Selection of morpho-agronomic descriptors for characterization of papaya cultivars

Abstract

This study was conducted to define a list of sufficient minimum descriptors to distinguish between papaya genotypes quickly and precisely. To this end, 30 quantitative and 21 multi-category descriptors related to plant characteristics, such as leaves, flowers, fruit and seeds were evaluated in 27 genotypes of papaya, including crops, local varieties and improved lines. The quantitative descriptors were subjected to principal components analyses using the Singh and direct selection methods, whereas a correlation analysis was conducted for the qualitative descriptors. Eighteen and fifteen quantitative descriptors were discarded by the Singh and direct selection methods, respectively. However, considering the simultaneous analyses of these methodologies, 60% of the descriptors were selected to maximize the total variation of the genotypes. Six of the multi-category descriptors were monomorphic, and two were highly correlated with other characteristics and were discarded. The minimum descriptors that were selected had high discrimination potentials when they were analyzed together. Thus, for the purposes of the protection of varieties and the classification of the genotypes of papaya, there were found to be 18 quantitative and 13 multi-category minimum descriptors that contributed significantly to the total variation and possessed low correlation with each other. The elimination of descriptors did not entail a loss of information. Those descriptors that contributed most significantly to the first three principal components were the stem diameters, fruit lengths and widths, inflorescence peduncle lengths, thickness of fruit skin, leaf widths, dry and fresh seed weights and ratio between fruit lengths and widths and between total soluble solids and total titratable acidities.