Evaluation of Conventional and Marker-Assisted Breeding Methods for the Improvement of Fiber Quality in Gossypium spp.

Excess supply coupled with domestic reliance on export markets dictates that the U.S. cotton industry compete globally in terms of both price and fiber quality. Two cycles of divergent selection for fiber upper-half mean length (UHML) and bundle strength (Str) were conducted within five genetically diverse populations. Realized heritability estimates for UHML and Str were calculated for each cycle, and correlated responses among fiber properties and lint percent (LP) were measured as they responded to selection for UHML and Str. The results suggest that early generation selection for UHML and Str was an effective strategy for the genetic improvement of fiber quality within four of the five populations at College Station, TX. There were consistent negative correlations between fiber properties and LP. However, several strains with simultaneously improved fiber quality and LP were identified within each population, providing evidence of repulsion phase linkage. Marker-assisted selection (MAS) may help mitigate some of the current challenges regarding the genetic improvement of fiber quality, such as low genetic diversity and the negative association between fiber quality and lint yield. A multitude of quantitative trait loci (QTL) for UHML and Str have been identified in the literature, but the use of MAS for the improvement of fiber quality is still rare in public cotton breeding programs. Validation studies are necessary to develop portable genetic markers and to identify QTL with stable effects on trait expression across environments and genetic backgrounds. The effects of previously reported microsatellite markers (SSRs) linked to QTL for UHML and Str were evaluated in three genetic backgrounds, and the efficiency of MAS for fiber quality utilizing SSRs linked to stable QTL for UHML and Str was investigated. Using the results of 31 published QTL mapping studies, six SSRs associated with stable QTL for UHML and six SSRs associated with stable QTL for Str were identified. In all but one case, the genetic gain achieved through marker-based selection of individual plants having four-to-six beneficial alleles for UHML or Str was similar to that achieved by phenotypic selection of the top 20%