We randomly selected 18 healthy Jianshui yellow–brown ducks (YB), including 7 males and 11 females, at the Jianshui yellow and brown duck breeding farm in Jianshui County, Yunnan Province. During the collection process, 5 mL of blood samples were taken from the wing veins of these ducks using vacuum blood collection tubes with EDTA anticoagulant. Ice packs were cryopreserved and transported back to the laboratory and stored in a −20°C refrigerator for subsequent extraction of whole gene DNA. In addition, 6 Jianshui yellow–brown ducks (3 males and 3 females) were randomly selected for slaughter, and 6 liver, 6 sebum, and 3 breast muscle tissue samples were collected, immediately place the tissue into a pre-prepared cryopreservation tube containing RNA preservation solution (the RNA preservation solution is RNA solid Stable Preservation Solution, Wuhan Servicebio Company), place it in liquid nitrogen, take it back to the laboratory and store it in a −80°C ultra-low temperature refrigerator for subsequent extraction of total tissue RNA. For detailed information on the collected samples of Jianshui yellow–brown ducks, please refer to Supplementary Table S1. All experimental procedures adhered to the regulations of the Administration of Laboratory Animal Affairs and were approved by the Ethics Committee of Yunnan Agricultural University (No. 202103035).

The Tiangen Total RNA Extraction Kit was used to extract total RNA from the liver, sebum and breast muscle tissue according to the manufacturer’s instructions. The quality-controlled samples were sent to Berry Genomics (Beijing, China) for sequencing, which will be performed on the Illumina novaseq 6,000 sequencing platform using paired-end sequencing (Paired-end sequencing, measuring 150 bp at each end, and the length of each read is 300 bp) with 10 G per sample (i.e., 1 billion base pairs per sample by transcriptome sequencing to ensure sufficient data for accurate gene expression analysis). The liver, sebum and breast muscle transcriptomic data of mallard and Pekin ducks used in this study were downloaded from public databases (NCBI and BIG Data Center, Bioproject ID: PRJNA645648, BIG accession codes PRJCA001307), transcriptome sequencing data downloaded from public databases also use paired-end sequencing, with 150 bp at each end and a read length of 300 bp. The same tissues from Jianshui yellow–brown ducks, mallards, and Pekin ducks were compared to identify differentially expressed genes (DEGs). The raw data underwent filtering and quality control using FASTP software with default parameters. The resulting clean data were aligned to the Pekin duck reference genes using HISAT2 (21) software (version: CAU_duck1.0). The parameters for mapping sequenced reads back to the reference genome using HISAT2 are as follows: --rna-strandness --new-summary -x genome −1 read1.fq.gz −2 reads2.fq.gz -S sample.sam. We set the parameter “--rna-strandness” mainly because we adopted the strand-specific library construction method when we sequenced the transcriptome of Jianshui yellow–brown duck tissues. In this way, we were able to obtain the orientation information of the RNA fragments to estimate the gene expression level more accurately. At the same time, we used the “-x” parameter to specify the reference gene of the selected species. The aim of this study was to look for differences between species by performing transcriptome analyses of the same tissue from different species. In order to control a single variable as much as possible and reduce the influence of sex factors on differentially expressed genes, we first removed the chromosomal information related to the sex of the species on the reference genome, and only aligned the reads back to the autosomes on the reference genome of the species. Expression quantification was performed on the alignment results using the “featurecounts” command of SUBREAD (22) software to obtain the raw expression matrix. Using the TPM (Transcripts Per Million) approach to normalize the raw expression matrix of transcriptomic data. TPM correction of raw counts prior to correlation analysis of samples is required in order to eliminate variability due to differences in gene length and sequencing depth. This ensures that comparisons of gene expression between samples are accurate and fair. Such standardization is essential for subsequent sample correlation analysis. Differential expression analysis was conducted on the raw expression matrix using DESEQ2 (23). We will use the raw expression matrix generated by featurecounts as the input file for DESEQ2. In this process, we set DESEQ2’s comparison grouping information, i.e., grouping according to varieties. After that DESEQ2 will calculate key information such as log₂|Foldchange|, value of p and padj for each gene. Significant DEGs were identified based on the criteria of |log2 Fold Change| > 1.5 and corrected p-values (P.adj) less than 0.001. The correction for false-positive detection due to multiple comparison tests was performed using the Benjamini–Hochberg (BH) method. Screening for differentially expressed genes (DEGs) by adjusted p-values reduces the rate of false positives and enhances the reliability of the study results. The number of DEGs in each population was then counted, and volcano plots were generated using R.

Joint analysis of selection signal and transcriptome

To further narrow down the candidate gene list and increase the robustness of the results, a joint analysis of selection signals and transcriptome data was performed. First, we screened the candidate genes related to the target trait (i.e., meat performance) from the selection signal results, and intersected the candidate genes from the selection signal with the differentially expressed genes obtained from the transcriptome analysis, and the intersected portion of the results was the result of the joint analysis.

Eighteen Jianshui yellow–brown ducks underwent resequencing, and the resequencing data of 92 ducks from nine different varieties were downloaded from the NCBI database, totaling 110 individuals from 10 populations. The quality of the sequencing data was assessed, and the 18 Jianshui yellow–brown ducks exhibited a base population quality with Q20 and Q30 values above 90%, normal CG distribution, an average sample alignment rate of 98.22%, an average sequencing depth of 18.94×, and an average sample coverage rate of 92.37% (Supplementary Table S3). The Q20 and Q30 values for the 92 resequencing data from the NCBI database also exceeded 90%, with an average alignment rate of 97.69%, an average sequencing depth of 8.14×, and an average coverage rate of 90.66% (Supplementary Table S4). Thus, the sequencing data demonstrated reliable quality and met the requirements for resequencing analysis. After the detection and filtering of population SNPs, a total of 19,795,912 high-quality SNPs were obtained from 110 individuals. The subsequent analysis focused on population structure, population genetic diversity, and population selection based on these population SNPs.

To examine the presence of outlier samples and inter-population genetic structure and relationships, PCA, phylogenetic tree construction, and population structure analysis were performed using whole-genome SNPs from 110 individuals representing 10 populations (Pekin, Cherry Valley Pekin, mallard, spot-billed, Fenghua, Shanma, Shaoxing, Jinding, Gaoyou, and Jianshui yellow–brown ducks). PCA results demonstrated that individuals within the same population tended to cluster together (Figure 1A). The 10 populations could be categorized into four groups: Gaoyou duck, Jinding duck, Shanma duck, and Shaoxing duck formed one group; mallard, spot-billed duck, and Fenghua duck formed another group; Pekin duck and Cherry Valley Pekin duck clustered together; and Jianshui yellow–brown duck formed a distinct individual population. The results from the NJ evolutionary tree (Figure 1B) were consistent with the PCA results, with the Jianshui yellow–brown duck population forming a separate cluster. Structure analysis indicated that the cross-validation error rate was minimized when K = 4 (Supplementary Figure S1). Figure 1C depicts the structure results when K = 2–4. When K = 2, the 10 populations were divided into two groups: one comprising domesticated varieties and the other consisting of wild varieties. Notably, the Fenghua duck and mallard populations showed signs of hybridization. When K = 3, Pekin and Cherry Valley Pekin ducks were separated from the domesticated varieties, resulting in three populations: the shelduck population (Shaoxing ducks, Gaoyou ducks, Jinding ducks, Shanma ducks, and Jianshui yellow–brown ducks), white-feathered meat-type duck population (Pekin ducks and Cherry Valley Pekin ducks), and wild duck population (mallard, spot-billed ducks, and Fenghua ducks). However, when K = 4, Jianshui yellow–brown ducks were further separated from the shelduck population. The results of genetic differentiation indicated that, among the 10 populations, the highest level of genetic differentiation was observed between Jinding ducks and Gaoyou ducks (Fst = 0.217), whereas the lowest level was observed between mallard and spot-billed ducks (Fst = 0.056; Supplementary Table S5). The Jianshui yellow–brown duck population exhibited the greatest degree of differentiation from the Gaoyou duck population (Fst = 0.209) and the smallest degree of differentiation from the Fenghua duck population (Fst = 0.139).

Nucleotide diversity values are indicative of the genetic diversity within a population. Supplementary Figure S2 illustrates the genetic diversity levels of the 10 populations. Mallard exhibited the highest genetic diversity, while Gaoyou ducks displayed the lowest. The two wild duck populations showed relatively rich genetic diversity. The genetic diversity of the Jianshui yellow–brown duck population was lower than that of mallard, spot-billed ducks, Shaoxing ducks, and Fenghua ducks but higher than that of Pekin ducks, Shanma ducks, Jinding ducks, Cherry Valley Pekin ducks, and Gaoyou ducks. Consequently, Jianshui yellow–brown ducks ranked fifth among the 10 populations in terms of genetic diversity.

In descending order, the values of genetic diversity were as follows: mallard, spot-billed ducks, Shaoxing ducks, Fenghua ducks, Jianshui yellow–brown ducks, Pekin ducks, Shanma ducks, Jinding ducks, Cherry Valley Pekin ducks, Gaoyou ducks. Two wild duck populations showed relatively rich genetic diversity. Among the eight domesticated populations, Jianshui yellow–brown ducks ranked third in terms of genetic diversity, after Fenghua ducks and Shaoxing ducks.

Meat performance is a crucial economic trait in the domestication of ducks. To investigate the selection of functional genes related to meat traits in varieties under different degrees of selection, the selection signals of Jianshui yellow–brown ducks, which undergo low selection, and Pekin ducks, which experience high selection, were analyzed with wild mallard as the control. Fst (population fixation index) and π (nucleotide diversity) were calculated for the whole genome regions of Jianshui yellow–brown duck, Pekin duck and mallards using the sliding window method with VCFTOOLS software. The window size and step size used in the calculation were in accordance with the parameters set in the section “2.2.3 Population selection analysis.” Based on the above method, the obtained Fst values were converted into Z-scores, and log₂θπ values were calculated for the compared populations. Figure 2 shows the results of the selection signals for Jianshui yellow–brown ducks and mallard. In performing the selection signal analysis, we screened genomic regions with ZFst and θπ values ranked in the top 1% as candidate regions that might be subject to selection. Specifically, we screened out regions in the genome of the Jianshui yellow–brown duck population that might be subjected to selection by setting a threshold of ZFst value >3.42 and log2(π_Mallard/π_{Jianshui yellow–brown duck}) ≥ −0.55. BEDTOOLS software was applied to merge the overlapping candidate regions, and in the results we identified 62 potential candidate genomic regions. After detailed annotation of genes within these regions, we identified a total of 136 genes that were subject to selection in Jianshui yellow–brown ducks.

To identify the selected genes in the Pekin duck population (Pekin duck vs. mallard), by calculating the thresholds for the ZFst and log₂θπ values, we considered the regions with ZFst values >3.26 and log2(π_Mallard/π_{Pekin duck}) ≥ −0.37 as the regions that might be subject to selective action in the Pekin duck population (Figure 3), 90 selected genes were annotated within 63 candidate regions for Pekin ducks. Pekin ducks are globally recognized for their meat characteristics, while Jianshui yellow–brown ducks are a meat-egg-type variety with a lower degree of selective breeding. To compare the genomic differences between these two varieties, selection signal analysis was conducted on Jianshui yellow–brown ducks and Pekin ducks to identify selected genes related to meat performance in the Jianshui yellow–brown ducks population and evaluate the selective breeding of Jianshui yellow–brown ducks for meat traits. Through gene function query and Tajima’s D calculation, 6 genes associated with growth and development and lipid metabolism – ELOVL2 (ELOVL Fatty Acid Elongase 2), ELOVL3 (ELOVL Fatty Acid Elongase 3), GDF10 (Growth Differentiation Factor 10), VSTM2A (V-Set And Transmembrane Domain Containing 2A), PHOSPHO1 (Phosphocholine Phosphatase 1), and IGF2BP1 (Insulin Like Growth Factor 2 MRNA Binding Protein 1) – were subjected to significant selection in the genome of Jianshui yellow–brown ducks (Supplementary Figures S3–S8). Furthermore, compared with Pekin and mallard, 6 genes related to muscle growth and fat deposition – GDF10, IGF2BP1, SCD (Stearoyl-CoA Desaturase), MAP3K20 (Mitogen-Activated Protein Kinase Kinase Kinase 20), TGFB3 (Transforming Growth Factor Beta 3), and WNT8B (Wnt Family Member 8B) – were identified (Supplementary Figures S9–S14). Compared with the Pekin duck population, 21 genes subjected to significant selection were observed in the Jianshui yellow–brown duck population (Supplementary Figure S15). However, based on functional annotation of these genes and comparison with the previously screened genes related to meat traits, none of these 21 genes were identified as candidate genes associated with muscle development and fat deposition.

Transcriptome sequencing technology facilitates the detection of mRNA expression in the same tissue across different duck varieties. This allows for the annotation and enrichment analysis of the functions of Differently Expressed genes (DEGs), which facilitates the identification of important functional genes related to characteristic traits in various varieties. In this study, transcriptome sequencing analysis was conducted on the liver, sebum and breast muscle tissues of Jianshui yellow–brown ducks and mallards. Additionally, DEGs were screened in the liver, sebum and breast muscle tissues between Pekin ducks and mallards. After screening and quality control of the raw data of transcriptome sequencing, and then comparing the high-quality reads back to the reference genome of ducks (CAU_duck1.0: https://ftp.ensembl.org/pub/release-110/fasta/anas_platyrhynchos_platyrhynchos/dna/Anas_platyrhynchos_platyrhynchos.CAU_duck1.0.dna.toplevel.fa.gz), the results showed that the transcriptome sequencing results of the tissue samples from the Jianshui yellow and brown duck population were of good quality, with an average Q20, Q30, and comparison rate of 92.61, 97.11, and 78.92%, respectively (Supplementary Table S6). The transcriptome sequencing results of Pekin duck and mallard duck downloaded from public databases also showed a high quality level, with average Q20, Q30 and alignment rates of 94.17, 97.48, and 81.01%, respectively (Supplementary Table S7). After quantification of gene expression, sample correlation analysis was performed on the normalized data using the TPM correction method. Subsequently, sequencing quality and sample data quality were preliminarily examined through analyses of sequencing base quality values, alignment rates, and sample correlation. Sample correlation analysis was performed with six sets of transcriptome data obtained from the same tissue among different duck varieties. The results revealed the presence of abnormal data (Supplementary Figures S16–S17). To improve the credibility of the results, sample correlation analysis was repeated after removing the abnormal data, and the results of this analysis suggested that the intra-population correlation was higher than the inter-population correlation, and samples from the same variety could be clustered together (Figures 4, 5). Samples with high confidence were then used to screen for DEGs, using threshold values of |log2FoldChange| > 1.5 and Padj < 0.001. According to the presentation in Supplementary Figures S18A–C, the transcriptomic data of Jianshui yellow–brown ducks and mallards showed that these two breeds exhibited different gene expression profiles in liver, sebum and breast muscle tissues. Specifically, there were 464 differentially expressed genes in liver tissues (251 up-regulated and 213 down-regulated), 1,435 differentially expressed genes in sebum tissues (958 up-regulated and 477 down-regulated), and 1,290 differentially expressed genes in breast muscle tissues (439 up-regulated and 851 down-regulated genes). After another transcriptome analysis of Pekin and mallard duck tissues (Supplementary Figures S18D–F), we identified 2,050 differentially expressed genes in liver tissues (532 genes were up-regulated and 1,518 genes were down-regulated), 473 differentially expressed genes in subcutaneous adipose tissues (including 295 genes up-regulated and 178 genes down-regulated), and 475 differentially expressed genes were screened out of breast muscle tissues (185 genes were up-regulated and 290 genes were down-regulated), differentially expressed genes between groups can be viewed in Supplementary Tables S8–S13.

Genotypic differences between populations within the region under selection may trigger changes in the expression of the gene. Multi-omics data were utilized to identify reliable artificial selection loci and functional genes associated with economically important traits. Five functional genes (ELOVL2, ELOVL3, GDF10, VSTM2A, and IGF2BP1) that may affect meat production and quality in Jianshui yellow–brown ducks were identified from genome-wide signal selection analysis. Transcriptome analysis of DEGs revealed significant differences in PHOSPHO1 gene expression in the breast muscle tissues of Jianshui yellow–brown ducks and mallards (Figure 6). In addition, the expression levels of ELOVL3 and GDF10 genes also showed significant differences in sebum tissue (Figures 7, 8). To identify the functional genes related to muscle growth and fat deposition traits in Pekin ducks, and provide references for the selective breeding of meat-type duck varieties, genomic selection signal detection was performed on Pekin ducks and mallards. This analysis led to the identification of six functional genes related to meat traits (GDF10, IGF2BP1, SCD, MAP3K20, TGFB3, and WNT8B). Data on DEGs indicated significant differences in TGFB3 gene expression in the liver tissue and SCD gene expression in both liver and sebum tissues (Figures 9, 10).

To comprehensively and accurately assess the germplasm resources of Jianshui yellow–brown ducks, resequencing was performed on 18 Jianshui yellow–brown ducks. The resequencing data from the ducks in the NCBI database were also used for population structure analysis. PCA was performed based on genome-wide SNPs. The results revealed that the Jianshui yellow–brown duck population was clustered into a single category, and individuals of the same species could be clustered together without any outlier samples in the PCA plot. In the evolutionary tree, 110 individuals from 10 populations were categorized into four groups. Among them, Shaoxing ducks, Gaoyou ducks, Shanma ducks, and Jinding ducks were clustered into one category, and all four groups were shelduck varieties, originating from East China. Thus, they were implicated to have a closer genetic relationship than the other six populations. Mallards, spot-billed ducks, and Fenghua ducks were clustered into a single category. Although differences in body size and appearance existed between mallards and spot-billed ducks, the degree of genetic differentiation between the two varieties was small. Therefore, no reproductive isolation occurred after mating, and separating the two populations in the evolutionary tree was challenging (5, 24). In the relationship tree, Fenghua ducks and mallards showed a closer genetic relationship. Fenghua ducks are directly domesticated from mallard ducks. Because of the short period of domestication, Fenghua ducks still retain some of the habits and characteristics of mallards (4, 25). Two white-feathered varieties, namely, Pekin duck and Cherry Valley Pekin duck, were clustered into a single category owing to their close genetic relationship. The Cherry Valley Pekin duck is a duck variety known for its high-quality meat, and this variety was improved and bred on the Chinese native Pekin duck by the British Cherry Valley Company. Therefore, the Cherry Valley Pekin duck and the Pekin duck exhibited a close genetic relationship (3, 26). In addition, the Jianshui yellow–brown duck population was clustered into a separate category, a finding similar to those of the distribution of the Jianshui yellow–brown duck population in PCA. Compared with PCA and evolutionary tree analysis, structure analysis can be used to determine the reasonableness of grouping and gene exchange among different varieties. In structure analysis, when the cross-validation error rate was minimum (i.e., K = 4), 10 populations were categorized into four groups. Jianshui yellow–brown ducks were separated from the shelduck population, forming a distinct population. This finding validated the results from PCA and evolutionary tree analysis, highlighting the differences in molecular genetic characteristics between Jianshui yellow–brown ducks and wild ducks, white-feathered meat-type ducks, and shelducks in East China. We hypothesized that the Jianshui yellow–brown ducks have developed a unique molecular genetic mechanism owing to the influence of natural and artificial selection.

Jianshui yellow–brown ducks, known for their tender meat, have become the key ingredient for the specialty dish “Qujiang Roast Duck” in Jianshui County, Yunnan Province, China. Identifying candidate genes that affect the traits of these varieties is crucial for understanding the genetic basis of their characteristics and conserving germplasm resources. In the selective breeding process, selection signal analysis was conducted on Jianshui yellow–brown ducks and mallard to explore the functional genes influencing meat production and meat quality traits. Several candidate genes related to muscle quality were identified, including members of the elongase of the very-long-chain fatty acid (ELOVL) gene family, such as ELOVL2 and ELOVL3, both of which were subjected to significant selection. The ELOVL family, as initiation and rate-limiting enzymes that catalyze fatty acid synthesis, plays an important role in fatty acid chain elongation (27). VSTM2A plays an essential role in regulating preadipocyte differentiation. VSTM2A overexpression increased the efficiency of fat synthesis, whereas VSTM2A knockdown inhibited fat formation (28, 29). Growth-related genes, including GDF10, IGF2BP1, and PHOSPHO1 were also identified. Gene polymorphisms in GDF10, also known as bone morphogenetic protein-3b (BMP-3b), have shown a significant correlation with some growth traits in cattle and have a possible involvement in skeletal muscle formation and development. The GDF10 gene is also involved in lipogenesis and metabolism (30). Based on the results of GWAS analysis, Zhou et al. identified IGF2BP1 as a candidate gene affecting the growth and carcass of domestic ducks (20). Moreover, the study by Wang et al. on domestic chickens, which was based on the pan-genomic data of domestic ckickens, showed that high expression of the IGF2BP1 gene influences the body size of ckickens (31). The PHOSPHO1 is a member of the phosphate dehydrogenase family and encodes an enzyme belonging to the acid phosphatase class that catalyzes the hydrolysis of inorganic phosphate esters. This gene is well known for its key role in bone growth and bone mineralization in organisms (32, 33). Recently, experimental results from Peng et al. have also shown that the PHOSPHO1 gene also plays an important role in the regulation of myogenic cell differentiation (34). These findings contribute to understanding the genetic mechanisms underlying the meat production and quality characteristics of Jianshui yellow–brown ducks.

Based on the analysis of population structure, it was found that the Jianshui yellow–brown duck and Pekin duck, two domesticated varieties, have significantly diverged from the wild mallard through domestication and selective breeding. In the selection signal analysis between the Jianshui yellow–brown duck and mallard, candidate genes related to meat traits of the Jianshui yellow–brown duck were identified. However, as a meat-egg-type local variety, the meat performance of Jianshui yellow–brown ducks still requires further improvement. Therefore, the meat-type Pekin duck was chosen as the research object and the mallard as the control, and the selection signaling was used to identify candidate genes affecting muscle growth and fat deposition traits in Pekin duck. Six functional genes (GDF10, IGF2BP1, SCD, MAP3K20, TGFB3, and WNT8B) related to meat traits were identified. Among them, SCD, a key enzyme required for the synthesis of monounsaturated fatty acids, plays a crucial role in the regulation of lipogenesis. Mice with SCD1 gene knockout have increased energy expenditure and fat consumption and decreased amounts of fat deposition (35). Bioinformatics approaches have been used for studying meat quality in livestock, and the results have suggested that the SCD gene is an important functional gene affecting lipogenesis and fat deposition (36, 37). TGFB3, a candidate gene affecting variation in rib count in pigs, also plays a role in avian lipogenesis and lipometabolism. Based on the GWAS-derived SNP association results affecting abdominal fat percentage in chickens, TGFB3 was identified as a candidate functional gene affecting fat deposition (38). Wang et al. identified TGFB3 as a target gene of miR-122, which is a regulator of adipose metabolism (39, 40). MAP3K20 and WNT8B, belonging to the MAPK and Wnt signaling pathways (41, 42), respectively, are involved in cellular signaling pathways crucial for muscle growth and development. These genes may have a regulatory role in the muscle growth of Pekin ducks.

Jianshui yellow–brown duck, as a meat-egg-type local variety, has undergone limited selective breeding for meat traits. In contrast, the Pekin duck is a meat-type variety undergoing extensive selective breeding. To explore the genetic differences between these two varieties and assess the selection potential of Jianshui yellow–brown duck for meat performance, a selection signal analysis was performed. Only 21 genes were subjected to selection in the Jianshui yellow–brown duck population compared with those in the Pekin duck. When using mallard as a control, 136 and 90 genes were selected in the Jianshui yellow–brown duck and Pekin duck populations, respectively. It is worth noting that during the functional annotation of these 21 genes (YB vs. PK) selected from the Jianshui yellow–brown duck population, we also found an interesting gene, ITGA4 (Integrin Subunit Alpha 4), which encodes the α4 integrin, which has been shown to play an important role in the regulation of inflammation and immune responses (43, 44). Jianshui yellow–brown ducks have always lived at relatively high altitudes and, due to their relatively short history of domestication, they exhibit strong environmental adaptations and disease resistance. Compared with the Pekin duck population, we observed that the ITGA4 was strongly selected for in the genome, and hypothesized that the ITGA4 plays an important role in both environmental adaptation and disease resistance in Jianshui yellow–brown ducks.

Among these selected genes, 6 genes were associated with muscle growth and fat deposition in Jianshui yellow–brown ducks, and 6 genes were related to meat traits in Pekin ducks. However, these genes were not detected in the Jianshui yellow–brown duck population when comparing selection signals between Jianshui yellow–brown ducks and Pekin ducks. Although the Jianshui yellow–brown duck variety has been bred for its meat characteristics for some time, the degree of its improvement in meat traits was not adequately high because of the short duration and low intensity of selective breeding. Therefore, further selective breeding efforts are needed to enhance meat yield and quality in Jianshui yellow–brown ducks.

The selection signaling method has gained attention for identifying important functional genes related to economic traits in ducks (6, 45–47). By screening genes within regions subject to selection, which exhibit reduced polymorphism, higher inter-population differentiation, and changes in allele frequency, the accuracy of results can be improved (48). Combining multiple selection signal assays enhances the reliability of the findings. Additionally, the availability of upgraded reference genomes facilitates gene annotation within selected regions (49–51). Transcription of DNA into mRNA is a key event for proteins to exert their functions. However, gene expression during the transcription process is spatiotemporal specific, wherein different growth periods and feeding conditions affect the degree of gene expression (52, 53). Therefore, joint analysis of selection signals and transcriptome data provides a preliminary understanding of the effect of allelic changes in the selected region on gene expression and facilitates the identification of reliable selection sites and functional genes (1, 20). In this study, ELOVL3, GDF10, and PHOSPHO1 were identified as important functional genes affecting muscle growth and fat deposition in Jianshui yellow–brown ducks through a joint analysis of selection signals and transcriptome data. Moreover, the selection signal analysis between Jianshui yellow–brown ducks and Pekin ducks revealed that genes related to meat traits in the Jianshui yellow–brown duck population were not strongly selected, indicating the need for further emphasis on selective breeding for meat traits. The TGFB3 and SCD genes identified in Pekin ducks, associated with meat performance, may serve as potential markers for the molecular breeding of meat-type Jianshui yellow–brown ducks in the future.