Using nuclear microsatellite data to trace the gene flow and population structure in Czech horses

Based on a data set comprising 2879 animals and 17 nuclear microsatellite DNA markers, we propose the most comprehensive in-depth study mapping the genetic structure and specifying the assignment success rates in horse breeds at the Czech population scale. The STRUCTURE program was used to perform systematic Bayesian clustering via the Markov chain Monte Carlo estimation, enabling us to explain the population stratification and to identify genetic structure patterns within breeds worldwide. In total, 182 different alleles were found over all the populations and markers, with the mean number of 10.7 alleles per locus. The expected heterozygosity ranged from 0.459 (Friesian) to 0.775 (Welsh Part Bred), and the average level reached 0.721. The average observed heterozygosity corresponded to 0.709, with the highest value detected in the Czech Sport Pony (0.775). The largest number of private alleles was found in Equus przewalskii. The population inbreeding coefficient FIS ranged from -0.08 in the Merens to 0.14 in the Belgian Warmblood. The total within-population inbreeding coefficient was estimated to be moderate. As expected, very large genetic differentiation and small gene flow were established between the Friesian and Equus przewalskii (FST = 0.37, Nm = 0.43). Zero FST values indicated no differences between the Czech Warmblood-Slovak Warmblood and the Czech Warmblood-Bavarian Warmblood. A high level of breeding and connectivity was revealed between the Slovak Warmblood-Bavarian Warmblood, Dutch Warmblood-Oldenburg Horse, Bavarian Warmblood-Dutch Warmblood, and Bavarian Warmblood-Oldenburg Horse. The breeds' contribution equalled about 6% of the total genetic variability. The overall proportion of individuals correctly assigned to a population corresponded to 82.4%. The posterior Bayesian approach revealed a hierarchical dynamic genetic structure in four clusters (hot-blooded, warm-blooded, cold-blooded, and pony). While most of the populations were genetically distinct from each other and well-arranged with solid breed structures, some of the entire sets showed signs of admixture and/or fragmentation.