Evaluation Of Cotton Chromosome Segment Introgression Lines And QTL Mapping For Boll Number Per Plant Using IL-19-10

In the present study, a set of CSILs using G. hirsutum TM-1as the recipient parent and G.barbadense Hai7124as the donor parent was evaluated in five environments in three years,and a F2population was also built to fine mapping QTL for boll number per plant in IL-19-10.The main results are as follows:1. A frequency distribution diagram was ploted for yield and fiber quality traits of CSILsby software SPSS17.0, phenotypic analysis and two-way ANOVA were also made. All traitsexhibited approximately normal distribution, with absolute value of kurtosis and skewnessless than1, indicating typical quantitative trait distribution and fit for QTL mapping. Theeffect of environment, genotype, interaction between environment×genotype on all traitswere significant, indicating that yield and fiber quality traits were significantly influenced byall the three factors. Environment factor had the most important effect on phenotypic variationfollowed by genotype and environment×genotype interaction contributed the least, andenvironment and area are the major sources of phenotype variation.2. Phenotypic traits were surveyed for yield traits such as boll number per plant, bollweight, lint percentage, etc and fiber quality traits such as strength, fiber length, fibermicronaire, etc in CSILs, and QTL mapping was performed using QTLIciMapping3.0. Atotal of117QTL for17traits were identified, including44QTL for yield traits and73QTLfor fiber quality traits, explaining2.8%-52.79%of phenotypic variation respectively. Amongall117QTL, some stable QTL were mapped, including1QTL for boll number per plant and1QTL for fiber strength which could be tested in three different environments, and3QTL forlint percentage,1QTL for lint yiled,1QTL for strength,1QTL for, fiber length1QTL forfiber micronaire,1QTL for Uniformity,3QTL for Elongation,3QTL for Yellowness, which could beidentified in two different environments. Results indicated that QTL Distribute unevenly onand sub-genome, the At sub-genome had more QTL. Several QTL clusters were identified onchromosome A5、A8、A9、 A1、A5、A8、D5、A5、A7、A10、A12and D11. QTLs detectedhere were divided into single QTL and cluster QTL, containing18QTL (17.6%) and99QTL(82.4%) respectively.3. A F2population of IL-19-10×TM-1was designed for fine mapping QTL for bollnumber per plant. A total of4QTL were identified, including2for boll number per plant and 1for lint percentage and1for boll weight respectively. One QTL for hot season boll numberexplained6%of the phenotypic variation, of which the Hai7124allele showed positive effectin the direction of increasing boll number. Another QTL for autumn boll number explained4%of the phenotypic variation, of which the Hai7124allele showed positive effect too.Several novel CSILs for each yield and fiber related trait were identified by this work,indicating G.barbadense is a valuable and important gene resources and construction of CSILpopulation is a useful way to detect such genes or chromosome fragments efficiently. TheseCSILs provide an useful platform to fine mapping and cloning such important G.barbadensegenes so as to molecular marker assisted breeding of G.hirsutum..