Integration of quantitative and qualitative descriptors for genetic diversity studies of watermelon accessions

Abstract

Most of genetic diversity studies are based only on quantitative descriptors. The aim of this study was to apply different analysis strategies with quantitative and qualitative descriptors to identify which is the most suitable to integrate descriptors for genetic diversity studies of watermelon (Citrullus lanatus) germplasm using different assays. A total of 20 watermelon accessions were evaluated in the State of Rio Grande do Norte and two commercial cultivars were used as witness (control) (Crimson Sweet and Charleston Gray) in two cultivation seasons (2014/2015), using 49 descriptors, in which 26 were quantitative and 23 were qualitative. The genetic diversity of the accessions was obtained using eight analysis strategies that combined the Mahalanobis Distance, Average Standardized Euclidean Distance (SED), and Weighted Average Euclidean Distance, followed by groupings using the Tocher optimization method. A wide morphological variability was observed between and within accessions. The eight different strategies of genetic dissimilarity analyses resulted in 36 different groups. The clustering strategy, by which dissimilarity between accessions calculated using SED (quantitative descriptors) and WED (qualitative descriptors) is the most suitable for integrating quantitative and qualitative descriptors in genetic diversity studies of watermelon germplasm.