Genetic Diversity and Candidate Selection Signatures in Hungarian and Romanian Carpathian Water Buffalo Inferred from Cross-Species SNP-Array Genotyping

The Carpathian water buffalo represents a locally adapted but under-characterized genetic group found in Central and Eastern Europe. Genome-wide information on its genetic diversity, population structure and potential adaptive variation remains limited, particularly for Hungarian and Romanian populations. In this study, we genotyped 263 water buffalo individuals from Hungary and Romania using the GeneSeek Genomic Profiler Bovine 100K SNP array to evaluate genetic diversity, the population structure, runs of homozygosity (ROH) and candidate genomic regions showing signatures of selection. After quality control, 214 Hungarian and 33 Romanian individuals and 6605 SNPs were retained for downstream analyses. Both populations showed moderate genetic diversity, with the Romanian population displaying higher minor allele frequency, observed heterozygosity and nucleotide diversity than the Hungarian population. In contrast, the Hungarian buffalo showed a higher burden of runs of homozygosity, including a larger proportion of long ROH segments, suggesting stronger recent autozygosity or a more restricted breeding structure. Principal component analysis and neighbor-joining phylogeny separated the two populations, whereas ADMIXTURE indicated shared ancestry and a within-population substructure rather than complete population-specific differentiation. The integration of standardized FST, absolute allele-frequency differences and ROH islands identified six candidate regions under a positive signature of selection in each population. These regions harbored genes previously associated with immune response, reproduction, growth, milk production and thermotolerance in bovids. Functional enrichment was limited, with significant Gene Ontology terms detected only in the Hungarian candidate regions. Our results provide a regional genomic baseline for the future conservation and breeding management of Carpathian water buffalo. Given the use of a cross-species SNP array and unequal sample sizes, the candidate selection signals should be interpreted as hypothesis-generating and warrant validation using higher-density buffalo-specific genomic data.