The research team, led by Alma Račkauskaitė of the Department of Animal Breeding and Reproduction, combined two very different kinds of data. The first was a pedigree dataset covering 22,666 horses drawn from the studbooks of all five populations, provided by the Lithuanian Agricultural Data Center. The second was a genetic dataset of 1,293 horses based on parentage verification tests using blood group and blood protein markers, a technology introduced in Lithuania in 1985 and applied for decades in breeding control. Although blood typing has been replaced in many countries by DNA-based methods, the standardized long-term dataset it produced allowed the researchers to evaluate genetic relationships within and between populations with a consistency that few national archives can match.

The pedigree analysis revealed a sharp divide between closed and open breeding systems. Pedigree completeness six generations deep was highest in the closed Zemaitukai population at 96.9 percent, followed by the Trakehner at 82.0 percent and the Arabian at 72.4 percent. The open Baltic Warmblood and Lithuanian Warmblood populations showed far lower completeness, at 39.9 percent and 54.5 percent respectively, reflecting the frequent use of imported animals whose pedigrees extend back only a few generations in national records. The number of founders told the same story from the opposite direction: just eight in the Arabian and 23 in the Zemaitukai, against 431 and 592 in the Baltic and Lithuanian Warmbloods, consistent with their open, internationally connected breeding programs.

Perhaps the most technically interesting result concerns the effective population size, a key measure of how fast a population is losing genetic variation. The researchers calculated it in two ways: from the numbers of breeding males and females, and from the rate of inbreeding per generation. For the Trakehner, with 78 reproductive mares and 21 stallions, the effective population size came out at 211, with moderate inbreeding of 0.0373 and a very low rate of inbreeding per generation of 0.0033, indicating controlled mating strategies. The open populations showed far higher effective sizes based on breeding animal counts, 851 for the Baltic Warmblood and 758 for the Lithuanian Warmblood, but these figures collapsed when calculated from the inbreeding rate, dropping to 311 and 216. That gap of more than 60 percent signals hidden relatedness and unbalanced parental contributions that a simple headcount of breeders conceals.

The endangered Zemaitukai showed the most dramatic discrepancy of all. Its effective population size based on breeding animals was 223, but based on the rate of inbreeding it fell to just 19, reflecting strong inbreeding pressure, with an inbreeding coefficient of 0.234. This pattern echoes findings in other endangered horse breeds, such as the Hucul and Cleveland Bay, where estimates based on inbreeding rates revealed far more severe inbreeding than counts of breeding animals suggested. The lesson, the authors argue, is that effective population size estimates are highly sensitive to the method used, to pedigree completeness, and to mating patterns, and that conservation decisions should never rest on a single metric or on basic thresholds alone.

On the molecular side, the blood typing data covered 49 alleles at 12 blood group loci, plus five blood protein systems including albumin, transferrin, and serum esterase. Across all horses, the average observed heterozygosity was 0.475 and the expected heterozygosity 0.395, indicating moderate diversity. The most polymorphic locus, EAD, carried 12 alleles with an expected heterozygosity of 0.864. Within the Trakehner population specifically, the mean number of alleles per locus was 3.833 across all animals and 3.417 in the reference group of currently breeding animals. Crucially, observed heterozygosity in the Trakehner rose from 0.460 in the full dataset to 0.488 in the reference population, while expected heterozygosity increased from 0.375 to 0.392, evidence that recent breeding efforts are successfully preserving, and perhaps even improving, genetic diversity despite the constraints of a partly closed studbook.

The open populations showed the highest heterozygosity overall, as expected from their constant crossbreeding with imported stallions, but the Baltic Warmblood reference group showed a slight decline in effective allele numbers, hinting at potential erosion of rare alleles under stabilized selection. The Arabian population, by contrast, exhibited the lowest genetic variability of all, with observed heterozygosity of 0.441 in the total population and 0.444 in the reference group, confirming a narrow genetic base. Although the Arabian carried a relatively high total allele count, it also had the highest proportion of rare alleles and lacked several variants found in the other breeds, a signature of long-term breeding within a closed population.

The relationships among the populations emerged clearly from genetic distances, assignment tests, and multivariate analyses. The Trakehner was most closely related to the Baltic Warmblood and Lithuanian Warmblood, with Nei’s genetic distances of just 0.008 and 0.009 in the full dataset, reflecting historical and ongoing gene flow from Trakehner stallions into these open sport horse populations. In the assignment test, only 57.9 percent of all Trakehner horses were correctly assigned to their own population, with the remainder assigned mainly to the Lithuanian Warmblood and the Arabian, the latter reflecting the breed’s permitted Thoroughbred and Arabian ancestry. In the reference group, however, correct assignment rose to 62.1 percent, suggesting that the actively bred Trakehner subpopulation is becoming more genetically homogeneous and internally consistent.

The Arabian and Zemaitukai, by contrast, formed clear, isolated clusters in every analysis, from principal component plots to hierarchical clustering heatmaps. Fully 89 percent of Zemaitukai and 83 percent of Arabian individuals were assigned to their own populations, indicating strong genetic integrity and minimal external introgression. The principal component analysis explained 36.2 percent and 22.3 percent of total variation on its first two axes, with no single dominant marker driving the separation. The heatmap analysis also revealed that reference populations of the Baltic and Lithuanian Warmbloods clustered separately from the full historical gene pools of their own breeds, evidence of recent shifts in allele composition driven by selection, drift, or targeted breeding strategies. The moderate but rising differentiation between the Trakehner and Arabian, with the fixation index increasing from 0.015 to 0.021 in the reference dataset, suggests that structured breeding in recent years is strengthening boundaries between breeds.

The authors conclude that population structure, historical gene flow, and management practices jointly shape genetic variability, and their results point to concrete management actions: optimizing mating plans by balancing sire and dam contributions, monitoring inbreeding and effective population size trends, managing gene flow among the related Trakehner, Baltic Warmblood, and Lithuanian Warmblood populations, and developing targeted conservation strategies for the genetically distinct Zemaitukai and Arabian. The Trakehner’s relatively long generation interval of 12 years, typical of sport horse breeding, may itself help stabilize diversity by slowing genetic turnover. Encouragingly, from 2025 Lithuania has introduced national support for genetic testing and parentage verification of breeding horses, a move expected to open the door to microsatellite- and SNP-based analyses that will build on this blood-group legacy. For a breed whose numbers are falling but whose genes remain resilient, that combination of historical depth and modern tools may be exactly what sustainable breeding requires.

Subject of Research: Genetic diversity and population structure of Trakehner and other horse breeds in Lithuania

Article Title: Genetic variability and associations of Trakehner and other horse populations in Lithuania

Article References: Račkauskaitė, A., Šveistienė, R., Razmaitė, V., Jatkauskienė, V., & Marašinskienė, Š. (2026). Genetic variability and associations of Trakehner and other horse populations in Lithuania. Archives Animal Breeding, 69(2), 323-336. https://doi.org/10.5194/aab-69-323-2026

Image Credits: AI Generated

DOI: 10.5194/aab-69-323-2026

Keywords: Trakehner, Zemaitukai, Lithuania, genetic diversity, inbreeding, effective population size, pedigree analysis, blood group markers, horse breeding, conservation genetics, heterozygosity, studbooks