ADDITIVE MAIN EFFECTS AND MULTIPLICATIVE INTERACTION ANALYSIS IN UPLAND COTTON (Gossypium hirsutum L.)

Abstract

Twenty genetically diverged genotypes of upland cotton (Gossypium hirsutum L.) received from Cotton Development Board, Dhaka were used to investigate their performances in three locations (Jessore, Rangpur and Dinajpur) over four consecutive years (2010-11, 2011-12, 2012-13 and 2013-14). At the beginning of investigation, the genotypes were characterized and evaluated on 12 morphophysiological characters including seed cotton yield (kg ha-1). Days to 1st flowering was long and ranged from 47-56 days, suggested prolong crop duration in cotton. The unbrust bolls plant-1 was a detrimental character for reducing yield potential of the genotypes. The PCV(Phenotypic Coefficient of Variation) value was higher than corresponding GCV(Genotypic Coefficient of Variation) for all characters, suggested polygenic controlled of the characters and thus had environmental effects (locations and years) for the full expression of the characters. Among the twelve characters, plant height, bolls per plant, boll weight and seeds per boll exhibited higher than 90% heritabilities but plant height itself showed wider difference between PVC and GCV, indicated the strong effect of environment. Therefore, except plant height other three characters might into consideration in selection or hybridization program to improve the ultimate character seed cotton yield. High heritability (h2 b=90.36%) coupled with high genetic advanced (58.49%) was estimated for bolls plant-1 hence the character might improve through simple selection which ultimately improve yield potential in cotton. A total of 55 pairs of characters combinations were drawn both at genotypic (rg) and phenotypic (rp) levels to estimate correlation coefficients. Vegetative branches plant-1 and unbrust bolls plant-1 exhibited negative and significant association and six characters namely secondary fruiting branches plant-1, main stem nodes plant-1, days to 1st flowering, bolls plant-1, boll weight and seeds boll-1 showed positive and significant association with seed cotton yield both at genotypic and phenotypic levels. But simultaneous consideration of genetic parameters, genotypic and phenotypic correlation coefficients and path analysis of genotypic correlation with seed cotton yield. Five characters might consider constructing selection indices for ultimate improve of seed cotton yield among the genotypes. The higher genotypic correlation coefficient in respect of related phenotypic correlation coefficient suggested inherent relationships of the characters either positive or negative directions with seed cotton yield.