Exploring the genetic diversity of South African indigenous and rural horse breed.

Abstract

South Africa boasts a diverse range of indigenous and rural horse breeds that thrive in both commercial breeding operations and rural farming communities. The infusion of Javanese war-horses by Dutch settlers marked the genesis of the Cape Horse, involving subsequent crossbreeding, with various equine stocks – some of which included the English Thoroughbred, the American Saddlebred and the Persian Arabian. The emergence of modern, South African horse populations stands as a testament to human ingenuity, natural selection, and adaptability. To date, there is a paucity of genomic research that has been conducted on indigenous Basotho pony and rural, non-descript South African horse breeds focusing on their multi-functional roles, genetic diversity and adaptive traits. This study aimed to characterize the genetic and functional diversity of the rural and indigenous horse populations in South Africa and providing essential data to counter potential threats of extinction, uncontrolled crossbreeding and exotic breed replacements. Data was collected from 110 surveys and 282 phenotypic records across indigenous (Basotho Pony-type (BSP)), rural (Eastern Cape Rural (ECR); East London Rural (ELR) and KwaZulu-Natal (KZN) rural) and commercial (SA Saddlebred (SAD); Thoroughbred (THR)) horse breeds - collected in KwaZulu-Natal, Eastern Cape and Gauteng provinces. Data was collected across different production systems and horse breed representatives within defined provisional locations, based on literature and consultations with officers belonging to the Department of Agriculture and Rural Development, ensuring selections were relative to local breed availability. Farmers located within rural areas were interviwed based upon voluntary participation, whereby some farmers owned either a minimum of one horse, to some owning multiple horses. Horses were selected based on distant relatedness to minimize genetic bias and therefore the number of horse samples per farmer was not proportionate. This approach was adopted to ensure an unbiased representation of the breed populations, increasing accuracy of genetic diversity estimations. Similarly, the commercial breeder and trainer also participated voluntarily, with horse selection based on both distant relatedness and the number of available horses within the breeding stud and training facility. Data categorized and structured from the detailed surveys and phenotypic records were analysed using IBM SPSS Statistics version 29.0.2.0. Crosstabulation analyses demonstrated that all four rural horse breeds (BSP, ELR, ECR and KZN), were significantly utilized for transportation and rural racing (p = 0.001). The BSP population also showed a unique functional role in meat production within Sotho and Tshwana cultural communities (p = 0.001), as supported by hierarchical cluster analysis (HCA), which placed BSP as the most distinct cluster. HCA revealed the closest similarity between commercial breeds SAD and THR, with an agglomeration coefficient of 0.000, indicating identical functional roles. CATPCA of six phenotypic traits identified five clusters, with rural breeds overlapping due to shared traits, while SAD and some THR individuals distributed separately, reflecting breedspecific phenotypic traits. Ordinal regression analysis found no significant socio-economic impact of these functional roles (p > 0.050), although their contributions improved livelihoods in rural communities through indirect means. The Illumina EquineSNP70k and EquineSNP80k BeadChips were used to analyze genetic diversity of the horse breed populations. Principal Component Analysis (PCA) analysis performed using Golden Helix SNP Variation Suite (SVS) version 8.5.0, identified four distinct clusters, with the commercial breeds (SAD and THR) and the ELR separating from BSP, ECR, and KZN. Global PCA showed all South African breeds clustering together within continental representative clusters. Ancestral proportions analysis (sNMF) conducted within R/Rstudio version 4.3.1, suggested shared ancestry among the clusters, linked to SAD and THR precursor breeds. Minimal differentiation between indigenous and rural breeds was confirmed by Pairwise Fixation Index (Golden Helix SNP Variation Suite (SVS) version 8.5.0) and AMOVA (ARLEQUIN version 3.1) results (low inter-population variation, (𝐹𝑠𝑡 < 0.060; high intra-population variation, F = 94.390%). Runs of Homozygosity (ROH) analysis conducted within R/Rstudio version 4.3.1, revealed shorter accumulative runs (<300 Mbps) in BSP, ECR, and THR populations, while ELR showed variable lengths (300–800 Mbps), indicating recent inbreeding or selective events. SAD and KZN populations exhibited higher total runs. ROH peaks (>0.70) highlighted adaptive selection in KZN and ELR for traits like fertility (NLRP13; NLRP5), immunity (FIZ1; SSC5D), and growth (EPN1; SNAI1). Commercial breeds showed selection for sporting traits but also contained deleterious genes (MAGI1; ADGRL3) in SAD. Inbreeding levels (𝐹𝑅𝑂𝐻 = 0.032–0.326) were highest in SAD, ELR, and KZN, and lowest in BSP, ECR, and THR populations. SNP data extracted from the 282 horse specimens was used to identify and evaluate genomic regions under positive selection within and between the six breed populations in order to provide insight into the adaptive genetic traits preserved. These findings thus contribute to conservation efforts and support the sustainable management of horse populations that hold cultural and economic significance. Strong selective sweeps from population-specific iHS analyses were identified on chromosomes 1, 2, 4, and 7, linked to traits such as fertility and reproduction (COBL, CFAP58, SPACA3), nervous system development (UNC13, ARHGAP35, DTNA, HTR1A), growth (FBN1), and feed efficiency (GLB1L2, B3GAT1, PPARA). Traits associated with nervous system development were found to be significant within these horse populations, highlighting their control system in shaping the behaviour, cognitive function and response to environmental stimuli, which collectively facilitates their rapid adaption to diverse working roles. Cross-population XP-EHH analyses revealed similar selection patterns, reinforcing the fixation of these genomic regions. Intersection of XP-EHH and iHS analyses highlighted unique selection signatures: hair follicle development (SOS1) in ECR, locomotory behaviour (ADAM22) in ELR, and fear response (SLC1A1) in ELR vs. KZN, reflecting adaptations to ecological niches. Additional traits related to blood vessel morphogenesis (SEMA3E, TSPAN12, JAM3) and olfactory responses (OR7D20, OR1M1) further illustrated environmental adaptation. These results underscore the functional and genetic importance of indigenous and rural horse breeds, advocating for conservation policies that sustain their role, in South African societies.