The genetic variation and population structure of a panel of 90 durum wheat (Triticum durum desf.) consisting of 62 varieties and breeding lines originating from two agro-ecological zones in Bulgaria (Northern and Southern Bulgaria) and 28 introduced varieties from South–western, Central and Eastern Europe, and the USA were determined by 34 microsatellite markers (SSR). The genetic diversity in the modern durum wheat was 0.5612 with 6.88 alleles per locus. Model-based population STRUCTURE analysis identified two sub-populations (K = 2) separating the South Bulgarian varieties (SP1) from all others (SP2), including South-Western, Central-Eastern European and North Bulgarian ones. Subsequent genetic structure analysis at K = 4 revealed an additional division of each sub-population into two (SP1-2, SP1-4, SP2-1, SP2-3). DAPC analysis and UPGMA dendrogram based on SSR data were in excellent agreement with the Subpopulations defined by STRUCTURE analysis. The Principal Component Analysis (PCA) revealed that among the studied 5 agronomically important traits the correlation between grain yield and plant height was the highest and distinguished 10 varieties and advanced breeding lines from Southern Bulgaria (SP1-2 and SP1-4) as most promising in regard to yield stability. This study showed a good relationship between the genetic and phenotypic population structures according to the division of the genotypes by their agro-geographical origin. It will be useful for both breeders and farmers and could serve as a fundament for durum wheat improvement programs under drought prone environmental conditions.

The genetic variation and population structure of a panel of 90 durum wheat (Triticum durum desf.) consisting of 62 varieties and breeding lines originating from two agro-ecological zones in Bulgaria (Northern and Southern Bulgaria) and 28 introduced varieties from South–western, Central and Eastern Europe, and the USA were determined by 34 microsatellite markers (SSR). The genetic diversity in the modern durum wheat was 0.5612 with 6.88 alleles per locus. Model-based population STRUCTURE analysis identified two sub-populations (K = 2) separating the South Bulgarian varieties (SP1) from all others (SP2), including South-Western, Central-Eastern European and North Bulgarian ones. Subsequent genetic structure analysis at K = 4 revealed an additional division of each sub-population into two (SP1-2, SP1-4, SP2-1, SP2-3). DAPC analysis and UPGMA dendrogram based on SSR data were in excellent agreement with the Subpopulations defined by STRUCTURE analysis. The Principal Component Analysis (PCA) revealed that among the studied 5 agronomically important traits the correlation between grain yield and plant height was the highest and distinguished 10 varieties and advanced breeding lines from Southern Bulgaria (SP1-2 and SP1-4) as most promising in regard to yield stability. This study showed a good relationship between the genetic and phenotypic population structures according to the division of the genotypes by their agro-geographical origin. It will be useful for both breeders and farmers and could serve as a fundament for durum wheat improvement programs under drought prone environmental conditions.