencing depth and genome coverage were calculated based on the alignment results, and individuals with a sequencing depth < 10× were removed. SNP quality control was performed using PLINK (v1.9) [18]. We retained only autosomal SNPs and then filtered them to exclude those with a detection rate ≤ 90%, minor allele frequency (MAF) ≤ 0.05, and outliers. Ultimately, 496 individuals were retained from the 510 individuals (Table S2).

Population clustering based on environmental variables

The high-resolution climate dataset (CRU TS v. 4.08) [19] comes from the Norwich Climate Research Unit (https://crudata.uea.ac.uk/cru/data/hrg/cru_ts_4.08/cruts.2406270035.v4.08/). This dataset has a spatial resolution of 0.5°×0.5° and spans from 1901 to 2023, excluding the Antarctic region.

The dataset uses seven climate variables, including cloud cover (cld), diurnal temperature range (dtr), frost day frequency (frs), potential evapotranspiration (pet), precipitation (pre), wet day frequency (wet), and vapor pressure (vap). To more accurately reflect the environmental differences in temperature and elevation, this study further used four variables from WorldClim (v2.1) [20]higher resolution climate data (0.1667°): mean temperature (tavg), maximum temperature (tmax), minimum temperature (tmin), and elevation (elev). Subsequently, ArcMap (v10.8) [21] was used to process the raw climate data into annual and monthly mean land area latitude and longitude grids in the GCS_WGS_1984 coordinate system. Pearson correlation coefficients among 11 environmental variables were calculated in R (v4.2.0) [22] to assess collinearity and interdependencies.

Afterwards, we used the R package “terra” [23] to statistically analyze the annual average and monthly average data of climate variables, as well as the elevation data corresponding to the sampling origins of the 29 Chinese goat populations. Eventually, a comprehensive environmental dataset containing 131 environmental parameters was obtained for principal component analysis (PCA) (Table S3). Then, we used a K-means clustering algorithm to cluster the goat populations based on the Euclidean distance of the principal component scores, and the 29 goat populations were divided into two groups (CL1 and CL2) based on the clustering results. To check whether the grouping based on environmental variables is consistent with the population genetic clustering pattern, we first used PLINK (v1.9) [18] to remove the high LD sites in the data set with the parameter “-indep-pairwise 50 5 0.2”, and the remaining 1,242,943 SNPs from 496 individuals of 29 goat populations (CL1 and CL2) were used for population genetic structure analysis. Subsequently, Genetic PCA analysis was performed using PLINK (v1.9). Neighbor-joining (NJ) phylogenetic trees were constructed using MEGA (v11.0.13) [24] based on genetic distance. Ancestor components were inferred using ADMIXTURE (v1.30) (K = 2–10). Maximum likelihood phylogenetic trees at the goat population level were constructed using TREEMIX (v1.13) [25].

Detection of genomic regions associated with environmental adaptation

To identify genomic regions related to goat environmental adaptation in goats, we used the genetic differentiation index method for selection signal analysis [26] and the latent factor mixture model (LFMM) [27] method for genome-environment association analysis to analyze SNP data of 29 goat populations. Specifically, the analysis was implemented using VCFtools (v0.1.16) [28], employing a sliding window method (window size of 50 kb, step size of 10 kb) to estimate the value for each SNP; LFMM analysis was performed using the “lfmm” in R package [29]. 131 environmental variables were used, and the optimal number of latent factors were inferred using the lfmm_ridge_CV function. 1000 iterations and 100 burn-in times were performed to determine the z scores of all SNPs, and the P values were corrected with the genomic expansion factor to reduce systematic errors and false positive rates. The 50 kb genomic intervals upstream and downstream of the top 1% significant SNP sites were defined as the genomic regions associated with environmental adaptability.

Detecting of genomic regions associated with high-elevation adaptation

To identify genomic regions associated with high-elevation adaptation in goats, we defined high elevation as areas above 3,000 m and low elevation as areas below 100 m. Based on this criterion, we initially selected goat populations from high-elevation areas and low-elevation areas. To reduce the potential bias caused by unequal sample sizes among breeds/populations, individuals were randomly sampled within each breed/population, and the final groups used for the high-elevation adaptation analysis included high-elevation areas (G-high, n = 22) and low-elevation areas (G-low, n = 25). (Table S4) Subsequently, we used VCFtools (v0.1.16) [28] to calculate the values and nucleotide diversity ratios ( ratio, (G-high/G-low)) among the two groups (G-high and G-low) to identify highly differentiated and significantly different genomic regions. The significant SNP site in the top 1% and their upstream and downstream 50 kb regions were defined as genomic regions associated with high-elevation adaptation in goats.

Detecting of genomic regions associated with environmental temperature adaptation

To identify genomic regions underlying extreme temperature adaptation in goats, we classified sampled goat populations into temperature-adaptive groups via equal-interval classification based on three meteorological variables: annual average temperature (tavg), maximum temperature (tmax) and minimum temperature (tmin). The value range of each variable was divided into equal intervals. Populations that fell into the high-temperature intervals for all three variables were defined as the hot group (G-hot, n = 80), while those consistently within the low-temperature intervals were assigned to the cold group (G-cold, n = 49). The Köppen–Geiger climate classification was only adopted to characterize the climatic background of the studied populations [30]. (Table S5) In addition, the values and ratios (G-high/G-low) methods were used to screen for selection signals related to extreme temperature adaptation, and the upstream and downstream 50 kb genomic regions of the top 1% significant SNP sites were defined as genomic regions related to temperature adaptation.

Detecting of genomic regions associated with arid adaptation

The Aridity Index of the United Nations Environment Programme () [31, 32]can be used as an important indicator for quantitatively assessing the degree of regional water shortage and the annual or seasonal aridity. The study used ArcMap (v10.8) [21] to calculate the of the origin of each goat population according to the formula below.

Where PRE represents annual average precipitation (mm), and PET represents potential evapotranspiration (mm). According to the standard, regional climate can be divided into five categories: hyper-arid (0–0.05), arid (0.05–0.20), semi-arid (0.20–0.50), dry sub-humid (0.50–0.65), and humid (above 0.65).

We systematically evaluated the values of 29 goat populations. Since the sample did not include populations in hyper-arid areas, populations distributed in arid, semi-arid, dry sub-humid, and humid environments were initially evaluated for comparison. Four populations located in dry sub-humid regions (= 0.50–0.65), comprising 98 individuals in total, were excluded from the comparison. In addition, to increase the environmental contrast between the arid and wet groups, only three populations with the highest values among the humid populations (> 0.65) were retained as the wet group. After removing the relevant populations, the samples were divided into an arid group (G-arid, n = 62) and a wet group (G-wet, n = 38) (Table S6). The G-arid group mainly includes populations distributed in cold semi-arid grassland climate (BSk), cold desert climate (BWk), and temperate cool summer and dry winter climate (Dwb); while the G-wet group includes populations distributed in subtropical dry winter climate (Cwa), humid subtropical climate (Cfa), and tropical monsoon climate (Am) regions. In addition, the values and ratios (G-arid/G-humid) methods were used to screen for selection signals related to extreme temperature adaptation, and the significant SNPs in the top 1% and their upstream and downstream 50 kb regions were defined as candidate genomic regions related to arid adaptation.

Identification and functional annotation of candidate genes

The genomic regions detected above are mapped to goat gene annotation files in the Ensembl database for gene annotation (https://ftp.ensembl.org/pub/release-112/gff3/capra_hircus/Capra_hircus.ARS1.2.chr.gff3.gz). To analyze the function of the target genes, we performed Gene Ontology (GO) [33] and Kyoto Encyclopedia of Genes and Genomes (KEGG) [34] functional enrichment analyses using the DAVID online software [35]. In this analysis, we retained the functional entries that passed Fisher’s exact test and had P < 0.05.

Result

Goat population quality control and annual environmental data statistics

After quality control and screening, we ultimately retained 496 goat individuals from 29 populations out of 510 individuals. Subsequently, we performed quality control on the SNP loci, ultimately retaining 13,128,570 SNP loci for subsequent analysis.

This study integrated 11 environmental variables, including cloud cover (cld), daily temperature range (dtr), frost frequency (frs), potential evapotranspiration (pet), precipitation (pre), wet day frequency (wet), vapor pressure (vap), mean temperature (tavg), elevation (elev), and maximum and minimum temperatures (tmax and tmin) (Fig. 1A-K). Temperature-related environmental variables (tmax, tmin, and tavg) show higher annual averages in southern and southeastern China, while they are lower in the Qinghai-Tibet Plateau and northern regions. The annual average frequency of frost days (frs) is higher in the Qinghai-Tibet Plateau and northern China, and lower in the southern region, showing an overall trend of gradually decreasing from northwest to southeast. The annual average water vapor pressure (vap) is higher in southern and southeastern coastal areas of China, and lower in the northwest and northern regions, showing an increasing trend from north to south.

Fig. 1.

Geographic distribution of annual average values for 11 environmental variables in China. A Cloud Cover (cld), percentage (%). B Diurnal Temperature Range (dtr), degrees Celsius (°C). C Wet Day Frequency (wet), days per month. D Potential Evapotranspiration (pet), millimeters per day (mm/day). E Precipitation (pre), millimeters per month (mm/month). F Frost Day Frequency (frs), days per month. G maximum temperature (tmax), degrees Celsius. (°C.) H Average Temperature (tavg), degrees Celsius. (°C.) I minimum temperature(tmin), degrees Celsius. (°C.) J Vapor Pressure (vap), hectopascals (hPa). K Elevation (elev), meters (m). L Pearson correlation coefficients among environmental variables, where red indicates positive correlations and blue indicates negative correlations, and color intensity reflects the strength of correlation

In the correlation analysis among the 11 environmental variables (Fig. 1L), the potential evapotranspiration (pet) had a weak correlation with most of the environmental variables, and the correlation with the frost day frequency (frs) was only − 0.02. Furthermore, water vapor pressure (vap) showed a significant positive correlation with precipitation (pre), average temperature (tavg), maximum temperature (tmax), and minimum temperature (tmin) (), indicating ticollinearity between water vapor pressure (vap) and multiple environmental variables. The frequency of frost days (frs) showed a significant negative correlation with temperature-related variables (tavg, tmax, tmin) (). Elevation (elev) showed a significant negative correlation with multiple environmental variables (), and only showed a positive correlation with the frequency of frost days (frs) and daily temperature range (dtr) ().

Cluster analysis of goat population environment

We performed PCA analysis on 131 environmental variables from 29 goat populations (Fig. 2A). The results showed that the eigenvalues of the first three principal components were 106.81, 24.56, and 18.18, respectively, explaining a total of 87.46% of the environmental variation (PC1 = 62.46%, PC2 = 14.36%, PC3 = 10.63%) (Fig. 2B). These 13 water vapor pressure (vap) variables are highly positively correlated with PC1 (mean loading: 0.92 ± 0.04), while the 13 daily temperature range (dtr) variables are strongly negatively correlated with PC1 (mean loading: -0.85 ± 0.04) (Fig. 2C and D; Table S7).

Fig. 2.

Environmental variable analysis of goat populations in China. A Distribution of goat populations along the first two principal components derived from environmental variables, were divided into two groups (CL1 and CL2) (B) Percentage of variance explained by each principal component based on environmental variables. C Contribution and direction of environmental variables to the principal components; arrow length represents the magnitude of contribution. D Loadings of 131 environmental variables on the first two principal components (PC1 and PC2)

Environmental genetic structure analysis of goat populations

Based on the PCA scores, we used Euclidean distance to perform hierarchical cluster analysis on the 29 goat populations, dividing them into two groups, CL1 and CL2 (Fig. 3A). Among them, the goat population in CL1 is mainly distributed in Southwest my country (Chongqing (n = 15), Sichuan (n = 70), Guizhou (n = 49), Yunnan (n = 38)), South China (Guangdong (n = 5), Guangxi (n = 32), Hainan (n = 15)), East China (Shanghai (n = 20), Jiangsu (n = 20)) and Central China (Hunan (n = 20)). The goat population in CL2 is mainly distributed in Northwest China (Shaanxi (n = 20), Gansu (n = 20), Ningxia (n = 20)), North China (Inner Mongolia (n = 20), Shanxi (n = 19)), Southwest China (Sichuan (n = 10), Tibet (n = 22)), as well as Liaoning (n = 20), Hubei (n = 20) and Shandong (n = 40) (Fig. 3B;Table S8).

Fig. 3.

Population genetic structure, geographic distribution and phylogenetic relationships of Chinese indigenous goats. A Heatmap of the top three principal components (PC1-PC3) from genome-wide PCA showing SNP loading values based on environmental variables. B Geographic distribution map of Chinese goat populations C ADMIXTURE ancestry components for K = 2–5. D PCA scatter plot of Chinese goat populations. E Treemix maximum likelihood population tree for Chinese goat populations. F NJ phylogenetic tree of Chinese goat populations

Genetic structure analysis of SNP data from 496 individuals in 29 goat populations (CL1 and CL2) revealed that, with increasing K values (K = 2–5), some goat populations in CL1 and CL2 exhibited varying degrees of ancestral component mixing, suggesting that some degree of gene exchange may have occurred during their evolutionary history (Fig. 3C). The PCA results showed a high degree of consistency with the ADMIXTURE results, with PC1 and PC2 explaining 6.48% and 4.57% of the genetic variation, respectively (Fig. 3D). Phylogenetic trees further support the above results. Among the 29 goat populations, most of the CL1 group clustered into one branch in the phylogenetic tree, while the CL2 group formed another major branch, showing obvious genetic differentiation. Among them, some individuals of the MTG and MGG populations in the CL2 group were observed in the CL1 group branch, indicating that these two populations may have had gene exchange with the CL1 group or recent interbreeding (Fig. 3E). The Treemix maximum likelihood population tree shows that the genetic drift values of most populations are mainly between 0.005 and 0.015, indicating a relatively concentrated branching structure. Furthermore, compared to other populations, YDG exhibits relatively long branch lengths and more pronounced genetic differentiation. (Fig. 3F).

Genomic regions related to environmental adaptation in goat populations

This study used and LFMM to detect genomic regions significantly associated with goat environmental adaptation. screened out 397 significant SNPs. The three SNPs with the highest values were mainly located on chromosome 6 (95.45–95.60 Mb), chromosome 3 (96.5-96.65 Mb), and chromosome 22 (51.7–51.8 Mb). 685 genes were annotated within a 50 kb range upstream and downstream of these SNPs (Table S9), covering key biological functions such as regulation of heart development (TBX20 [36]), male reproductive, glycosylation (SRD5A3 [37, 38]), ion transport and protein glycosylation modification (TMEM165 [39]), intracellular signal transduction (RGS7 [40, 41]) (Fig. 4A).

Fig. 4.

Selection signatures and functional enrichment related to environmental adaptation in Chinese indigenous goats. A Genome-wide values with candidate genes labeled. B Genome-wide distribution of z-scores of SNP loci in the LFMM test. The dashed line indicates the top 0.1% threshold level; C Shared genes detected by and LFMM. D Functional enrichment analysis of 262 genes identified using at two methods

LFMM results showed that, based on a determined number of latent factors (K = 5), environmental variable-genome association analysis identified 5,754 SNPs that were significantly associated with environmental variables. The three SNPs with the highest z-values were located on chromosome 28 (41.61 Mb), chromosome 19 (28.1 Mb), and chromosome 6 (95.43–95.56 Mb), respectively. The regions containing these SNPs annotated 2,664 genes (Table S10), covering biological functions such as angiogenesis (MMRN2 [42]), cell migration (EPHA2 [43], FMNL1 [44]), and cell cycle regulation (CDC25A [45]) (Fig. 4B).

The method (n = 685) and the LFMM method (n = 2664) detected 262 shared genes across the gene sets (Fig. 4C; Table S11). A permutation test with 10,000 iterations showed that the observed overlap was significantly higher than expected by chance (expected overlap = 67.25 genes, empirical P = 9.999 × 10⁻⁵), suggesting that the shared genes were unlikely to result from random overlap. Subsequent functional enrichment using DAVID revealed that these genes were significantly enriched in 19 biological processes (BP), 12 cellular components (CC), 25 molecular functions (MF) categories, and 2 KEGG pathways (P < 0.05) (Fig. 4D; Table S12).

Genomic regions associated with high-elevation adaptation in goat populations

This study used the and ratio method to detect selection signals associated with high-elevation adaptation in goats. Using a 1% threshold, detected a total of 398 SNP sites, covering 607 genes (Table S13). The three sites with the highest values are located on chromosome 13 (22.15-22.35 Mb), chromosome 6 (95.45-95.55 Mb) and chromosome 20 (38.8-38.9 Mb). The regions where thes [46]e SNPs are located are annotated with these genes, covering genes related to hematopoietic development (SKIDA1 [47]) and reproduction and metabolism (PRLR [48]) (Fig. 5A). The ratio method detected a total of 399 SNP sites and annotated a total of 883 genes (Table S14). Among them, the three sites with the highest ratio were located on chromosome 5 (18.1-18.2 Mb), chromosome 6 (95.45-95.55 Mb), and chromosome 10 (94.35-94.45 Mb), covering genes related to key adaptive traits, including hair growth regulation (FGF5 [49]) and lipid transport (COL4A3BP [50]) (Fig. 5B). A total of 95 genes were detected by both methods (Fig. 5C; Table S15). They were significantly enriched in 13 biological processes (BP), 4 cellular components (CC), 9 molecular functions (MF) categories, and 1 KEGG pathways (P < 0.05) (Fig. 5D; Table S16).

Fig. 5.

Selection signatures and functional enrichment associated with high-elevation adaptation in Chinese indigenous goats. A Genome-wide distribution of values between G-high and G-low. B Genome-wide distribution of ratio between G-high and G-low. C Shared genes detected by ratio and . D Functional enrichment analysis of 95 genes identified using at two methods

Genomic regions associated with drought adaptation in goat populations

To explore the regulatory mechanisms by which goats adapt to arid environments, we screened goat populations located in arid and humid regions based on the United Nations Environment Programme Drought Index () and the Köppen-Geiger climate classification (Fig. 6A). We used the method and the ratio method to explore selection features related to drought adaptation in goats. At the top 1% threshold, the method detected 390 SNPs covering 660 genes (Table S17). The three sites with the largest outlier values of are on chromosome 6 (95.45-95.7 Mb), chromosome 7 (0.35-0.45 Mb), and chromosome 10 (94.3-94.45 Mb). The genes in the regions where these three sites are located cover signal transduction and kinase (CAMK4 [51] ), lipid transport (COL4A3BP [50]) and hair growth regulation (FGF5 [52]) (Fig. 6B).

Fig. 6.

Selection signatures and functional enrichment associated with aridity adaptation in Chinese indigenous goats. A The UNEP aridity index map of China. B Genome-wide distribution of ratio between G-aird and G-wet. C Shared genes detected by ratio and . D Functional enrichment analysis of 162 genes identified using at two methods

The ratio method detected 404 significant SNPs, of which the three sites with the highest ratio were located on chromosome 6 (95.4-95.55 Mb), chromosome 10 (94.35-94.5 Mb), and chromosome 12 (51-51.1 Mb). The 809 annotated genes cover biological functions such as lipid transport (COL4A3BP [50], HMGCR [53]), nucleoplasmic transport (XPO4 [54]), metabolic regulation (BRSK2 [55]), and cell cycle regulation (LATS2 [56], MOB2 [57] ) (Fig. 6C; Table S18). In addition, COL4A3BP was not only identified as a drought adaptation-related locus in goats, but was also identified in high-elevation adaptation analysis.

Functional enrichment was performed on 162 candidate genes detected by both methods (Fig. 6D; Table S19). The results showed that they were significantly enriched in pathways such as “lipid metabolic process”, “positive regulation of cold-induced thermogenesis”, “potassium chloride symporter activity” and “Glycerophospholipid metabolism”. These results indicate that the enrichment of lipid metabolism and membrane transport-related functions may help goats cope with drought by maintaining cell membrane homeostasis and osmotic regulation balance (P < 0.05) (Fig. 6E; Table S20).

Genomic regions related to environmental temperature adaptation in goat populations

We used the method to detect the selective characteristics of goats in hot and cold environments. Under the first 1% threshold, 373 SNP sites were detected, covering 612 candidate genes. Among them, the three sites with the largest outlier values were located on chromosome 6 (95.45-95.6 Mb), chromosome 7 (0.35-0.45 Mb), and chromosome 22 (56.45-56.55 Mb). The candidate genes in the regions where these three sites are located cover signal transduction and kinase (CAMK4 [51]), neural development (EXOC2 [58] ) and hair growth regulation (FGF5 [49]) (Fig. 7A; Table S21).

Fig. 7.

Selection signatures and functional enrichment associated with temperature adaptation in Chinese indigenous goats. A Genome-wide distribution of values between G-hot and G-cold. B Genome-wide distribution of ratio between G-hot and G-cold. C Shared genes detected by ratio and . D Functional enrichment analysis of 74 genes identified using at two methods

Further testing of the two populations using the ratio method revealed that 653 candidate genes were annotated within the regions covered by the top 1% of 362 SNPs. Among them, the three loci with the highest ratios were located on chromosome 6 (69.45–70.80 Mb), chromosome 16 (79.05–79.15 Mb), and chromosome 18 (57-57.1 Mb). These candidate genes are involved in RNA transcription and processing regulation (SCAF1 [59], FIP1L1 [60], PRMT1 [61]), protein transport (IPO9 [62], TIMM17A [63]), protein hydrolysis (RNPEP [64]), and neurodevelopment regulation (PRR12 [65]) (Fig. 7B; Table S22). Furthermore, the CAMK4 gene appears not only in genomic regions associated with drought adaptation but also with temperature adaptation. Meanwhile, FGF5 was detected in genomic regions associated with temperature, drought, and high-elevation adaptation, suggesting it is a core candidate gene for goats to cope with multiple extreme environments.

Candidate genes obtained by the method (n = 579) and the ratio method (n = 538) were integrated, and a total of 74 shared candidate genes were screened (Fig. 7C; Table S23). Functional enrichment analysis showed that these genes were significantly enriched in pathways such as “antibacterial innate immune response”, “keratinization”, “extracellular region”, and “NOD-like receptor signaling pathway” (P < 0.05). (Fig. 7D; Table S24).

Discussion

Understanding how Chinese goats adapt to diverse and complex ecological environments is crucial for deepening our understanding of evolutionary biology and promoting the sustainable development of animal husbandry. This study, based on a comprehensive analysis of whole-genome resequencing data and high-resolution environmental variables, systematically explores the adaptive genetic basis of Chinese goats under different ecological gradients at the genomic level. The results indicate that the environmental adaptation of Chinese goats exhibits significant multi-gene regulatory characteristics, involving synergistic responses to multiple key ecological factors such as elevation, temperature, and drought. The clusters in different environments showed good consistency, accompanied by significant genomic differentiation patterns, which together indicate that local adaptation played a dominant role in shaping the genetic structure of Chinese goat populations [10, 66]. These findings further reveal the complex adaptation strategies developed by Chinese goats under long-term natural selection pressure, providing important evidence for elucidating the molecular mechanisms of their environmental adaptation.

Environmental factors such as high temperature and drought play an important role in the evolution of goats in China. In the PCA of environmental variables, the mean temperature (tavg) and water vapor pressure (vap) are highly correlated with PC1, indicating that heat and drought gradient together constitute the main environmental level affecting the ecological differentiation of Chinese goat populations [67]. According to the Köppen climate classification [30], the CL1 population is mainly distributed in tropical to subtropical regions, while the CL2 population is concentrated in temperate regions, which suggests that long-term environmental differences may have affected the genetic structure of the two. Chinese goats were introduced from the Fertile Crescent of West Asia around 6,000–7,000 years ago [3, 68]. During their eastward spread, they gradually differentiated along the north-south climate gradient of China, forming a differentiated adaptation pattern to the hot and humid environment in the south and the dry and cold environment in the north [69]. The CL1 and CL2 differentiation pattern observed in this study is the result of the synergistic effect of long-term geographical diffusion and climate selection, reflecting the dynamic evolution of the genetic structure of Chinese goat populations driven by the environment.

In the genetic structure analysis, some goat groups in CL1 (Banjiao Goat (BJG) [70], Chongming White Goat (CMG) [71] and CL2 (Ziwuling Black Goat (ZBG) [72] and Zhongwei Goat (ZWG) [73]) all showed genetic characteristics dominated by a single ancestral component, suggesting that their genetic background is relatively stable and the level of gene exchange is low. This characteristic is influenced by its geographically isolated environment (such as mountains or islands), its long-term relatively closed breeding pattern, and the continuous natural selection under specific ecological conditions (such as arid or semi-desert environments), which together shape its relatively stable population genetic structure. The MTG and MGG goats in the CL2 population showed a certain degree of mixed characteristics in genetic structure analysis, suggesting that they may be influenced by different genetic backgrounds. Meigu goat (MGG) was formed by crossbreeding two introduced breeds with local goat breeds and gradually developed through long-term natural selection and spontaneous breeding by the masses, reflecting a clear background of human intervention [11]. Based on its genetic structure characteristics, it is evident that both human introduction and the natural environment jointly promoted gene exchange and introgression across taxa. In the phylogenetic tree, some individuals of the Matou Goat (MTG) [74] and Meigu Goat (MGG) [75] did not strictly cluster in the CL2 branch, but were embedded in or close to the CL1 branch, reflecting the complexity and mixing of their genetic origins. The Treemix maximum likelihood population tree shows that most goat populations gradually differentiate along the main trunk, exhibiting a hierarchical structure consistent with the CL1 and CL2 grouping results. Some populations cluster in relatively close branches, indicating a close genetic distance. It is worth noting that the Youzhou dark goat (YDG) exhibits a significantly longer branch length, which is distinct from other groups, indicating that it may have undergone strong genetic drift and a relatively independent evolutionary process, which is consistent with its distribution in karst landform areas, long-term geographical isolation and limited gene flow [76, 77].

Gene-environment association analysis revealed the adaptive evolutionary mechanisms of goats in a variety of complex environments. The limited overlap between and LFMM detected genes may be mainly attributed to their different statistical frameworks. identifies loci with strong genetic differentiation among predefined population groups, whereas LFMM detects genotype–environment associations while accounting for latent population structure. Nevertheless, the permutation test showed that the observed overlap was significantly higher than random expectation, supporting the shared genes as a conservative candidate set for environmental adaptation. Our findings suggest that the adaptation of Chinese goats does not depend on isolated responses to single stressors, but rather on the coordinated regulation of multiple physiological systems. In terms of metabolism and energy balance, CEBPZ [78] and RXRA [79] play a role in energy storage and body temperature maintenance by regulating adipocyte differentiation and lipid deposition. RGS7 [40] and CACNG4 [80] participate in G protein signaling and neural pathways, enabling them to quickly sense environmental changes and precisely regulate the body’s stress response, maintain homeostasis, and thus enhance their ability to adapt to complex environmental stresses. Defense and homeostasis were supported by the enrichment of OGG1 and chemotaxis-related genes [81], as well as the role of SCPEP1 [82] in blood pressure regulation. In summary, these results indicate that goat adaptation is a systematic and multi-pathway process that integrates metabolic, immune, homeostatic, and behavioral mechanisms.

High-elevation goat populations exhibit significant selection signals in several key genes related to physiological regulation, reproduction, and environmental adaptation, reflecting their comprehensive adaptation strategies to extreme environments such as hypoxia, low temperature, and high ultraviolet radiation. At the basic physiological level, SKIDA1 [47] enhances the hematopoietic function of high-elevation goats by participating in the hematopoietic process, thereby improving blood oxygen transport capacity and alleviating hypoxia stress. PRLR [48], KITLG [83] is involved in reproductive and developmental processes, suggesting the importance of reproductive adaptation in high-elevation environments; COL4A3BP [50] is involved in lipid transport and metabolism, and may improve an individual’s ability to survive in a resource-scarce environment by optimizing energy storage and utilization. Overall, these genes work synergistically across multiple levels, from hematopoietic regulation and energy metabolism to reproductive development and coat regulation, to construct a systematic adaptation mechanism for high-elevation goats to extreme environments. This pattern is consistent with the multi-system co-evolutionary features observed in other high-elevation adapted species [84–86].

As a complex climate phenomenon, drought needs to be assessed by comprehensively considering multiple dimensions such as precipitation and potential evapotranspiration [6]. Although temperature is closely related to the degree of drought and is often included in the calculation of drought indices (such as the Global Standardized Precipitation Evapotranspiration Drought Index [87]), this study still focuses on the independent analysis of the drought adaptation mechanism of goats. The results showed that several key genes related to energy regulation, water maintenance and cellular homeostasis were identified, suggesting that drought-stricken goats cope with long-term water scarcity and high-temperature stress through multi-system synergistic regulation. Among them, CAMK4 [88], as an important signal transduction and kinase molecule, participates in regulating cellular stress response and metabolic balance; COL4A3BP [50] and HMGCR [89] participate in lipid transport and cholesterol synthesis to optimize lipid metabolism and energy reserves. Meanwhile, POLK [90] suggests that DNA damage caused by oxidative stress or environmental pressure in arid environments requires more efficient repair mechanisms to maintain genome stability. The role of FGF5 [52, 91]in regulating hair growth may be to enhance adaptability to arid environments by adjusting hair length and density, reducing evaporative heat loss or minimizing water loss. SEN1 [92] is upregulated in drought-resistant animals such as camels, and participates in antioxidant and cell protection processes, helping to alleviate oxidative damage caused by dehydration. Meanwhile, ZEB1 [93] further transmits external drought signals to the cellular response level by coordinating cell contraction, tissue remodeling, and osmotic pressure regulation, thereby promoting the occurrence and maintenance of the overall adaptive response. Overall, these genes work together at multiple levels, from signal transduction and lipid metabolism to DNA repair and cell cycle and phenotypic regulation, to construct a comprehensive adaptation strategy for drought-stricken goats to cope with water shortage and environmental stress.

In studies on extreme temperature adaptation, genes such as CAMK4 [51, 88], KIT [94], PRRG2 [95] and RRAS [96] are involved in cellular stress response and signal transduction processes, suggesting that the body can quickly sense temperature changes and regulate downstream physiological responses. As an important regulator of cell proliferation and differentiation, RAF1 [97] may enhance the body’s adaptability to extreme environments by regulating cell growth and repair processes. In addition, MKRN2 [98], LNX1 [99] and NOSIP [100] are involved in the ubiquitination regulation process, which helps maintain the dynamic balance between protein degradation and renewal under temperature stress. Regarding immunity and cell survival, IRF3 [101] is involved in the regulation of immune response, and BCL2L12 [102] is associated with the process of cell apoptosis, indicating that extreme temperatures may drive the body to enhance its immune capacity by inducing inflammatory responses or cell damage. In addition, CPT1C [103] and STARD4 [104] are involved in lipid metabolism and energy supply, suggesting that under extreme temperature conditions, the body needs to regulate lipid utilization to meet higher energy demands; LMOD1 [105] is involved in muscle contraction regulation, which may be related to maintaining the ability to produce heat in cold environments. Meanwhile, keratin family genes (KRT78, KRT79, KRT18 [106–108]) are often highly expressed in epithelial tissues such as skin and hair, maintaining the integrity of cell and tissue structures and buffering structural damage caused by low temperature freezing under extreme temperature conditions. Overall, these candidate genes work synergistically across multiple levels, including signal transduction, protein homeostasis, immune defense, and energy metabolism, to jointly construct a comprehensive adaptation mechanism for goats to extreme temperature environments.

Conclusions

This study systematically analyzed the selection signals of Chinese goats under different extreme environments (such as high elevation, hot, and arid), revealing that environmental pressures shape the adaptive evolutionary pattern of their genome through multidimensional selection. By integrating whole-genome resequencing data with high-resolution climate variables, we analyzed the association between environmental heterogeneity and genetic variation at the population scale, and identified a series of candidate genes related to energy metabolism, developmental regulation, immune response and morphological adaptation. Among them, genes such as CAMK4, COL4A3BP, CEBPZ and FGF5 showed selection signals in different environmental gradients, suggesting that they may play an important role in key adaptation processes such as regulating physiological homeostasis, signal transduction and energy metabolism. This study not only deepens our understanding of the evolutionary mechanism of environmental adaptation in Chinese goats, but also provides important theoretical basis for goat stress resistance breeding and germplasm resource protection. Against the backdrop of intensifying global climate change, exploring and utilizing the genetic diversity inherent in local breeds is of great significance for enhancing the environmental adaptability of livestock systems.

Authors’ contributions

Methodology, W.W, and K.C; Software, W.W, K.C and J.D; Visualization, W.W and K.C; Validation, L.Q and K.Y; Supervision, L.Q and W. L; Formal Analysis, K.Y, and J.L; Investigation, K.Y and J.D; Funding Acquisition, K.Y and W.L; writing—original draft preparation, W.W and K.Y; writing—review and editing, W.L and K.Y. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Science and Technology Innovation Program of Shanxi Agricultural University (2020BQ75) and Open Project Fund of Shanxi Key Laboratory of Livestock Genetic Resources Exploration and Biotech Breeding (2025).

Data availability

In this study, we analyzed 510 goats, of which the whole-genome resequencing data of 491 goats were obtained from the National Center for Biotechnology Information (NCBI) and the China National Center for Bioinformation (CNCB), while the remaining 19 goats [109]were sequenced by our research group. The detailed accession numbers of each individual sample are listed in Table S1.

Declarations

Ethics approval and consent to participate

This study was reviewed and approved by the Institutional Animal Care and Use Committee of Shanxi Agricultural University (approval number: SXAU-EAW-2022G.WX.0116211).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.