Runs of homozygosity and selection signals analysis reveals domestication traits and divergence in local domestic duck breeds

Abstract

Artificial selection can enhance desirable traits but may also lead to excessive homozygosity, negatively affecting genetic diversity. Monitoring inbreeding levels in poultry is crucial for maintaining the sustainable development of economic traits. This study investigates the genetic diversity of local domestic duck breeds including Weishan Partridge (WS), Matahu (MT), and Wendeng Black (WD) in Shandong province, with a focus on functional genes associated with economically important traits. We assessed the number and distribution of runs of homozygosity (ROH) and calculated the inbreeding coefficient (F_ROH) in three local duck breeds. The Weishan Partridge duck exhibited high genetic diversity, while the WD showed a high inbreeding level, likely due to intensive anthropogenic selection for meat production. Meanwhile, we compared the shared and unique genetic traits of the three breeds at the genomic level based on ROH Islands to obtain significant pathways and candidate genes associated with economically important traits, such as growth and development, lipid metabolism, and immune response. Comparative analysis of genomic selection signatures across breeds revealed distinct patterns of genetic differentiation: the MT breed exhibited selective pressures primarily targeting genes associated with metabolic regulation pathways, the WD breed showed evidence of selection for loci governing muscle development and myogenesis, while the WS breed demonstrated enrichment of selection signals in genomic regions linked to immune system function and anti-inflammatory response mechanisms. Altogether, this study provides valuable insights into the genetic diversity and differentiation of local duck breeds including MT, WS and WD in Shandong Province. The findings are crucial for developing more effective breeding strategies, which will aid in the conservation and optimal utilization of these indigenous breeds.

Introduction

Ducks provide abundant products such as eggs, meat and feathers for human life and are one of the poultry with important economic value. Local duck breeds are an important part of biodiversity and also have a profound impact on the development of agricultural production economy and traditional food culture (Yang et al., 2025). According to the 2024 edition of the National List of Livestock and Poultry Genetic Resource Varieties, there are a total of 66 local domestic duck breeds and their supporting lines in China, of which 42 are local domestic duck breeds (https://www.gov.cn). Local duck breeds are the result of long-term natural selection and artificial domestication, carrying unique selection imprints, showing strong survival and adaptability, and having advantages in disease resistance, roughage tolerance, egg production and egg quality (Feng et al., 2021). However, the conservation and development of most of the local domestic ducks are still in the primary stage, and the production performance and economic value are limited because they have not been systematically selected and bred. At the same time, the small population size and the backwardness of selective breeding methods make the local domestic duck breeds also face the risk of confused genetic background and reduced genetic diversity (Zhu et al., 2024). Therefore, it is particularly important to accurately assess the conservation effects of current local duck breeds to optimize breeding strategies.

In livestock and poultry breeding, the practice of mating within a limited population, especially involving related individuals, is common. This occurs because superior-performing individuals are selected more frequently than others, leading to increased inbreeding and potentially reduced genetic diversity and production performance. Inbred animals may also become more susceptible to specific diseases (Makanjuola et al., 2021). Assessing the level of inbreeding can therefore help to intervene in a timely manner to prevent the loss of genetic diversity in a population. One effective method for assessing inbreeding is the analysis of runs of homozygosity (ROH), which represent regions of the genome where loci are identical by descent, inherited from a common ancestor (Xie et al., 2019). While conventional inbreeding coefficients, based on genealogical information (F_PED), may be affected by inaccuracies or missing data, ROH-based inbreeding estimates (F_ROH) are based on actual genomic data, offering a more reliable measure of inbreeding (Gomez-Raya et al., 2015). Furthermore, the number and length of ROHs can provide insight into the population’s effective size and its ancestral history. Populations with smaller effective sizes tend to exhibit more ROHs, while larger populations typically show fewer ROHs (Liu et al., 2024). Long ROHs may indicate recent inbreeding, whereas shorter ROHs can suggest more distant common ancestry.

Natural selection and artificial selection are evolutionary mechanisms that operate through the differential retention of favorable genetic variation. Natural selection involves the adaptive evolution of populations in response to environmental pressures, favoring alleles that enhance fitness through improved survival and reproductive success. Artificial selection, in contrast, represents deliberate human-directed breeding programs that systematically select for specific phenotypic traits of economic, aesthetic, or functional value. Both processes require heritable genetic variation and result in changes to allele frequencies within populations, ultimately leading to phenotypic changes that reflect the selective pressures applied (Tian et al., 2023). Research has shown that regions of the genome with frequent ROH occurrences are often associated with genes under positive selection. The advancement of sequencing technologies has facilitated the use of ROH detection methods across various animal species, including pigs, chickens, cows and horses (Tian et al., 2023; Fang et al., 2021; Wirth et al., 2024). Additionally, selection signal analysis, which examines allele frequency shifts, linkage disequilibrium, and population differentiation, provides insights into the mechanisms of anthropogenic selective breeding. Methods such as XP-CLR, Fst, and XP-EHH are useful for identifying traits under selection, offering new perspectives on phenotypic variation and the search for genes associated with important traits (Qi et al., 2024).

Wendeng Black, Matahu, and Weishan Partridge are native duck breeds from Shandong province, all of which are egg-laying ducks. Over time, these breeds have undergone natural and artificial selection, resulting in distinct genetic characteristics. However, there is limited research on the genome-wide characterization and functional gene exploration of local duck breeds in Shandong Province. Despite having been bred for a long time, the conservation status of their genetic resources remains unclear, which is detrimental to genetic diversity conservation and sustainable development. To address this gap, we employed sequential pure fragment and selection signal analysis to identify key genes associated with production traits and explore the differential genomic characteristics of these local duck breeds. In this study, we evaluated their inbreeding levels based on F_ROH and detected ROH islands to identify candidate genes associated with economic traits, and also compared the genetic differences among the three of them by selection signal analysis. This study provides valuable insights into the selection and breeding of high-performance local breeds of ducks and contributes to the conservation of local germplasm resources.

Materials and methods

Sample collection information

The samples of Shandong local domestic duck species involved in this experiment were collected in the pre-laboratory, including 30 Wendeng black ducks, 30 Weishan Partridge ducks and 29 Matahu ducks (Table 1).

Table 1.

Test sample collection information.

Duck breeds	ID	Number	Station
Weishan Partridge	WS	30	Jining Weishan Xinhe Laying Duck Breeding Co.
Matahu	MT	29
Wendeng Black	WD	30	Weihai Qinghe Wendeng Black Duck Original Breeding Farm

Sequencing typing and quality control

The sequencing raw data involved in this experiment were quality controlled using FASTP software to remove reads with junctions, excessive N content, and low quality (Chen et al., 2018). The clean data was compared to the reference genome (Genome assembly ZJU1.0) of Anas platyrhynchos (mallard) by BWA software, and then BAM files were generated after sorting and de-duplication steps (Li and Durbin, 2009). The HaplotypeCaller module of GATK software was used to detect variants in multiple samples from the processed comparison files, and the detected variants were filtered by VariantFiltration module, and the detected SNPs were functionally annotated by ANNOVAR software (Li and Durbin, 2009; Wang et al., 2010). In order to reduce the error rate of SNP detection, SNPs were filtered: QD < 2.0, FS > 60.0, MQ < 40.0, SOR > 3.0; and then SNPs were strictly filtered: snp cluster filtering (no two snp within 5 bp); SNP filtering near indel (SNPs within 5 bp of indel were filtered out); adjacent SNP filtering (SNPs within 5 bp of indel were filtered out); and SNP filtering (SNPs within 5 bp of indel). INDEL filtering (the distance between two indels cannot be less than 10 bp); for loci with GQ (Genotype Quality) less than 20.0, the typing quality of the samples was labeled as lowGQ. The VCF file was further filtered to obtain high-quality SNPs using VCFTOOLS software with the parameters set to –minDP 4 –max-missing 0.3 –maf 0.05 (Danecek et al., 2011).

ROH testing and classification

The ROH assay was performed on the three populations using the -homozyg parameter of the PLINK software, with the following parameters set: (1) a sliding window size of 50 SNPs; (2) a minimum length of 100 kb for the ROH fragments; (3) the number of heterozygotes allowed for each sliding window was no more than 1; (4) the number of genotypes allowed to be missing in each sliding window was no more than 5; (5) the maximum interval between two consecutive SNPs in the ROH fragments is not more than 1000 bp; (6) the minimum density of SNPs in the ROH fragments is 1SNP/50 kb, and (7) the hit rate (sliding window threshold) of all scanning windows that contain SNPs must be at least 0.05 (Purcell et al., 2007). Statistics on the number and length of the ROH fragments and the distribution of the ROHs at the genome and chromosome levels were obtained for each sample data. distribution of ROH at the chromosome level. Meanwhile, in order to further understand the population history of local duck breeds, the obtained ROHs were categorized into six groups based on ROH length: <1Mb, 1∼2Mb, 2∼3mb, 3∼4Mb, 4∼5Mb and >5Mb. Define ROH fragments <1 MB as short ROH fragments, ROH fragments in the length range of 1–5Mb as medium ROH fragments, and ROH fragments >5 MB as long ROH fragments.

Calculation of population inbreeding coefficient based on ROH

The ROH-based inbreeding coefficient F_ROH is used to assess the degree of inbreeding by calculating the length of the ROH in the genotype, and F_ROH is calculated as $FROH=\sum \frac{LROH}{LAUTO}$. where L_ROH is the total length of ROH for each individual autosomal chromosome and L_AUTO is the total length of autosomal SNPs.

ROH island identification

ROH islands are regions of the genome that consist of long pure segments at high frequencies. In the analysis process, the sliding window method is often used to detect ROHs by calculating the frequency of each SNP involved in ROHs in an individual and calculating the proportion of SNP occurrences in ROHs. All detected ROHs were collated and analyzed, and ultimately the top 1 % of SNPs involved in composing ROHs were selected as the threshold for ROH islands. Genomic segments in which all SNPs exceeding this threshold are linked and occur in at least 30 % of individuals are called ROH islands. Here, we counted the genes involved in the ROH of each breed and performed genes functional enrichment analysis.

Selective signal analysis

In order to understand the germplasm resource differences among local domestic duck breeds in Shandong, this study analyzed the genomic data of any two of the three breeds for selection signal using three methods: Fst, Pi ratio and Dxy. The above analysis was done using PIXY software (Korunes and Samuk, 2021). During the analysis, the top 1 % of the genome was first screened for genomic regions subject to selection and the corresponding genes were annotated, and secondly, the overlapping results of any two of the three methods were defined as candidate genes and functionally enriched. The window size of all selection-clearing analyses was uniformly set to 10 kb.

Candidate gene functional annotation

Genes within the ROH islands were functionally annotated by Gene Ontology (GO) analysis and KyotoEncyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. A significant enrichment of the host pathway was indicated when p < 0.05.

Results

The ROH analysis in MT, WD, and WS

According to the statistical results, the total number of ROH observed in the MT, WD, and WS groups were 7369, 7867 and 6437, respectively (Fig. 1a). Correspondingly, the total length of ROH in these groups was 5,483,452 bp, 5,838,050 bp and 4,534,039 bp for MT, WD and WS respectively. Notably, chromosomes 17, 31, and 33 were excluded from the analysis due to their short length, which was insufficient to detect any ROH (Fig. 1b; d). Analysis revealed a positive correlation between chromosome length and ROH abundance, which can be attributed to several genomic mechanisms. Larger chromosomes provide greater physical space for ROH formation and accumulation due to their extended genomic territory. Additionally, longer chromosomes experience more recombination events per meiosis, creating increased opportunities for ROH generation when inbreeding occurs. Chromosomal regions with reduced recombination rates, particularly pericentromeric areas, exhibit enhanced ROH retention due to suppressed crossing-over activity. Conversely, smaller chromosomes demonstrate limited capacity for extensive ROH formation due to their restricted physical length. Furthermore, small chromosomes often harbor essential functional genes, and excessive ROH accumulation in these regions may impose deleterious fitness effects, resulting in their elimination through purifying selection.

Fig. 1.

Runs of Homozygosity (ROH) detection in three populations: (a) number of ROH of different length types; (b) distribution of number of ROH in different chromosomes; (c) distribution of number of different length types on chromosomes; (d) density of distribution of ROH at the genome chromosome level; and (e) proportion of short, medium and long ROH numbers.

The highest total number of ROH fragments was observed on chromosome 1, with 5080 fragments, followed by chromosome 2, with 3741 fragments and chromosome 13 with 13 fragments. The identified ROH fragments were categorized into six length groups: <1 Mb, 1-2 Mb, 2-3 Mb, 3-4 Mb, 4-5 Mb, and >5 Mb, with a 1 Mb interval threshold (Fig. 1c; e). More than 82 % of the ROH in all three populations were <1 Mb in length, totaling 17,763 fragments. Medium-length ROHs, ranging from 1 to 5 Mb, along with those >5 Mb, accounted for 17 % and 0.3 % of the total, numbering 3,843 and 59 fragments, respectively. These findings suggest that inbreeding within the local domestic duck breeds of Shandong is primarily a result of distant ancestral generations.

Overall, the total number of ROH detected on chromosomes decreased with the reduction in chromosome length, and the distribution of ROH across different length categories followed a similar pattern (Table 2). In general, a higher proportion of longer ROH fragments correlates with stronger inbreeding in recent generations, suggesting reduced genetic diversity. The percentage of ROH <1 Mb exceeded 80 % in all three populations (MT, 80.96 %; WD, 82.39 %; WS, 82.69 %), a phenomenon reflecting ancient inbreeding signals that may have originated from prolonged periods of small-scale reproduction or multiple bottlenecking times in the population history. In addition, the accumulation of recombination time cutting long ROH into short ROH could also lead to the dominance of short ROH. The proportion of long ROH (>5 Mb) was extremely low in the three populations, but the group differences were significant (MT, 0.204 %; WD, 0.394 %; WS, 0.249 %), and long ROH tended to represent recent inbreeding events, such as consanguineous mating in the last couple of generations, suggesting that there may have been more frequent recent mating in the WD population. WS, on the other hand, had the least long ROH, reflecting the effectiveness of its genetic diversity conservation.

Table 2.

Statistics of ROH type and number by length.

Breed	Type	Number	Percentage	Mean size (Mb)	Total length (Kb)	Genome coverage	Proximity factor (Froh)
MT	<1MB	5966	0.8096	0.5124	3057020	2.7138	2.7138
WD		6482	0.8239	0.5132	3326800	2.9532	2.9532
WS		5323	0.8269	0.5069	2698070	2.3951	2.3951
MT	1-2MB	1068	0.1449	1.3587	1451090	1.2882	1.2882
WD		1013	0.1288	1.3509	1368510	1.2148	1.2148
WS		889	0.1381	1.3511	1201180	1.0663	1.0663
MT	2-3MB	220	0.0299	2.3781	523171	0.4644	0.4644
WD		235	0.0299	2.3838	560186	0.4973	0.4973
WS		169	0.0263	2.3465	396566	0.3520	0.3520
MT	3-4MB	77	0.0104	3.4058	262249	0.2328	0.2328
WD		79	0.0100	3.4086	269280	0.2390	0.2390
WS		31	0.0048	3.4080	105648	0.0938	0.0938
MT	4-5MB	24	0.0033	4.4966	107919	0.0958	0.0958
WD		28	0.0036	4.6114	129120	0.1146	0.1146
WS		10	0.0016	4.3542	43542	0.0387	0.0387
MT	>5MB	15	0.0020	5.4669	82004	0.0728	0.0728
WD		31	0.0039	5.9403	184150	0.1635	0.1635
WS		16	0.0025	5.5641	89026	0.0790	0.0790

The individual inbreeding coefficients for the WS population ranged from 0.1120 to 0.1764, for the MT population from 0.1055 to 0.2074, and for the WD population ranged from 0.1367 to 0.2326 (Fig. 2). Notably, the inbreeding coefficients of the WD were higher than those of the MT and WS. These results suggest that WS has lower inbreeding coefficients and exhibit higher genetic diversity, whereas WD has higher reproduction coefficients, which may be related to being subjected to more stringent anthropogenic selection.

Fig. 2.

Violin plots of genomic inbreeding coefficients for the three populations.

ROH Island assay and candidate gene analysis

Anthropogenic selection pressures have significantly shaped the genomic landscape, resulting in distinct selective imprints across various populations. This phenomenon is particularly evident in the tendency of specific genomic regions to form consecutive pure segments. To identify the genomic regions associated with ROH in the three populations, we selected the TOP 1 % SNPs as candidate markers. The detection rates for these SNPs varied across populations, with the lowest detection rates observed at 56.67 %, 68.97 %, and 63.33 % for the WS, MT, and WD populations, respectively, and the highest rates at 93.33 %, 96.55 %, and 100 % (Fig. 3a). From this analysis, we identified 73, 60, and 81 ROH Islands in the WS, MT, and WD populations, respectively. A total of 316, 304, and 488 genes were annotated in these regions (Fig. 3b; Table S1). Notably, 196 genes were identified in at least two populations, and 55 genes were common to all three populations.

Fig. 3.

Statistics of ROH island assays and functionally annotated genes: (a) manhattan plot on ROH assays; (b) statistics of annotated genes for the three populations of ROH islands.

In the subsequent analysis, GO and KEGG functional enrichment analyses were conducted to explore the genetic characteristics unique to each of the three populations (WS, MT and WD) based on the genes identified within the ROH islands. Significant GO terms and KEGG pathways were selected using a threshold of P.adjust < 0.05 (Table S2). For the WS, MT, and WD populations, 314, 244, and 323 GO terms, and 61, 44, and 25 KEGG pathways, respectively, were identified. Notably, 44 GO terms and 8 KEGG pathways were common across all three populations (WS, MT and WD). Here, we screened the top 20 of GO project and KEGG signaling pathway in ascending order according to the size of P.adjust value for analysis to discuss the genetic characteristics of the three breeds, respectively (Fig. 4).

Fig. 4.

Functional enrichment analysis of ROH island genes: (a) GO item enrichment; (b)KEGG signaling pathway.

The GO terms for MT focused on lipid metabolism (especially phospholipid catabolism, phospholipase activity), organelle transport (Golgi), metal chaperones, synaptic modulation, and bone mineralization, which may indicate that MT has a strong selective pressure in lipid metabolism, cellular transport, and neural or muscle regulation. The GO terms of WD were clearly related to muscle structure and function. In addition, the enrichment of items related to salivary secretion, digestive system, and fluid secretion suggests that there are selective adaptations in muscle motor and secretory functions, which may be related to feeding, digesting, or exercising in WD. In contrast, the GO terms for WS focus on lipid metabolism (similar to MT), purine metabolism, angiotensin receptors and their related processes (blood pressure regulation and vascular homeostasis), which may be related to cardiovascular system adaptation or energy utilization.

For MT, addiction-related pathways such as nicotine, morphine, and cocaine may be associated with neuromodulation, and phospholipid metabolism and arachidonic acid metabolism as well as thyroid hormone pathways may be associated with lipid and energy metabolism. WD is mainly enriched in pathways such as pancreatic secretion, salivary secretion, and insulin secretion, which are associated with digestive and secretory functions. In addition, the cGMP-PKG signaling pathway and calcium signaling pathway, may be involved in muscle contraction and secretion regulation, which in turn affects WD motility, digestion, and egg-laying processes. The enrichment of lysosomal, cancer pathway, VEGF and NF-κB pathway signaling pathways may suggest an advantage of WS in anti-inflammatory response and immune regulation.

Differences in the enrichment of GO terms and KEGG signaling pathways reflect the differences in physiological functions of the three duck breeds, such as different energy metabolism requirements, locomotor abilities, or adaptation mechanisms to the environment. Here, we counted GO terms and KEGG signaling pathways associated with economic traits for potential functional genes(Table S3).

We further compiled the information related to ROH islands shared by local duck breeds in Shandong Province, and identified genes associated with inflammatory responses that were annotated multiple times, such as PLA2G4A, PTGS2, SOCS1 and CLEC16A. In addition, TENM3, GRIA2, CPB1 and PDC may be associated with nerve signaling and energy metabolism processes (Table 3).

Table 3.

Shared fragments of ROH islands and related genes in three local duck breeds.

Chr	Starting position	Ending position	Length (bp)	Related gene name
Chr1	66639986	67053049	413063	LMO3
Chr3	89317419	89390024	72605	ADGRB3
Chr4	21299519	21489163	189644	TENM3
Chr4	21727593	21843746	116153	TENM3
Chr4	29847478	30089412	241934	GLRB, GRIA2
Chr8	14161552	14604334	442782	PLA2G4A, PTGS2
Chr8	14617147	14694254	77107	PDC
Chr8	14694445	15100882	406437	HMCN1, ODR4, PDC, PRG4, TPR
Chr9	26091097	26495654	404557	AGTR1, CPB1, PIGZ
Chr15	597500	901989	304489	CLEC16A, LITAF, RMI2, SOCS1

Analysis of differences in genetic characteristics among three breeds

To investigate genetic differentiation among the three varieties, each of the three varieties was compared to the other two by selection signal analysis to screen for selection-driven genetic differences. We used three methods of Fst, Pi ratio and Dxy to take the intersection and finally obtained 1405, 1483 and 1389 genes in MT, WD and WS breeds, respectively, which were detected by at least two of the three methods (Table S4). In addition, 317, 459 and 283 genes were detected by all three methods in MT, WD and WS breeds(Fig. 5a). We subsequently analyzed these genes for GO entries and KEGG signaling pathways (Table S5). The first 20 GO terms and KEGG pathways were selected based on the ascending order of adjusted P-values (P-adjusted) (Fig. 5b; c). Similarly, we counted GO terms and KEGG signaling pathways associated with economic traits for potential functional genes (Table S6). Collectively, the distinct genetic profiles of the MT, WD, and WS populations in terms of muscle function, metabolic regulation, immune response, reproductive performance, and calcium metabolism underscore their potential applications in waterfowl farming.

Fig. 5.

Results of selection signal analysis of three populations of individual varieties in Shandong Province with two other varieties: (a) Wayne plots of the number of genes obtained by different analysis methods; (b) GO enrichment results; (c) KEGG enrichment results. MT and WD populations exhibited significant genetic divergence in pathways associated with muscle cell proliferation and differentiation.

We found that MT was significantly enriched in neurotransmitter receptor activity (e.g., acetylcholine receptor, G protein-coupled receptor) and synapse-associated GO entries, which may reflect that MT was selected for neural signaling efficiency. In addition, salivary secretion somatic secretion and tissue secretion suggest adaptations in the MT digestive system or exocrine function that may be related to feed conversion rates or environmental adaptations. GO terms in WD are mostly related to muscle structure and contractile function, such as cellular components and regulatory processes like troponin complex, myofibrils, and myonodules, which may have a close relationship with WD muscle development or athletic ability. The enrichment of ribosomal subunits, on the other hand, suggests enhanced protein synthesis to support the metabolic demands of muscle telltale growth. Leukocyte apoptosis regulation and interleukin-1 production, on the other hand, are associated with disease resistance or inflammatory response. The active presence of GO terms such as those related to the cytoskeleton, such as the actin skeleton, stress fibers, adhesion patches, and cellular junctions, as well as those related to digestive secretion, such as salivary secretion, and digestive system processes, further emphasizes the optimization of tissue structural stability and digestive efficiency, reflecting a certain degree of mechanical stress and feed-specific adaptations in WS. Cell adhesion, on the other hand, is associated with tissue barrier function, reflecting the dominance of WS in pathogen defense processes.

KEGG signaling pathways associated with cardiovascular development and signaling regulation, such as Arrhythmogenic right ventricular cardiomyopathy, Cardiac muscle contraction, and Cholinergic synapse, were significantly enriched in MT, which was associated with ducks' exercise performance and stress tolerance were closely related. In WD, Type II diabetes mellitush and growth hormone synthesis, secretion and action affect energy metabolism and storage, which in turn regulate the rapid development and weight gain of ducks. In contrast, GnRH secretion can modulate the reproductive axis or the reproductive cycle, which in turn improves egg production. The core selective pressures of WS, on the other hand, focus on immune defenses and digestive adaptations, with significant enrichment of cell adhesion and calcium signaling pathways suggesting an advantage in disease resistance and feed conversion rates.

Discussion

Comparative analysis of inbreeding levels across three breeds

The ROH represent a powerful genomic approach for quantifying population genetic diversity, reconstructing demographic history, and identifying genomic signatures of domestication (Hewett et al., 2023). ROH-based analyses enable precise estimation of genome-wide inbreeding coefficients (F_ROH), providing robust metrics for assessing population-level inbreeding intensity and genetic diversity (Di Gregorio et al., 2023).

Our comparative analysis revealed significantly elevated inbreeding levels in WD individuals relative to MT and WS breeds. This pattern likely reflects the combined effects of reduced effective population size, intensive selective breeding practices, and potentially closed breeding management systems within the WD population. These findings corroborate previous linkage disequilibrium analyses that similarly demonstrated elevated genomic inbreeding within this breed. MT exhibited intermediate inbreeding coefficients, while WS displayed the lowest inbreeding levels, consistent with maintained genetic diversity.

Notably, the WD population exhibited substantial heterogeneity in individual inbreeding coefficients, including several individuals with extreme inbreeding values, indicative of uneven breeding practices and variable genetic backgrounds within the population. Conversely, MT and WS populations demonstrated more uniform distributions of inbreeding coefficients, suggesting more consistent breeding management strategies. Furthermore, these findings imply that the WD population has undergone a higher intensity of artificial selection, potentially linked to its higher meat production traits. The combination of factors such as extensive inbreeding, geographic isolation, small population size, and strict selective breeding likely contributed to the elevated allele frequencies and reduced genetic diversity observed in the WD population (Liu et al., 2022). To mitigate the risks associated with inbreeding, we recommend avoiding prolonged centralized selection and the continued use of the same genetic lines. Enhancing genetic diversity could be achieved through strategies such as expanding the population size, introducing new genotypes, and incorporating extragenic genes during the breeding process. These measures could help prevent the genetic diseases associated with high inbreeding levels while maintaining the population's productivity and health (Chaudhari et al., 2023). In addition, it is crucial to emphasize the importance of continuous monitoring and periodic evaluation of the population's genetic status. Such practices are essential for developing rational conservation strategies that promote long-term genetic sustainability.

Functional analysis of ROH islands and genomic selection signatures

We conducted functional enrichment analysis of genes located within ROH islands across the three duck breeds and performed comparative genomic scans to identify selection signatures associated with economically important traits and breed-specific genetic adaptations. Notably, ROH island analysis and selection signature detection yielded concordant results, revealing distinct patterns of genetic differentiation that reflect breed-specific selection pressures. Our findings demonstrate that genetic differentiation among the three breeds corresponds to divergent selection for distinct functional categories: MT breeds exhibited enrichment for genes involved in metabolic regulation and neurological pathways, consistent with the high energetic demands of sustained egg production. WD breeds showed selection signatures predominantly associated with muscle development and endocrine regulation, supporting their utility for meat production systems. In contrast, WS breeds demonstrated enrichment for immune defense mechanisms, stress response pathways, and anti-inflammatory processes, reflecting adaptation to diverse environmental challenges. Performance data corroborate these genomic findings: WD individuals exhibited superior body weight at 8 weeks of age compared to MT and WS breeds, along with enhanced carcass characteristics and meat quality parameters, supporting their potential development as specialized broiler lines (Ren et al., 2024). Convergent selection patterns were evident in ROH islands shared across all three breeds, which were significantly enriched for genes regulating inflammatory responses, behavioral traits, and metabolic processes. This functional convergence likely reflects common selective pressures for improved production efficiency, stress tolerance, and metabolic optimization—core objectives in contemporary duck breeding programs (Huang et al., 2025).

Enhanced growth rates, metabolic homeostasis, and extended reproductive cycles represent fundamental objectives in modern poultry breeding programs. Meat production efficiency and quality are determined by two primary biological factors: skeletal muscle development patterns including fiber type composition, and lipid deposition characteristics encompassing intramuscular fat content and phospholipid profiles (Cho et al., 2019). Several candidate genes contribute to skeletal muscle development through distinct regulatory mechanisms. TNNT3, VCL, PRKCB, MYO16, AGTR1, and FGF14 collectively regulate muscle contractile apparatus assembly, cytoskeletal organization, myocyte proliferation and differentiation, and intracellular signaling cascades essential for muscle fiber development (van de Locht et al., 2021; Zhang et al., 2022; Guo et al., 2020; Kengyel et al., 2015; Che et al., 2025; Hu et al., 2013). Genes involved in adipocyte biology significantly influence meat quality characteristics. PLCG2, PLA2G4A, and PLA2G10 modulate intramuscular adipogenesis through complex signaling networks involving PPARγ, diacylglycerol (DAG), and protein kinase C (PKC) pathways, thereby regulating adipocyte differentiation, metabolic activity, and lipid accumulation (Ma et al., 2024; Xiao et al., 2021; Huang et al., 2024). Genes regulating digestive function directly impact growth performance through enhanced nutrient utilization. CHRM3 encodes a muscarinic acetylcholine receptor mediating parasympathetic control of digestive secretions, gastrointestinal motility, and appetite regulation via brain-gut axis signaling, thereby promoting feeding behavior (Zhang et al., 2025a). Additionally, ADCY8, ATP2B1, KCNMA1, and CPB1 coordinate the secretion of pancreatic enzymes, gastric acid, and bile, while regulating intestinal digestive rhythms to optimize nutrient digestibility, protein absorption, and overall growth efficiency (Kong et al., 2019; Nie et al., 2020; Honzlová et al., 2022; Whelan et al., 2024).

A robust immune system is essential for minimizing pathogen-induced growth disruption and maintaining optimal health status, ultimately contributing to enhanced body weight gain and reproductive performance. Key immunoregulatory genes including CIITA, PTPRC, TRAF2, PTGS2, and LITAF function as critical modulators of immune recognition, adaptive immune responses, antimicrobial defense, and inflammatory resolution. These genes collectively regulate pathogen clearance kinetics, inflammatory tissue damage extent, and recovery processes in livestock species (Chen et al., 2021; Al Barashdi et al., 2021; Martín-Vázquez et al., 2023; Forlani et al., 2023; Albini et al., 2025). Additional immune-modulatory genes including PLCG2, PIK3CB, FOXP1, and ZEB1 regulate intracellular signal transduction cascades and metabolic homeostasis, thereby indirectly influencing immune response magnitude and duration (Ye and Zeng, 2022; Lu et al., 2022; Chen et al., 2023; 2025). The NRXN and SLITRK gene families, along with NCAM2, primarily mediate neuroimmune communication networks, including stress-induced immune responses and neurogenic inflammatory processes (Marchese et al., 2021; Puranik and Song, 2024; Zhang et al., 2025b). Reproductive performance is regulated by genes affecting hypothalamic-pituitary-gonadal axis function. DIO3, EGFR, HRAS, and CREB3L4 modulate hormone secretion through activation of GnRH, FSH, and cAMP-mediated signaling pathways, controlling gonadal development, sexual maturation timing, folliculogenesis, and granulosa cell differentiation during oogenesis (Artini et al., 2017; Bai et al., 2020; Pan et al., 2022; Yuan et al., 2022). Metabolic support for reproductive function is provided by PIK3CB, MAP2K4, and PRKCB, which regulate energy metabolism and lipogenesis through PI3K/Akt, MAPK, and insulin signaling pathways to maintain reproductive energy homeostasis (Zhu et al., 2022; Ran et al., 2023; Muhammad et al., 2025).Eggshell quality is determined by calcium homeostasis genes including ATP2B1, RYR2, and CACNA1C, which regulate calcium ion transport within the oviduct and control calcium carbonate deposition and eggshell mechanical properties (Gloux et al., 2019; Medina et al., 2015; Li et al., 2021).

The majority of the genes identified in this analysis are associated with key physiological traits, including immune health, metabolic balance, and reproductive performance. These traits are crucial for the economic value of livestock and poultry, impacting growth performance, reproductive efficiency, and disease resistance.

Conclusions

This study characterized genome-wide ROH distribution patterns across three indigenous duck breeds from Shandong Province, quantified population-level inbreeding coefficients, and identified genomic signatures of selection through ROH island and selection signal analyses. Our findings revealed distinct inbreeding hierarchies among breeds: WD exhibited significantly elevated inbreeding levels, MT demonstrated intermediate values, while WS maintained the lowest inbreeding coefficients. These results underscore the critical importance of implementing strategic mating systems and genetic management practices to preserve population genetic diversity in domestic duck breeding programs. Functional enrichment analyses of ROH islands and selection signatures revealed breed-specific patterns of genomic selection corresponding to distinct biological pathways. MT breeds showed enrichment for genes involved in neurological signaling and metabolic regulation, consistent with selection for enhanced egg production traits. WD breeds demonstrated selection signatures associated with muscle development and locomotor function, reflecting their optimization for meat production characteristics. WS breeds exhibited enrichment for immune defense mechanisms and anti-inflammatory pathways, indicating selection for disease resistance and environmental adaptation. Through these complementary genomic approaches, we identified numerous candidate genes associated with economically important traits including meat quality, immune function, stress tolerance, and digestive efficiency. These findings provide comprehensive insights into the genetic architecture underlying production traits in Shandong's indigenous duck populations and establish a foundation for implementing genomic-assisted breeding strategies in future genetic improvement programs.

Ethics approval and consent to participate

The current study and the use of all ducks were approved by the Institutional Animal Care and Use Committee of Liaocheng University, Liaocheng, China (AP2024061217).

Consent for publication

Not applicable.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its additional files, or in the following public repositories. Data has been submitted to the public database under the following accession numbers: whole genome re-sequence data [PRJNA1145904].

Funding

This research was funded by the Key R&D Program of Shandong Province, China (2024LZGC020 and 2024LZGC002), the Program of Fujian Key Laboratory of Animal Genetics and Breeding (FJXQKFJJ2023), the Shandong Province Livestock and Poultry Genetic Resources Preservation Farm and Gene Bank Protection Project (K22LC0701 and K23LC1301), the National Student Innovation and Entrepreneurship Program (202410447021), and the University-level Student Innovation and Entrepreneurship Project (CXCY2024302).

CRediT authorship contribution statement

Pengwei Ren: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization, Writing – review & editing. Yadi Jing: Resources, Writing – review & editing. Liu Yang: Resources, Writing – review & editing. Muhammad Zahoor Khan: Writing – review & editing. Meixia Zhang: Resources, Writing – review & editing. Xiang Liu: Writing – review & editing. Weiqing Ma: Resources, Writing – review & editing. Zhaoyan Ding: Resources, Writing – review & editing. Xiangfan Li: Resources, Writing – review & editing. Chao Qi: Resources, Writing – review & editing. Zhansheng Liu: Resources, Writing – review & editing. Shuer Zhang: Resources, Writing – review & editing. Zhiming Zhu: Funding acquisition, Writing – review & editing. Nenzhu Zheng: Resources, Writing – review & editing. Mingxia Zhu: Conceptualization, Supervision, Project administration, Funding acquisition, Writing – review & editing.

Disclosures

All authors declare that they have no competing interests.