Biplot Analysis of G×E Interactions for Fruit Yield and Component Traits in Melons (Cucumis melo L.)

Genotypic evaluation in multi-environments is essential to better understand the nature and magnitude of genotype by environment (G×E) interactions for any desirable trait. Melon (Cucumis melo L.) trials were conducted to determine stability of fruit yield and component traits. Nine genotypes including four commercial hybrids were planted at three locations (College Station, Uvalde, and Weslaco, TX) in two years (2010 and 2011). Total fruit yield (TFY; t·ha^-1), marketable yield (MFY; t·ha^-1), percent marketable yield (PCMY), number of fruits per vine (FN) and average fruit weight (FW) were recorded. Data was analyzed by the combined analysis of variance and trait stability by bi-plot analysis. Location had significant effect on all the traits except PCMY. Location × year interaction was significant for TFY, PCMY, and FN. Location × genotype interaction significantly affected TFY and FN, while Location × year × genotype interaction was significant only for MFY and PCMY. Biplot analysis indicated that ‘Orange Casaba’ was the highest mean performing genotype for TFY, MFY, ‘Oro Duro’ for FN, and ‘Mission’ for PCMY. Further, the ideal genotypes for specific traits were as follows: TAMU F39 (TFY), Mission (MFY, FN), and Sol Real (PCMY). Mission was the most stable and average performing genotype for marketable yield across all environments, while the Uvalde location was the most ideal environment for selecting varieties adapted to south-central Texas.