ABSTRACT. Species of the genus Aquilegia are exceptionally diverse in their floral morphology and color, commonly known as columbine. They are widely planted ornamentals and are highly attractive for hummingbirds. However, little is known about their genetic diversity. We examined the genetic diversity of the species and cultivars using amplified fragment length polymorphism (AFLP) markers. Sixteen EcoRI/MseI AFLP primer combinations produced 327 informative polymorphic bands, with a mean of 20.4 bands scored per primer. Jaccard's coefficient of similarity varied from 0.61 to 0.93, indicative of high levels of genetic variation. Cluster analysis using the unweighted pair group method with arithmetic mean algorithm placed the 64 accessions into two main clusters, each divided into two sub-clusters. The AFLP variability was significantly associated with the geographic origins, as the Asian species and the North American species grouped into two distinct clusters. The genetic diversity found among Aquilegia demonstrated the potential value of Chinese germplasm for cultivar improvement and for widening the genetic basis of breeding programs and breeding material selection. We concluded that AFLPs are informative and can provide significant insights for genetic diversity research in columbine species.

The objective of this study was to determine quantitative trait loci (QTL) underlying ten floral and related traits in Aquilegia. The traits assessed were calyx diameter, corolla diameter, petal length, petal blade length, sepal length, sepal width, spur length, spur width, plant height and flower number. These are important traits for ornamental value and reproductive isolation of Aquilegia. QTL analysis of these traits was conducted using single‐marker analysis and composite interval mapping (CIM). We used an F2 population consisting of 148 individuals derived from a cross between the Chinese wild species Aquilegia oxysepala and the cultivar Aquilegia flabellata ‘pumila’. Resulting CIM analysis identified 39 QTLs associated with these traits, which were mapped on seven linkage groups. These QTLs could explain 1.22–53.28% of the phenotypic variance. Thirty‐one QTLs, which explained more than 10% of the phenotypic variation, were classified as major QTLs. Graphical representations of the QTLs on seven linkage groups were made. Our research provides the potential for future molecular assisted selection breeding programmes and the cloning of target genes through fine mapping.