Conservation of biodiversity is a critical goal in the management of livestock genetic resources. Italy serves as an exemplary model due to its rich biodiversity among domesticated species, fostered by its diverse historical, environmental, climatic, and agricultural traditions (Senczuk et al., 2020). Goats, in particular, represent a significant component of this biodiversity, with over 30 indigenous breeds thriving in varied climates and farming systems, many of which operate under low-input agricultural practices (Cortellari et al., 2021a).

The Alpine goat breeds, extensively farmed in northern Italy, particularly in the Alpine regions, are subject to a range of farming practices, from intensive indoor systems to extensive grazing. These breeds are highly valued for their dairy production, known for both quantity and quality of their milk. The Camosciata delle Alpi (CAM) is a dairy goat breed originating from the Alpine regions of Italy, Switzerland, France, and Austria. Well-adapted to mountainous environments, it is characterized by its resilience, strong physique, and excellent milk production (Pegolo et al., 2025). The milk of the CAM breed is rich in fat and protein, making it particularly suitable for high-quality cheese production, a key factor in its economic relevance (Agradi et al., 2022). Similarly, the Saanen (SAA) breed, originally from Switzerland, is recognized as one of the world’s most productive dairy goats. It is renowned for its high milk yield, adaptability to various climates, and widespread use in commercial dairy production (McSweeney and McNamara, 2022). However, in recent years, crossbreeding with Dutch white genetics – intended to further enhance milk production – has introduced significant challenges. The introduction of Dutch genetics into Italian Saanen populations is supported by practical evidence from breeders. Many Italian breeders report to breed their flocks using semen from Dutch Saanen lines to improve milk yield and other productive traits. Therefore, assessing the genetic diversity of these breeds is crucial to understand the impact of such crossbreeding practices on their resilience, productivity, and overall health characteristics.

In Italy, the Associazione Nazionale della Pastorizia (AssoNaPa, 2025) is the breed society responsible for genetic programs in goats and sheep, in accordance with the EU “Animal Breeding Regulation” (Regulation (EU) 2016/1012, 2016), including the management of herdbooks for the Saanen and Camosciata delle Alpi goat breeds. In Lombardy, intensive (and some semi-intensive) indoor systems are common in lowland and valley areas, while semi-extensive and extensive modalities persist in upland and marginal terrain, producing differing management-driven selection pressures across herds. National selection tools, such as estimated breeding values (EBVs), and the gradual introduction of genomic monitoring, are available through AssoNaPa. While AssoNaPa routinely estimates EBVs for animals recorded in the herdbook with genealogical information, the reliability of pedigree data provided by farmers (AssoNaPa, 2025), a fundamental input for genetic evaluation, is often incomplete or inaccurate. In many herds, in fact, a substantial proportion of animals lack fully recorded or correct genealogical information, particularly when natural mating occurs in groups with multiple bucks. As a result, selection decisions are often based primarily on farmer-recorded phenotypes, rather than on EBVs, which limits the efficiency of genetic improvement programs at both national and regional levels. In parallel, national monitoring and conservation programmes remain essential to track genomic diversity and safeguard smaller local breeds. These nationwide efforts provide context for interpreting herd-level genetic variability and inbreeding.

Inbreeding, in fact is a key aspect of genetic diversity and a major concern in livestock breeding, as it increases homozygosity and may reduce fitness. The inbreeding coefficient (F), traditionally estimated from pedigree data, is now often derived from Runs of Homozygosity (ROH), which provide a more accurate measure of autozygosity (Szpiech et al., 2013; Peripolli et al., 2017). ROH are continuous homozygous segments within the genome (Gibson et al., 2006) that arise through inbreeding, genetic drift, or selection, and their length distribution offers insights into demographic history: short segments typically reflect ancient inbreeding, whereas long ones indicate more recent events (Keller et al., 2011). Thus, ROH analyses provide information on both the genetic health of populations and on regions of the genome shaped by selection (Purfield et al., 2012; Kim et al., 2013; Dixit et al., 2020). Beyond individual-level inbreeding estimates, the distribution of ROH across a population can highlight shared regions of homozygosity, known as ROH islands (Nothnagel et al., 2010). These genomic hotspots, where positive selection has increased the frequency of specific haplotypes, have been used to identify selection signatures in several livestock species (Zhang et al., 2015; Bertolini et al., 2018; Grilz-Seger et al., 2018; Mastrangelo et al., 2018; Peripolli et al., 2018; Xie et al., 2019). While such processes can favor beneficial mutations, they may also increase deleterious variants through genetic hitchhiking (Szpiech et al., 2013; Makino et al., 2018). Assessing ROH patterns in Alpine (CAM and SAA) goats can therefore provide valuable insights into how breeding practices and management strategies have shaped their genetic structure as already done in other species (Punturiero et al., 2023).

The aims of this study were: i) to characterize ROH patterns in both Alpine goats to understand how different breeding and management practices impact genetic diversity, and to identify genomic regions under putative selection; ii) to assess the impact of recent crossbreeding and reproductive strategies on the genomic integrity and structure of these populations. Details will then be reported at the individual farm level, providing both a general overview of the two breeds and highlighting the specific selective choice of each farm.

Alpine dairy goats such as CAM and SAA hold a pivotal role in sustaining genetic diversity and local dairy traditions in northern Italy. These breeds are extensively reared not only in mountain and pre-Alpine areas, where they contribute to landscape preservation and rural livelihoods, but also in lowland commercial farms, where their high productivity and adaptability support large farm milk production systems (Battaglini, 2007; Cortellari et al., 2021b).

By combining PCA, admixture analyses, ROH profiling, and inbreeding‐coefficient estimates, we obtained a comprehensive picture of the genomic structure of CAM and SAA goats across the ten Lombardy farms. These complementary approaches reveal how herd-specific management strategies and historical breeding choices have shaped the genomic landscape of both breeds, influencing levels of homozygosity, inbreeding, and differentiation among herds. Herds enrolled in this study operate independently, with breeding and mating plans based on the goals and reproductive management practices of each farmer. The clear separation between CAM and SAA in the combined PCA reflects breed-level genetic differentiation, in line with previous findings in these goat breeds (Brito et al., 2017). Within CAM, most herds formed tight clusters, indicative of relatively homogeneous breeding lines. However, Herd_06 showed a broader dispersion, likely reflecting the use of bucks from diverse genetic origins, which may have introduced greater variability. In contrast, all SAA herds exhibited greater genetic dispersion, possibly due to the introgression of Dutch genetic lines introduced to enhance milk production, as reported by breeders at the time of herd book registration (Supplementary Table S5). ADMIXTURE analysis with K = 4 revealed a complex ancestral structure within and across herds, despite the dataset including only two breeds (CAM and SAA). This suggests the presence of multiple genetic backgrounds within each breed, likely due to the use of bucks from different national or commercial lines, as well as herd-specific breeding strategies (Punturiero et al., 2023). In CAM herds, distinct ancestry components point to limited exchange of genetic material among farms and selection focused on different traits or production goals. These differences indicate that CAM herds do not share a uniform genetic profile but instead reflect herd-specific selection histories or the use of bucks from distinct genetic sources (Waineina et al., 2020). The consistent ancestral signal observed in most herds suggests a shared genetic foundation, likely due to the use of similar or closely related breeding lines. Regarding the SAA, Herd_08 stands out with a slightly more admixed profile respect Herd_09 and Herd_10, showing larger contributions from ancestors 1, 3, and 4, may be because the presence of CAM goats in past generations. Similar dynamics have been reported in other European dairy goat populations, where intensive selection, the importation of genetic material of selected bucks, and constrained within-herd mating practices have contributed to fine-scale population structuring (Peripolli et al., 2017; Mastrangelo et al., 2021).

When we compared the number of shared bucks across FA herds (using the official animal identification code recorded in the herdbook provided by farmers), we observed a very limited level of common FA bucks use (Supplementary Table S5). The same buck was used in a maximum of three herds, and typically sired between two and six half-sibling daughters per farm. This restricted sharing pattern suggests that male selection in these herds is largely independent, with each farm relying on its own breeding strategy and limited external male input. This low level of buck sharing likely contributes to the variability observed in the admixture profiles of herds using instrumental insemination. Unlike the two SAA herds, which displayed consistent ancestral signals, the CAM herds showed more heterogeneous patterns. The observed differences in admixture can be explained by the use of distinct and often unrelated sires, resulting in herd-specific genetic backgrounds. Independent mating strategies and limited gene flow between herds have likely shaped these unique genetic structures, as reflected in the varying degrees of admixture identified across farms.

Inbreeding estimates and ROH.

The moderate overall F_ROH (0.060 in CAM; 0.041 in SAA) and negative mean F_HOM (-0.023 in CAM; -0.017 in SAA) (Supplementary Table S2 and Supplementary Table S3) indicate that, despite some recent inbreeding, most herds is keen to maintain genetic variability varying at best the male used in reproduction, similar to other well−managed dairy goat populations (Luigi-Sierra et al., 2022; Vostry et al., 2024). The high correlation observed between F_ROH and F_HOM (Figure 6B) suggests that homozygosity is not randomly distributed across the genome, but rather tends to cluster within ROH.

The genome-wide ROH identified in this study, revealed a predominance of short segments (<2 Mb), which are typically indicative of ancient inbreeding events or past population bottlenecks, where increased homozygosity results from remote common ancestors. Over generations, recombination fragments these regions, leading to shorter homozygous tracts (Keller et al., 2011; Peripolli et al., 2017). Interestingly, ROH segments between 8–16 Mb were observed only in a few individuals. This pattern may reflect occasional recent inbreeding events, which could occur unintentionally in large-scale herds where the management of mating groups to control inbreeding becomes more complex. While small herds often apply well-structured breeding strategies, larger operations may face challenges due to the lack of mandatory paternity testing and the potential for misassigned sires across mating groups. In such cases, the true pedigree structure can become blurred, leading to undetected related matings and, consequently, longer ROH in some animals. The near absence of ROH >16 Mb, however, suggests that very close kin matings (e.g., parent-offspring) remain rare.

The distribution of ROH observed in Alpine goats in this study aligns with findings in other goat breeds, such as Egyptian and various Italian breeds, where a predominance of short ROH segments has been reported, suggesting historical selection and inbreeding events. Recent studies on Alpine and Mediterranean goat breeds have further highlighted breed-specific ROH patterns, reflecting the influence of breeding practices and environmental adaptations (Manunza et al., 2023; Pegolo et al., 2025). Similar patterns of ROH length distributions, reflecting both ancient and recent inbreeding, have also been reported in other livestock species such as cattle and sheep, highlighting the general influence of herd management, mating strategies, and demographic history on genomic homozygosity (Strillacci et al., 2020; Benedetti del Rio et al., 2025). The integration of ROH length distribution with the analysis of ROH islands adds further resolution to these patterns. While short ROHs point to ancient demographic events (Doekes et al., 2019), ROH islands reveal signatures of both shared and herd-specific selective pressures. Several genomic regions are common to both CAM and SAA, consistent with their shared ancestry and exposure to similar selection regimes, whereas others appear breed-specific, reflecting divergent breeding goals and management choices. Moreover, some ROH islands are consistently shared across multiple herds (Table 4), which may indicate the diffusion of particular haplotypes through the widespread use of popular sires or parallel selection schemes. Conversely, herd-specific ROH islands highlight unique management histories or localized selection pressures. Taken together, these findings demonstrate how both historical demographic processes and herd-level breeding dynamics have shaped the genomic landscape of Alpine dairy goats in Lombardy.

Common ROH islands – CAM and SAA.

Regarding ROH islands, when grouping herds by breed, we identified three shared homozygous regions: on CHR8 (74 Mbp), CHR12 (49 Mbp), and CHR14 (16 Mbp). These regions harbor 19 genes, of which several appear in the STRING/Cytoscape network (Supplementary Figure S4), highlighting functional connections among genes related to stress adaptation (ZMYM2, PARP4), metabolic homeostasis (AQP3, ATP12A), and fertility (RNF17, PSPC1). Notably, many of these genes show convergent selection signals in small ruminants. For example, ZMYM2, PARP4, PSPC1, MPHOSPH8, ATP12A, and RNF17 have been found in ROH islands in Italian Garfagnina goats and Mediterranean breeds selected for resilience and productivity (Dadousis et al., 2021; Serranito et al., 2021). The gene RNF17, crucial for spermatogenesis and gamete interactions, was also identified in Egyptian Barki goats, strengthening its role in reproductive efficiency and environmental resilience (Sallam et al., 2023). The aquaporin genes, AQP3 and AQP7, are essential for water, glycerol, and urea transport, enabling thermoregulation and heat-stress adaptation. Additionally, AQP3 has been linked to reduced reproductive performance in goats under repeated estrus synchronization (Sun et al., 2023). In addition, the AQP3, AQP7, UBE2R2, NFX1, DCAF12, and UBAP2 genes appear within heterozygosity-rich or selected regions across sheep and goat breeds (Mészárosová et al., 2022). Recent studies in livestock, including goats, have characterized heterozygosity-rich regions (HRRs) that overlap with ROH islands. If any of our ROH islands (for example, on CHR8 at 74 Mbp) (Chessari et al., 2024) overlaps with a known HRR, this would indicate that the region is homozygous in certain populations but highly variable in others. This pattern is a hallmark of balancing selection, where functionally important variants (e.g., related to immunity, stress adaptation, fertility) are maintained as both fixed and polymorphic across different lineages.

The VPS13B gene emerged as one of the most consistently detected across analyses: it was present in both CAM and SAM groups and found in 9 out of 10 herds examined. VPS13B has functional relevance supported by associations in both cattle and goats. In Bos taurus, VPS13B resides within a QTL on CHR14 linked to female fertility and milk production traits (Liu et al., 2018). In addition, VPS13B has been identified as under selection in Mediterranean and Chinese goats and is linked to leg morphology, a trait frequently correlated with fertility and milk yield (Amiri Ghanatsaman et al., 2023).

Breed specific ROH_islands – CAM.

Among the genes unique to CAM herds, several exhibit strong cross-species associations with economically important traits. In the domain of stress resilience and immune function, AMFR was identified in a dairy cattle GWAS as a candidate for heat-stress response and milk fatty acid composition, while ATG16L2, an autophagy-related gene, has been linked to immune resilience and mastitis recovery in Holsteins (Welderufael et al., 2018). Additional genes involved in disease resistance and immune modulation include ARHGEF17, has been associated with disease resistance in cattle (Ghoreishifar et al., 2020) and linked to gastrointestinal nematode resilience through IgA regulation in sheep (Atlija et al., 2016). A similar immunological role has also been proposed for the RELT and FCHSD2 genes.

In terms of growth and skeletal development, ARAP1 is located within QTL regions relevant to viability and meat productivity in sheep hybrids (Zlobin et al., 2023), while MATN3 variants are robustly tied to skeletal integrity and stature in cattle (Lopdell and Littlejohn, 2018).

The CAPNS2 gene, which encodes calpain-2, plays a significant role in follicular development and has been specifically implicated in goat reproductive physiology, as shown in a study comparing gene expression in the ovaries of polytocous and monotocous dairy goats (An et al., 2012).

Breed specific ROH_islands – SAA.

Among the genes unique to SAA herds, several have well-characterized roles in functional and production traits across livestock. A key finding is the association of APTX with the marbling score in Hanwoo cattle: co-expression network analysis highlighted APTX as a hub gene correlated with intramuscular fat content, supporting its potential role in meat quality (Lim et al., 2013). In dairy cattle, ATP6V0D1 was identified via integrated transcriptomic–proteomic analysis in Holstein mammary glands, suggesting its involvement in ATP production in the mammary tissue (Dai et al., 2018).

Multiple genes also contribute to environmental adaptation, especially in heat and thermal stress response. DNAJA1, an HSP40 chaperone, was highly expressed in sheep mammary tissue and linked to milk synthesis, while also being repeatedly associated with heat-stress response in cattle (Freitas et al., 2021). Similarly, CRYL1 emerged from ROH analyses in Chinese sheep and Garfagnina goats adapted to diverse climates, pointing to a role in environmental resilience (Abied et al., 2020; Dadousis et al., 2021).

Genes involved in immune and resilience traits also surfaced prominently in SAA-specific ROH. GJA3, GJB2, GJB6, and MAB21L1 were highlighted in Ugandan goat breeds under selection for adaptability and in the Boer breed, GJB2 and GJA3 are indicated as candidate genes associated with production traits, such as body size and growth (Onzima et al., 2018). The genomic region harbouring GJA3, GJB2, and GJB6 also represents a selection signature previously identified in the Garfagnina goat breed.

Additional genes previously identified in goats include UBAP1, which is associated with a QTL for cannon bone circumference (QTL ID 314136), suggesting a role in skeletal development, and ZDHHC1, which overlaps with QTLs for somatic cell score (QTL IDs 296969 and 296970), indicating its potential involvement in udder health and mastitis resistance in caprine production.

This study provides a snapshot of the genetic landscape of Alpine dairy goats (CAM and SAA) raised under commercial and semi-extensive systems in Lombardy, using SNP chip genotypes from a large number of individuals. By integrating genome-wide SNP data with ROH-based inbreeding and population structure analyses, we reveal insights of direct relevance for herd management and conservation. Despite historical selection pressures, overall genetic diversity remains within acceptable ranges, particularly in CAM, reflecting effective management strategies that preserve variability even without formal pedigree control. However, marked differences emerge across herds, shaped by specific reproductive histories and decisions. In SAA, the introduction of Dutch genetics increased heterozygosity and ROH variability, supporting productivity in the short term but potentially diluting original lines and threatening long-term resilience. ROH islands shared across or specific to breeds pointed to candidate genes for stress adaptation, metabolism, fertility, and immune response, traits crucial for sustainable production. This work highlights the need to integrate genomic tools into herd monitoring.

In conclusion, balancing productivity and genetic diversity is feasible through data-informed decisions. Genotyping all herd animals provides actionable information to improve reproductive efficiency, preserve genetic resources, and ensure the long-term sustainability of Alpine goat farming in Lombardy.