Evaluation of the heritability of key architectural traits and yield components in the Beninese cotton breeding program

Abstract

Classical plant breeding relies upon the high heritability of traits under selection. The objective of this study was to evaluate the heritability of important agronomic and morphological plant descriptors which are routinely recorded in the Beninese cotton (Gossypium spp.) breeding program. These descriptors collected through plant mapping are: plant height (HT), height of the first fruiting branch (HPBF), height of the top fruiting branch (HDBFC), position of first fruiting branch (NPBF), length of the longest vegetative branch (LBV), length of the longest fruiting branch (LBF), number of vegetative branches (NBV), number of bolls on the longest vegetative branch (NCV), and number of bolls on the longest fruiting branch (NCF). Data were collected from micro-trials conducted from 2013 to 2016, using a total of 74 accessions. Broad-sense heritability was estimated using variance components from a linear mixed model. Results reveal that height-related traits (HT, HPBF, HDBFC) were highly heritable (H2>0.60) and positively correlated to the seed cotton yield. NBV and NPBF also were highly heritable. These traits are considered useful descriptors. Other architectural traits (LBV and LBF) are less heritable, they were less consistently evaluated, and their utility in a classical plant breeding program is suspect. NCV and NCF are characterized by high coefficients of variation (CV) and inconsistent estimates for heritability. Other descriptors may be more useful as yield components, since seed cotton yield was not a reliable descriptor.