Genetic diversity and population structure of five Iranian Paeonia taxa revealed by SSR markers and morphological traits

Background Genetic diversity and population structure remain poorly characterized in Iranian Paeonia despite their importance for breeding and conservation. This study aimed to characterize multilocus genetic diversity and differentiation among five labeled Iranian taxa using an analytical framework suitable for uncertain allele dosage and to descriptively compare genetic and morphological patterns. Results Seventy-two genotypes were analyzed at nine polymorphic simple sequence repeat (SSR) loci using a dosage-agnostic, polyploid-aware allele presence/absence framework. The SSR data were descriptively integrated with a previously generated dataset of locality-level means for 15 morphological traits. Analysis of molecular variance attributed 31.35% of SSR variation to differences among labeled taxa (ΦST = 0.3135, p = 0.001) and 68.65% to variation within taxa. The first two principal coordinate axes accounted for 38.93% of the reported PCoA variation and separated P. wendelboi from the P. tomentosa/P. wittmanniana complex. In the primary nine-locus DAPC, in-sample maximum-posterior label concordance was 66 of 72 genotypes (91.7%), with all discordant assignments involving P. tomentosa and P. wittmanniana . Forty-one threshold-defined multilocus genotype groups were recovered. Genetic diversity was lower in P. wendelboi (Simpson’s index = 0.785; 8 groups/32 samples) than in P. tomentosa (0.941; 19 groups/24 samples). Morphological principal component 1 explained 68.02% of the sample-weighted variation, and the exploratory genetic–morphological matrix correlation was r = 0.510. Mean Sørensen dissimilarity across three pairwise taxon comparisons was 0.535, with turnover exceeding nestedness-resultant dissimilarity. Conclusions Substantial within-taxon SSR variation was detected. The broad genetic separation of P. wendelboi persisted after global exclusion of Loc6 and Loc9 in a seven-locus sensitivity analysis, whereas fine-scale DAPC separation between P. tomentosa and P. wittmanniana was marker-set sensitive. Inference for P. archibaldii and the field-assigned P. mascula accession was limited by sample size. These findings provide a reproducible baseline for conservation sampling and support the need for expanded cytogenetic and genome-wide studies in Iranian Paeonia species.