| Cotton is grown in more than80countries, and contributes to the world economy as a leading natural fiber crop in the textile industry and a source of oil and protein from cottonseeds. The genus Gossypium consists of approximately45diploid (2n=2x=26) and5tetraploid (2n=4x=52) species, including four cultivated species, G. arboreum L. G. herbaceum L., G. barbadense L. and G. hirsutum L.. G. hirsutum L., commonly known as upland cotton, is the most important species and provides accounting for about95%of the world’s cotton fiber production.Genetic improvement of fiber yield and quality is always the top priority goal in cotton breeding programs. However, due to the typically negative association between fiber yield and quality and the narrow genetic base of modern cotton cultivars, it is a major obstacle for cotton breeders to continuously improve fiber quality and simultaneously maintain its productivity, and the progress to break the negative association between fiber yield and quality only using conventional breeding methods has been limited. Thus, innovative approaches have to be incorporated in cotton breeding.Great advances in molecular marker technologies make it possible for breeders to find a rapid and precise alternative approach to conventional selection schemes. Genetic maps with high marker density lay a good foundation for facilitating marker-assisted selection (MAS) of quantitative traits, identifying quantitative trait locus (QTL) and map-based cloning. However, due to the low polymorphism of DNA marker resulting from narrow genetic base of modern upland cotton, the present intraspecific genetic maps are characterized with less loci and larger average distance between two adjacent markers, and are barely satisfactory for MAS and map-based cloning. Thus, it is imperative to construct entirely genome-covered and high-density upland cotton maps and to identify QTL precisely and completely.Based on the genetic map from (CCRI35×Yumian1) F2:6recombinant inbred line population in our laboratory, the present study increases the marker density of previous genetic map using new available SSR and detects QTL controlling fiber quality traits. The main results are as following:1. The phenotypic analysis of fiber quality traitsFiber strength of CCRI35was3.6-7.1cN/tex lower than that of Yumian1in the five environments from2008to2012, whereas the other fiber quality traits had no significant difference between two parents. The phenotypic data of fiber quality traits for recombinant inbred line population presented a wide range of variation and distributed continuously.Correlation analysis showed that the fiber strength was extremly significant positive-correlated with fiber length and fiber length uniformity, the fiber length uniformity was significant positive-correlated with fiber Micronaire reading; the fiber length, fiber strength was extremly significant negative-correlated with fiber elongation and fiber Micronaire reading, respectively; no significant correlation was observed between fiber length and fiber length uniformity, fiber length uniformity and fiber elongation, fiber elongation and fiber Micronaire reading.2. Maker polymorphism and genotypingA total of8232pairs of SSR markers were used to screen for polymorphism between the mapping parents CCRI35and Yumian1, and404polymorphic primer pairs were obtained, accounting for4.9%of total primers. The polymorphic primers were used to genotype180recombinant inbred lines, and generated420loci.A total of1380loci including420loci in the present study and960loci obtained by our lab in previous were used to perfor χ2test, and518loci distorted from the expected Mendelian segregation ratio1:1(P<0.05), accounting for37.5%of the total loci. Among the518distorted loci,452loci were excessive for Yumian1alleles, whereas66loci were excessive for CCRI35alleles.3. Updated genetic mapAll1380loci were employed to perform the linkage analysis, and a genetic map containing1273loci was obtained. The map spanned3076.5cM with an average distance of2.42cM between two adjacent markers. A-genome contained500loci, spanning1462.6cM with an average distance of2.93cM between two adjacent markers, whereas D-genome contained773loci, spanning1613.9cM with an average distance of2.09cM between two adjacent markers.4. QTL controlling fiber quality traitsBased on the updated genetic map and the phenotypic data of fiber quality traits from five environments, a total of60QTL were detected, including15QTL controlling fiber length distributed on14chromosomes, explaining between6.1%and13.4%of the phenotypic variation;11QTL controlling fiber length uniformity distributed on11chromosomes, explaining between6.3%and11.4%of the phenotypic variation;9QTL controlling fiber strength distributed on8chromosomes, explaining between6.1%and26.5%of the phenotypic variation,10QTL controlling fiber elongation distributed on9chromosomes, explaining between6.4%and11.1%of the phenotypic variation; and15QTL controlling fiber Micronaire reading distributed on13chromosomes, explaining between6.3%and15.4%of the phenotypic variation.Among the60QTL,11QTL were detected in two or more environments and13QTL were consistent with previous studies.In addition,38favorable alleles originated from Yumian1, and22favorable alleles originated from CCRI35. |
