Regression and correlation of yield and yield components of rice (Oryza sativa L.)

Abstract

Yield is a complex trait influenced by multiple yield components; therefore, indirect selection using correlated traits is essential in rice breeding. This study aimed to evaluate the relationships between yield and yield components through correlation and regression analyses in an F7 upland rice population. The experiment used a single-plant design with 150 genotypes derived from a cross between Situ Bagendit and Cibogo. Observed traits included grain yield (weight of filled grains per plant) and several yield components. The findings demonstrated a positive correlation between yield and yield components. Correlation analysis showed that grain yield was strongly and positively correlated with number of productive tillers (r = 0.91), total number of grains per plant (r = 0.90), and number of filled grains per plant (r = 0.95). Regression analysis confirmed the importance of these traits, where the number of productive tillers strongly predicted total number of grains (Y = 169.69x + 54.03; R² = 0.84) and number of filled grains (Y = 166.21x − 104.10; R² = 0.94). Grain yield was best explained by the number of filled grains (Y = 0.028x − 0.042; R² = 0.92). These results indicate that number of productive tillers, total number of grains, and number of filled grains per plant are reliable selection criteria for improving yield in upland rice breeding.

Keywords: upland rice, selection, characters relationship, yield component