Analysis of decision-making coefficients of the lint yield of upland cotton (Gossypium hirsutum L.)

Increasing crop yield is a major challenge in modern agriculture. Thus, the relationships between yield and its impact traits in cotton breeding need to be studied. A genetic model with additive, dominance, and their interaction effect with environment was used to analyze the two-year data of 20 parents and 100 F1 hybrids from a 10 × 10 diallel cross (North Carolina II design) in intraspecific upland cotton (Gossypium hirsutum L.) hybrids of high planting density cases. The decision-making analysis of the other 11 traits on lint yield was performed based on correlation and path analyses. Results showed that the order of significant relationships of additive effect in terms of size between lint yield and other traits was as follows: boll weight; boll number of the top three fruit-bearing branches; total boll number; fiber length; micronaire; and plant height. The decision-making traits of lint yield were boll weight and total boll number, fiber length was a main restricting trait. For the genotypic value, the decision-making traits were boll weight and total boll number. The increased phenotypic values of the almost traits could increase lint yield, but high phenotypic values should be selected for total boll number and boll weight to obtain better results. Thus, the use of the decision-making coefficient approach to conduct indirect selection is more applicable than correlation and path analyses in breeding practices.