Principal components, correlations, path analysis and some descriptors in melon (Cucumis Melo L.) genotypes

Abstract

In Mexico, melon production in 2018 was approximately 595 000 t. Faced with an increasingly demanding market, melon producers have chosen to plant hybrid varieties, which have higher production costs. The objective of this research was to identify the genotypes with the highest yield and quality value, and thus determine if there is a relationship between yield and fruit quality. For this purpose, five lines and one control (Cruiser hybrid) were used, and a random complete block design with three repetitions and 12 variables, recommended as descriptors, were evaluated. For data analysis, a principal component analysis, a correlation analysis and a path analysis were performed. In principal component analysis, CP1 accounts for 53.85% of the variance and CP2 25.97%. Two groups were found highly positively correlated: the first was between the variables fruit weight (WEIGHT), equatorial diameter (DE), polar diameter (DP), pulp thickness (EP) and polar cavity diameter (CPD). In the second group are the variables flavor (SAB) and degrees brix (BRIX). In the path analysis, the variable PESO showed a positive high direct effect, while the DP and EP variables had high negative values. The BRIX variable presented a negative high coefficient value and the SABOR variable a positive high coefficient. Trail analysis showed that the PESO variable is a good indicator for calculating performance. According to the path analysis, for the variables BRIX and SABOR it is recommended to establish another method for the calculation of sugars.