A total of 24 Tunisian Barbarine sheep blood samples (BARB dataset) were obtained from selected farms that were part of a Cooperation project (2018–2022) led by Tamat (Italian NGO https://tamat.org) (Figure 1). The animals were selected among the families of breeders based on phenotypic features and information provided by farmers to ensure the collection of unrelated individuals. DNA extraction from blood was performed using the Illustra blood genomic Prep Mini Spin kit (GE Healthcare, Little Chalfont, United Kingdom). The samples were genotyped using the Illumina OvineSNP50 BeadChip, resulting in a total of 53,516 SNPs (Illumina, Inc., San Diego, CA, United States).

PLINK ver. 1.9 (Chang et al., 2015) was used to filter data and perform the quality control. Chromosomal coordinates of SNPs were referred to the OAR4.0 genome assembly (NCBI accession number GCA_000298735.2). Markers on unknown and sexual chromosomes were removed. The quality parameters applied were the following: a minor allele frequency ≥0.05, a genotype call rate for a SNP ≥0.95 and an individual call rate ≥0.99, resulting in 43,630 variants and 24 sheep (BARB dataset). To analyze the breed in a Mediterranean context, the raw data of Tunisian Barbarine were combined with genotypic data of 59 sheep breeds genotyped with Illumina Ovine SNP50K BeadChip array (hereafter called MED_POP dataset) (Kijas et al., 2012; Ciani et al., 2014; Ciani et al., 2015; Manunza et al., 2016; Gaouar et al., 2017; Mastrangelo et al., 2017; Mwacharo et al., 2017; Mastrangelo et al., 2018; Ben Jemaa et al., 2019; Ciani et al., 2020; Chessari et al., 2023). The random sampling selection procedure implemented in the BITE R package (Milanesi et al., 2017) was used to select a maximum of 30 representative individuals per breed. PLINK ver. 1.9 (Chang et al., 2015) was also used to perform quality control on MED_POP dataset: a minor allele frequency ≥0.05, a genotype call rate for a SNP ≥0.95 and an individual call rate ≥0.99. Furthermore, markers with high linkage disequilibrium (LD) were removed by applying the function--indep-pairwise with an LD threshold of r ² > 0.2, a window size of 50 SNPs and a window step size of 10 SNPs. The filtered dataset contained a total of 33,049 variants, 60 breeds and 1,151 animals, from 10 countries: Albania, Algeria, Cyprus, Egypt, Greece, Israel, Italy, Libya, Spain, Tunisia. The breeds of the final dataset used for the population analyses are summarized in Supplementary Table S1.

BARB dataset was examined through PLINK ver. 1.9 (Chang et al., 2015) to estimate observed (H_O) and expected (H_E) heterozygosity, the inbreeding coefficient (F_IS) and the average minor allele frequency (MAF). Additionally, historical trends in effective population size (N_e) were assessed using SNeP v1.1 software (Barbato et al., 2015), while contemporary N_e was also determined using NeEstimator ver. 2.1 (Do et al., 2014), following the random mating option, within the linkage disequilibrium method proposed by Waples and Do (Waples and Do, 2010).

The analysis of Runs of Homozygosity (ROH) was conducted using the consecutive runs method implemented in the R package detectRUNS (Biscarini et al., 2018). Specific parameters were defined for the inclusion of a ROH: (i) a minimum of 15 SNPs, (ii) no opposite genotypes (maxOppRun = 0), (iii) no missing genotypes within a ROH (maxMissRun = 0), (iv) the maximum allowed gap between consecutive SNPs within a ROH was set at 250 kb, and (v) the minimum length of a ROH was established at 1 Mb.

ROH segments were categorized into five length classes following the nomenclature in literature (Kirin et al., 2010; Ferenčaković et al., 2013): 1–2, 2–4, 4–8, 8–16, and >16 Mbp. The mean number of ROH per individual (N_ROH) and chromosome (NC_ROH), as well as the average length of ROH in Mb per individual (L_ROH) and chromosome (LC_ROH) were computed. Furthermore, the genomic inbreeding coefficient (F_ROH) was calculated by the total length of the genome comprised in the ROH for each individual divided by the total autosomal genome length (∼2.4 Gb).

The identification of markers present in the rich homozygous regions (ROH islands) was performed through the computation of the standard normal z-score from the incidence of all SNPs within ROHs and deriving p-values. The markers within the top 0.1% were considered to constitute ROH islands. The genomic coordinates of markers were examined using NCBI Genome Data Viewer according to Assembly OAR4.0 (GCA_000298735.2). QTL information from the Animal QTL Database for the OAR4.0 assembly, release 48, was used.

Population relationships and structure were studied using merged datasets (MED_POP dataset).

To explore individual similarities within the matrix, multidimensional scaling (MDS) analysis on pairwise identity-by-state (IBS) distances was performed using PLINK ver. 1.9 (Chang et al., 2015), with the—cluster and—mds-plot functions.

Additionally, Arlequin ver. 3.5.2.2 (Excoffier and Lischer, 2010) was employed to estimate population relatedness through pairwise F_ST, and the results were visually presented using a heatmap generated by the R package ggplot2 (Wickham, 2016).

To further visualize relationships, a Neighbour-Joining tree, based on pairwise Reynolds’ genetic distances, computed with Arlequin ver. 3.5.2.2 (Excoffier and Lischer, 2010), was created using SplitsTree4 ver. 4.14.8 (Huson and Bryant, 2006). Patterns of ancestry and admixture were explored using the ADMIXTURE software v1.3. (Alexander et al., 2009), with the unsupervised model-based clustering algorithm, which estimates the individual ancestry proportions given a K number of ancestral populations. The most likely number of clusters was estimated following the cross-validation procedure, estimating the prediction errors for each K value. Circle plots were visualized using the membercoef. Circos function in the R package BITE ver. 1.2.0008 (Milanesi et al., 2017) with an unsupervised model-based clustering algorithm.

To evaluate the variability in Tunisian Barbarine breed, genetic diversity indices were computed (Table 1).

The results revealed a moderate level of observed and expected heterozygosity (H_O = 0.390 ± 0.141; H_E = 0.388 ± 0.109). Additionally, the value of MAF was 0.298 ± 0.124, while the genomic inbreeding coefficients showed low values (F_IS = −0.007 ± 0.031; F_ROH = 0.017 ± 0.028).

Regarding N_e results, a continuous decline in effective population size can be observed across generations (Supplementary Figure S1). Specifically, the N_e estimated for 13 generations ago was 140. Furthermore, according to NeEstimator, the contemporary effective population size is N_e = 577.

The ROH analysis resulted in a total of 388 ROHs in all 24 individuals.

The N_ROH was 13.50 ± 7.65, while NC_ROH reported a mean value of 14.92 ± 10.33.

67% of ROHs had a length between 1 Mb and 2 Mb. ROHs of 2–4 and 4–8 Mb had a similar incidence of 15% and 13%, respectively, while long segments (>8 Mb) represented only 5% of the total. The L_ROH was 1.83 ± 1.02 Mb and the LC_ROH was 2.58 ± 0.98 Mb.

The number of ROH per chromosome displayed a specific pattern with the highest numbers found for the first three chromosomes, consistent with expectations, as OAR1 is the longest chromosome. The top 0.1% of the SNPs-in-run percentile distribution were selected in Tunisian Barbarine breed to identify ROH islands. The results, as illustrated in a Manhattan plot (Figure 2), recognized the presence of a ROH island on OAR13, with 36 SNPs and 14 different genes, with QTLs referring to fat tail deposition and milk yield. Furthermore, a second ROH island on OAR2 with 11 SNPs and 4 genes was identified, but it does not appear to be associated with quantitative trait loci (QTLs) (Table 2).

To unravel the genetic relationships of the Tunisian Barbarine sheep with the Mediterranean breeds considered in the analysis, an MDS analysis based on pairwise IBS distances was performed on the MED_POP dataset (Figure 3).

The first two principal components, collectively capturing 23% of the overall variance (C1 = 16.19% and C2 = 7.27%), effectively delineate a coherent geographic distribution of the breeds.

All European breeds grouped together, giving rise to two distinct branches: the first consisting of breeds from Israel and Cyprus areas, intricately linked to a cluster comprising Greek and Albanian breeds, acting as connecting nodes to the European cluster; the second with Egyptian and Libyan (also Libyan Barbarine) breeds form a discernible cluster, closely linked to the group characterized by Tunisian Barbarine, North African (Tunisian, Algerian), and Spanish breeds.

To enhance the description of the relationship between Tunisian Barbarine and other breeds within the same cluster, a MDS plot was applied to the same dataset, with breeds grouped by names (Supplementary Figure S2). The cluster characterized primarily by the breeds from North Africa and Tunisian Barbarine, includes the Spanish Canaria de Pelo, which stands out as the outermost point in the group. Additionally, closely overlapped within this cluster are Tunisian Barbarine and Algerian Barbarine, other Tunisian breeds like D’man, Queue Fine de l’Ouest, Algerian D’men, Algerian Ouled Djellal, Rembi, Sidaoun, Hamra, Berber and Tazegzawt. The Lybian Barbarine breed does not overlap with the North African breeds or with Tunisian Barbarine. Egyptian breeds, including Farafra, Saidi, Souhagi, Ossimi and Aburamad-Halaieb-Shalateen constituted the other African cluster. Pairwise F_ST values calculated among all the 60 breeds ranged from 0.001 to 0.224 (Supplementary Figure S3) and identify the following pairs of breeds as the most distant: Sardinian Ancestral Black vs. Awassi (F_ST = 0.203), Tunisian D’man vs. Awassi (F_ST = 0.204) and Chios vs. Tunisian D’man (F_ST = 0.201). In general, Awassi obtained the highest values of F_ST among of the analyzed breeds. Tunisian Barbarine showed the lowest values of genetic differentiation with the following breeds: Barbarine (from a different sampling) (F_ST = 0.002), Algerian Barbarine (F_ST = 0.001), Algerian Ouled Djellal (F_ST = 0.004), Queue Fine de l'Ouest (F_ST = 0.008), Rembi, (F_ST = 0.003), while with the Lybian Barbarine (F_ST = 0.017); this higher value is also confirmed by clusters in MDS plot. Moreover, of interest are the relatively low values of F_ST showed by Tunisian Barbarine with Spanish Rasa Aragonesa (F_ST = 0.020), Algerian Sidaoun (F_ST = 0.021), the Egyptian Saidi (F_ST = 0.021) and the Italian Bagnarola (F_ST = 0.021). The Neighbor-Joining tree depicted in Figure 4, using Reynolds’ genetic distances, places Tunisian Barbarine in alignment with the populations of origin grouped by country, confirming the shared phylogenetic root between Tunisian Barbarine and Algeria explained by the F_ST results.

The ADMIXTURE analysis, ranging from 2 to 60 K, provided insights into the population structure to identify ancestral components shared among different breeds (Figure 5). The breeds were clustered by geographical distribution. The results suggest that K = 31 is the most likely number of inferred populations. At K = 2, a clear separation based on geographical origin is evident, with a predominant distinction between North African breeds (including Tunisian Barbarine) from Israel and Cyprus characterized by red predominance, and European breeds (Italy, Spain, Albania, and Greece) represented by blue presence. Moving to K = 10, a shared genetic background is confirmed between Tunisian Barbarine and other North African breeds, indicating a similar internal substructure. Within the geographical regions, discernible internal substructures are also observed within breeds from Italy and Albania. Other geographical clusters (such as North African breeds) demonstrate a dominance of yellow, pattern also visible in Tunisian Barbarine, and a dominance of brown, which characterizes Spanish breeds. The internal substructure of Tunisian Barbarine and the other North African breeds is confirmed at K = 60, despite each population tending to distinguish with its own cluster. Noteworthy, besides Tunisian Barbarine, there are many other breeds exhibiting internal substructure include Awassi, Souhagi, Altamurana, Leccese, Noticiana, Sardinian Muflon, Queue fine de l’Ouest, Algerian Barbarine, Algerian Ouled Djellal and Rembi.

The study aimed to unravel the genetic structure of the Tunisian Barbarine sheep breed through comprehensive genomic analyses. Genetic diversity, which indicates the extent of genetic differences within a population, was a crucial aspect examined in this research. Actively observing and preserving this genetic diversity is essential to ensure the sustained viability of traditional breeding methods in the future, as emphasized by Asmare et al. (2023). The investigation provided valuable insights into the genetic diversity of the Tunisian Barbarine breed and its relationships with other sheep breeds. These findings shed light on population dynamics and their potential implications for breeding programs.

Notably, the observed heterozygosity (H_O) and expected heterozygosity (H_E) in Barbarine indicated a moderate level of genetic diversity. The differences between H_O and H_E in BARB dataset, even if low, can be attributed to several factors. Firstly, the sampling activity was contingent upon the participation of breeders within the project, thereby influencing the availability of animals for sampling. Additionally, each breeder typically maintained a small flock comprising a limited number of breeding sheep, breeding rams and their offspring. To mitigate the risk of inbreeding, only the breeding sheep and rams were included in the sampling process. Consequently, these factors collectively contributed to the relatively small size of our sample and may have influenced the observed results of genetic diversity within the Tunisian Barbarine sheep sample analysed. In comparison to previous studies on Tunisian Barbarine sheep, our results revealed higher diversity estimates than those observed in another Barbarine sample (Ahbara et al., 2022) (H_E = 0.301 ± 0.015 and H_O = 0.303 ± 0.015), whereas Bedhiaf-Romdhani et al. (2020), showed mean values of genetic diversity indices similar to the present work. This consistency across studies underscores the robustness of these findings and contributes to a comprehensive understanding of the genetic diversity within the Tunisian Barbarine population.

The present study’s results align with those of an analysis on Algerian Barbarine and seven local Algerian sheep breeds (D’men, Hamra, Ouled-Djellal, Rembi, Sidaoun, Tazegzawt, Berber): the individual H_O ranged from 0.220 to 0.370, with an average of 0.350, while the average H_E was 0.370 (Gaouar et al., 2017). Moreover, the findings are in line with those of other studies on Mediterranean breeds and southern and western European breeds reported by other authors (Kijas et al., 2012; Ciani et al., 2014). The observed values of H_E in Tunisian Barbarine sheep highlight a potential effect of genetic dilution due to gene flow with other thin-tailed breeds as already reported in previous studies (Gaouar et al., 2017; Bedhiaf-Romdhani et al., 2020; Ahbara et al., 2022); the crossbreeding tendencies resulted as a consequence of a shift in preference and the economic value of tail fat (Bedhiaf-Romdhani et al., 2008; Bedhiaf-Romdhani et al., 2020).

Regarding the F_IS results, Tunisian Barbarine showed a negative value indicating a low inbreeding. Similar findings were obtained in other studies conducted on different samples of Tunisian Barbarine (F_IS = −0.030 ± 0.010) (Bedhiaf-Romdhani et al., 2020), Algerian and Moroccoan breeds (F_IS mean = −0.040, s.d. = 0.060) (Belabdi et al., 2019), and other Algerian breeds (mean F_IS = − 0.070) (Belabdi et al., 2019).

In examining genetic inbreeding via ROH, Tunisian Barbarine exhibited a value aligned to that observed in another sample of Barbarine sheep (0.020) (Baazaoui et al., 2021). Interestingly, this value is significantly lower than the F_ROH identified by Ahbara et al. (2022) in different Tunisian Barbarine samples. Moreover, it remains lower than the values recorded for Queue Fine de l’Ouest and Noir de Thibar (Baazaoui et al., 2021).

The effective population size (N_e) computed using SNeP v1.1 software (Barbato et al., 2015), showed a value of 140 at the 13th generation, surpassing the minimum acceptable N_e threshold of 100, crucial for population conservation (Mastrangelo et al., 2017; Persichilli et al., 2021).

This result was higher than that observed in Noticiana (N_e = 76) and lower than those observed in various Mediterranean sheep breeds like Comisana, Leccese, Sardinian Ancestral Black and Altamurana (N_e = 571, N_e = 512, N_e = 303; N_e = 224 respectively) (Ciani et al., 2014; Chessari et al., 2023).

This decline in Ne over consecutive generations underscores potential challenges on application of efficient mating plans and the need of implementation of effective conservation strategies on small local populations. It is important to note that this decline could also be influenced by sampling constraints. Sampling strategy prioritized adult animals to mitigate inbreeding, primarily due to project limitations and limited collaboration among breeders.

However, the result obtained from NeEstimator (Do et al., 2014) indicated a higher value compared to that observed in another Barbarine sample (N_e = 376) but lower than that observed in Noir de Thibar (N_e = 706) (Baazaoui et al., 2021). The analyses suggest that the genetic diversity observed in the Tunisian Barbarine breed, despite its endangered status and small population size, does not indicate significant inbreeding.

The genetic diversity indices and N_e values support this observation, pointing towards a relatively preserved genetic diversity within the breed. Notably, the differences between N_e values depend on the two different approaches used. While SNeP (Barbato et al., 2015) focuses on estimating historical Ne trends, providing insights into the genetic diversity over time, NeEstimator (Do et al., 2014) evaluates contemporary unbiased N_e, offering a more comprehensive perspective on the genetic diversity conservation of the breed.

The evidence strongly supports the conclusion that historical crossbreeding has played a significant role in maintaining genetic diversity in Tunisian Barbarine. This is consistent with its breeding history, which involves interactions with various other breeds over time (Bedhiaf-Romdhani et al., 2008), and with results from ROH analysis.

The method based on ROH is considered one of the most powerful approaches to estimate genomic inbreeding in the livestock species, and in particular for local populations (Addo et al., 2021).

Long ROH segments typically indicate recent inbreeding events, whereas short ROH segments are associated to ancient inbreeding, due to the increased likelihood of recombination events over successive generations (Selli et al., 2021). In Tunisian Barbarine, results of L_ROH and LC_ROH, lower than 3 Mb, with a 67% showing ROH length lower than 2 Mb, could suggest that inbreeding events occurred not recently, even if the count of ROH resulted quite high (388) (McQuillan et al., 2008; Kirin et al., 2010; Purfield et al., 2017). Different findings emerged among sheep breeds in Northwest Africa, including another sample of Tunisian Barbarine, Tazegzawth, Algerian Hamra, Sidaoun, and Moroccan D’Man, revealing variations in higher number and lower average length of ROH (Belabdi et al., 2019). Tazegzawth and the Tunisian Barbarine showed several long ROH fragments, in accordance with high values of F_ROH (i.e. 0.220 and 0.340) (Belabdi et al., 2019). In contrast, other studies reported over 45% of detected ROH falling within the 1 to 2 Mb range or a prevalence of ROH in the shortest length category (Mastrangelo et al., 2018; Getachew et al., 2020; Selli et al., 2021). It is worth noting that the utilization of low-density SNP array for ROH detection may lead to an overestimation of ROH shorter than 4 Mb (Ferenčaković et al., 2013). Despite this, Tunisian Barbarine breed exhibited a considerable number of ROH when compared to counts observed in other breeds (Selli et al., 2021).

In conducting a comprehensive analysis of the genomic relationship and structure between Tunisian Barbarine and Mediterranean sheep populations, various statistical approaches were employed. Multidimensional scaling analysis (MDS) and population differentiation measurement, along with a Neighbor Network, confirmed findings from previous studies (Bedhiaf-Romdhani et al., 2020; Baazaoui et al., 2021). Tunisian Barbarine stands out for its remarkably low genetic differentiation, reflected in the lowest F_ST values, especially when compared to Algerian Barbarine, Algerian Ouled Djellal, Queue Fine de l’Ouest, and Rembi. These breeds overlap significantly in the same cluster on the MDS plot (Supplementary Figure S2). The close genetic proximity between Tunisian Barbarine and Algerian Barbarine aligns with the geographical proximity of the two countries. This closeness is further supported by the observed genetic similarity between Algerian Barbarine and Ouled Djellal, suggesting potential dilution of genetic identity due to crossbreeding with thin-tailed Ouled-Djellal sheep. The decline in the Barbarine population in Algeria by 60% between 1990 and 2000 underscores this influence (Laaziz, 2005; Gaouar et al., 2017).

The F_ST value between Tunisian Barbarine and Rembi is consistent with the results observed for Algerian Barbarine (F_ST = 0.030), supporting the possibility of crossbreeding with Rembi, another thin-tail breed (Gaouar et al., 2017). Genetic closeness between Tunisian Barbarine and Queue Fine de l’Ouest is substantiated by multiple studies, attributing the size reduction in the Tunisian Barbarine population to extensive crossbreeding between these two breeds (Bedhiaf-Romdhani et al., 2020; Baazaoui et al., 2021). The transition from fat-tailed Tunisian Barbarine sheep to thinner-tailed varieties primarily stems from preferences of butchers. Challenges in marketing fat from Tunisian Barbarine tail carcasses led to a decline in the Tunisian Barbarine population, impacting income dynamics. Additionally, the uncontrolled reproduction management and the absence of conservation programs further contribute to this scenario (Khaldi et al., 2011; Baazaoui et al., 2021). The highest F_ST value observed between Tunisian Barbarine and Awassi supports the notion that Awassi has not been used as a thin-tailed breed in the crossbreeding of Tunisian Barbarine. This differs from observations in Ethiopian fat-tailed sheep breeds and Awassi crossbreeding (F_ST = 0.004) (Getachew et al., 2020). In contrast, the F_ST value between Tunisian Barbarine and Libyan Barbarine (0.090) is consistent with the MDS plot, indicating relative closeness but distinct clusters. Higher F_ST value between Tunisian Barbarine and Noire de Thibar (F_ST = 0.070) suggests that, despite being a thin-tailed breed present in Tunisia (Baazaoui et al., 2021), Noire de Thibar did not significantly contribute to Barbarine’s dilution, unlike other North African breeds. In terms of genomic relationships, it can be inferred that, in the Mediterranean context, North African breeds, particularly Algerian breeds with a thin-tailed phenotype, are genetically closest and have been involved in crossbreeding plans. The results of the admixture analysis provided valuable insights into the genetic structure of these populations, and corroborated the patterns observed in the MDS plot and the F_ST values. Specifically, the connection through clusters between Tunisian Barbarine and Queue fine de l'Ouest, within the broader African context, was confirmed. This genetic association was consistent across various K values, with K = 10 revealing a shared ancestral component with the African cluster, characterized by a prevalence of yellow. The close genetic relationship persisted as K increased to 60, where each population formed its distinct cluster, emphasizing the robustness of the connection between Tunisian Barbarine and Queue fine de l’Ouest within the African breeds. Some Italian breeds, as already identified by other authors (Ciani et al., 2014; Chessari et al., 2023), reflect their genetic originality with homogeneous clusters.

The analysis conducted on Tunisian Barbarine enabled an exploration of its ROH pattern, revealing homozygosity hotspots associated with candidate genes and QTLs primarily linked to fat tail deposition and milk yield. The distribution of ROH on chromosomes showed a notable concentration on OAR2 and OAR3 (10%), aligning with expectations given that OAR2 is the second longest chromosome (Nosrati et al., 2021). One specific region on OAR2 (114.38–115.51 Mb) exhibited 11 SNPs and 4 genes (Table 2), with functions not corroborated in the literature. On OAR13 one island was discovered (47.16–49.61 Mb) containing 36 SNPs, 14 genes and 6 QTL related to milk yield and fat tail deposition (Table 2).

The genes LOC101117953, LOC101118207, LOC101110166, and the SNPs (rs412097174, rs417470080, rs422598859) were identified by Gerlando et al. (2019) in a genome-wide study on the Valle del Belice breed and annotated among those potentially influencing milk production. Regarding tail fat deposition QTL, several genes found in the ROH island in Tunisian Barbarine were identified as candidates for fat deposition.

This signal on OAR13 (47.16–49.61 Mb) is shared by other breeds, including Libyan Barbarine, Cyprus Fat-Tail, Chios, Ossimi, Italian Laticauda, and Chinese breeds, indicating a conserved fat-tail signature (Wei et al., 2015; Yuan et al., 2017; Mastrangelo et al., 2018; Baazaoui et al., 2024).

Within this region, the bone morphogenetic protein 2 (BMP2) gene was identified as a candidate gene for fat tail deposition by other authors (Yuan et al., 2017; Baazaoui et al., 2024) and seems to be involved in bone morphology and body shape development (Kijas et al., 2012). This is particularly intriguing given the importance of fat deposits as crucial components of adaptative physiology to survive extreme environments and conditions such as prolonged droughts, cold, and food scarcity (Ahbara et al., 2019).

Long-tailed sheep, which typically have more caudal vertebrae, tend to accumulate more fat, however it is not yet clear whether BMP2 plays a direct role in tail fat deposition or indirectly influences fat accumulation by regulating the number of caudal vertebrae in sheep (Lu et al., 2020).

The BMP2 gene’s impact on body size and muscle development has also been reported in cattle (Ghoreishifar et al., 2020). Additionally, LOC101117953, present in Tunisian Barbarine’s OAR13 island, is a newly identified gene copy resulting from a retro-transposable event originating from the protein phosphatase 1 catalytic subunit gamma (PPP1CC) gene, also located on ovine chromosome 13 (Pan et al., 2022). These findings underscore the importance of identifying unique traits in response to biodiversity loss. Despite genetic dilution with other thin tailed breeds, Tunisian Barbarine sheep still exhibits physiological mechanisms related to fat storage, typical of fat-tailed sheep breeds that inhabit tropical environments. This mechanism of storing and mobilizing body reserves is potentially advantageous in coping with heat stress and limited food and forage availability (Atti et al., 2004).