Multidimensional scaling analysis of sensory characteristics and quantitative traits inwild apricots

Abstract

The Cappadocia Region in Central Anatolia, Turkey is rich in wild apricot genetic resources. In this study, a multivariate analysis method of metric and nonmetric multidimensional scaling analysis (MDS) was used to classify 43 wild apricot genotypes and 5 standard cultivars by using Euclidean distance. MDS was applied, based on dissimilarities, to quantitative traits of weight, height, width and thickness of fruit, stone and kernel weight, flesh/stone ratio, total soluble solids content, titratable acidity and pH in fruit juice, and sensory attributes of fruit size and shape, cavity depth, suture (cheek line), fruit apex, fruit attractiveness, skin pubescence, ground and over color (blush), eating quality, aroma, firmness and juiciness of flesh, uniformity of ripening of fruit, flesh color and texture, skin cracking and pit burn susceptibility, separation of stone, stone size and shape (lateral view), stone surface and kernel taste. The results showed that most of the wild apricots (91.66%) had small sized fruits ranging from 10.5 g (type #53) to 79.3 g (type #68) with an average of 25.27 ± 2.44 g. The genotypes had fair or good eating quality and aroma, good fruit attractiveness and light bitter or sweet kernel taste in general. The harvest date showed quite a wide range between late June and early September among the genotypes. The sufficient number of dimensions was determined by stress value and pseudo-R2 statistics, then relative positions of genotypes were displayed on graphics. The Aprikoz was the most distinct cultivated apricot from the wild genotypes. Both MDS analyses revealed that most of the wild genotypes had similar fruit characteristics while genotypes of #13, #20, #39, #60, #64, #68 differed the most. The wild apricot individuals with interesting features could be used as a parent in apricot breeding programs for the development of commercial cultivars.