Combining ability, heterosis, and genetic distance among nine elite American Pima cotton genotypes (Gossypium barbadense)

Abstract

American Pima cotton (Gossypium barbadense L.) is grown in the southwest United States, and no Pima cotton hybrid is commercially grown. In this study, six commercial Pima cultivars (Pima S-7, Pima S-6, Phy 76, BR 007, DP 340, and DP 744) and three elite germplasm lines (SI 1331, 93252, and 8810) were crossed in a diallel mating scheme without reciprocals, and their 36 F2 and F3 progeny were planted in replicated tests at two locations (L) in 2004–2005. Yield and fiber traits were analyzed using Griffing’s Model 1, Method 2 diallel analysis. Genetic distance (GD) of parent lines was determined using 661 AFLP molecular markers. Across locations, significant genotype (G) and general combining ability (GCA) effects were detected for yield, boll size, lint percent, micronaire, 2.5% fiber length, fiber strength, and elongation. Significant specific combining ability (SCA) effects were detected for yield, micronaire, and strength. Lint yield and the yield components- boll size and lint percent exhibited significant G × L interactions. Significant GCA × L interactions occurred for yield and lint percent. Yield, boll size, and micronaire also had significant SCA × L interactions. No significant GCA or SCA by L interactions were detected for fiber length, strength, and elongation. The mean mid-parent heterosis (MPH) of the 36 hybrids in lint yield was low in F2, but non-existent in F3. Among the nine parents, Pima S-7 and DP 340 had the highest and significant positive GCA effects for lint yield, lint percent, and boll weight across tests. Only several F2 hybrids with Pima S-7 and DP 340 as one parent displayed some levels of useful heterosis in lint yield in the two locations, as compared to Pima S-7 as the check. MPH for other agronomic and fiber quality traits was low to non-existent in both generations. The average GD of the parents were not significantly correlated with their GCA effects, and GD between parents was also not correlated with SCA and MPH of their hybrids, with the exception of lint yield at the Las Cruces location where a significant negative correlation was obtained. However, GD of the parents was correlated with the performance of hybrids in lint yield, fiber length, fiber strength and elongation in one or two tests. This study represented one of the first studies to estimate heterosis, combining ability, and their relationship with genetic distance of parents in a representative subset of Pima cotton genotypes.