QTL Mapping and Heterosis Analysis for Fiber Quality Traits Across Multiple Genetic Populations and Environments in Upland Cotton.

An "immortalized F" (IF) population and two reciprocal backcross (HSBCF and MARBCF) populations were constructed to investigate the genetic bases of fiber quality traits in upland cotton across four different environments. A relatively high level of heterosis for micronaire (MIC) in IF population as well as fiber length (FL) and MIC in MARBCF population was observed. A total of 167 quantitative trait loci (QTLs) were detected in the three related experimental populations and their corresponding midparental heterosis (MPH) datasets using the composite interval mapping (CIM) approach. An analysis of genetic effects of QTLs detected in different populations and their MPH datasets showed 16 (24.24%) QTLs of partial dominance, and 46 (69.70%) QTLs of overdominance were identified in an IF population; 89 (62.68%) additive QTLs, three (2.11%) partial dominant QTLs, and 49 (34.51%) over-dominant QTLs were detected in two BCF populations. Multi-environment analysis showed 48 and 56 main-QTLs (m-QTLs) and 132 and 182 epistasis-QTLs (e-QTLs), by inclusive composite interval mapping (ICIM) in IF and two BCF populations, respectively. Phenotypic variance explained by e-QTLs, except for MARBCF population, was higher than that by m-QTLs. Thus, the overdominant, partial dominant, and epistasis effects were the main causes of heterosis in the IF population, whereas the additive, overdominant, and epistasis effects were the primary genetic basis of heterosis in the two BCF populations. Altogether, additive effect, partial dominance, overdominance, and epistasis contributed to fiber quality heterosis in upland cotton, but overdominance and epistasis were the most important factors.