| As the leading natural fiber source, cotton genus comprises of about 46 diploid and 5 tetraploid species, two tetraploid species, G. hirsutum L. and G barbadense L., account for 90 and 8%, respectively, of the world's cotton production. Compared to sea-island cotton, upland cottons generally have low-quality fibers, i.e., short or coarse fibers of relatively low strength. With the advances in spinning technology and the needs of the people, the better fiber quality will be required. These requirements of fiber quality have attracted a lot of efforts of governments and scientists to improve cotton fiber quality, especially that of upland cotton. Conventional genetic strategies have been employed to enhance the fiber properties of upland cotton for over half a century. However, it had limited efficiency to improve cotton. Recent advances in DNA markers offer plant breeders a rapid and precise alternative approach to conventional selection schemes to improve quantitative traits. In upland cotton, the genetic linkage maps have been developed and used to identify QTL for agronomic and fiber quality traits. Unfortunately, so far, these maps have covered only a relatively small part of cotton genome (11.6%19.5%) and more detailed linkage maps covering a wide range of the genome is needed to map QTL precisely and completely. Though widely testing AFLP polymorphism among different upland cotton varieties, it was found that the AFLP polymorphism was much higher between high quality cultivar Yumianl and multiple dominant gene line T586 than other cultivars tested. A genetic linkage map was constructed with SSR, AFLP and morphological markers using 117 F2 plants developed from a cross between upland cotton cultivar Yumianl and T586. The present genetic linkage map was used to identify and map the quantitative trait loci (QTL) affecting fiber-related traits in 117 F2:3 family lines. The mainly results were as following:DNA polymorphism between mapping parentsThe 64 AFLP primer combinations yielded a total of 149 polymorphic bands. Considering a total of 3980 bands, the polymorphism between the two upland cotton parents as revealed by AFLP is 3.7%. An average of 2.3 (between 1 and 7) informative AFLP per primer pair was counted. In 2002, a total of 73 polymorphism primer pairs were found among 291 cotton SSR primer pairs (217BNL primer pairs and 74 JESPR primer pairs), and the polymorphism primer pairs accounted for 26.0% of the total cotton SSR primer pairs. Four polymorphism primer pairs were found among 315 rice SSR primer pairs (RM primer), and the polymorphism primer pairs accounted for 1.3% of the total rice SSR primer pairs. In 2004, another 1526 cotton SSR primer pairs were detected, a total of 434 polymorphism primer pairs were found among a total of 1817 cotton SSR primer pairs, and the polymorphism primer pairs accounted for 22.3% of the total primer pairs. Compared with the other intraspecific upland cotton population, the polymorphism, between mapping parents Yumian 1 and T586, was higher than other mapping parents, and the population develomentd from the cross of Yumian 1 and T586 was an ideal mapping population.Trait performances of mapping parents and F2:3 family linesHigh quality cultivar Yumianl had 45.7% for lint percent, 31.8mm for 2.5% fiber span length, 88.0 for fiber length uniformity, 34.5 cN/tex for fiber strength, 6.8 for fiber elongation, and 4.8 micronaire units for fiber fineness. Multiple dominant gene cultivar T586 had 13.5% for lint percent, 21.0mm for 2.5% fiber span length, 81.3 for fiber length uniformity, 15.9 cN/tex for fiber strength, 9.9 for fiber elongation, and 2.1 micronaire units for fiber fineness, and cultivar T586 was characterized with low yield, poor fiber strength and and short fiber lemgth. The lint percent, fiber length, fiber strength and micronaire reading for Yumianl were 3.4, 1.5, 2.2 and 3.7 times more than that of T586, respectively. The fiber elongation of T586 was 1.5 times than that of Yumianl. The two mapping parents were markedly different in fiber-related traits.The lint percent of 117 F2;3 family lines ranged from 2.53 to 45.52%, with a mean of 28.39%. The 2.5% fiber length of F2:3 family lines ranged from 20.9 to 31.8mm. with a mean of 25.83mm. Fiber length uniformity of F2:3 family lines ranged from 78.6 to 86.8%, with a mean of 83.14%. Fiber strength of F2:3 family lines ranged from 15.7 to 31.3 cN/tex, with a mean of 23.92 cN/tex. Fiber elongation of F2;3 family lines ranged from 5.8 to 10.3, with a mean of 7.84. Fiber fineness of F23 family lines ranged from 2.1 to 6.1 micronaire units, with a mean of 3.41 micronaire units. For lint percent and fiber length uniformity, transgressive segregation was observed falling beyond T586, for micronaire units, transgressive segregation was observed falling beyond Yumianl, and for fiber elongation transgressive segregation was observed falling beyond both Yumianl and T586. Thedifferences of lint percent and fiber quality from (Yumianl XT586)F23 families exceeded the level of the present intraspcecific upland cotton populations, and reached or exceeded the level of the developed interspcecific populations.Linkage mapSeventy-seven SSR primer pairs revealed a total of 177 polymorphismic loci between twoparents. Nine AFLP primer pairs produced 17 polymorphic loci. A total of 202 loci, including 177 SSR, 17 AFLP and 8 morphological markers, were employed to perform linkage analysis using the intraspecific population (Yumianl X T586)F2 of upland cotton. One hundred and seventy-three markers were distributed into 28 linkage groups, while twenty-nine markers remained unlinked, which included 18 SSR, 6 AFLP and 5 morphological loci. The genetic map contained 159 SSR, 11 AFLP and 3 morphological markers (￡2°, Yi and Ts). The linkage groups varied from 1.3 to 215.9 cM with the average distance about 7.9 cM between two markers, and covered 1360 cM or approximately 30.6% of the total recombination length of the cotton genome. Up to date, this map is the most detailed intraspecific map of upland cotton, and a good basis for developing a more detailed linkage map and locating QTL more precisely and completely.QTL effects and origins in upland cottonBased on interval mapping, one QTL was identified for lint percent, explaining 22.6% of the lint percent variance. Three QTL for 2.5% fiber length were identified, explaining 21.7-23.0% of the fiber length variance, collectively explaining 66.7% of the trait variance. One QTL for fiber strength were identified, explaining 10.9% of the fiber strength variance. Four QTL for fiber elongation were identified, explaining 8.5-16.4% of the fiber elongation variance, collectively explaining 50.5% of the trait variance. One QTL for fiber fineness were identified, explaining 14.6% of the fiber fineness variance.Based on composite interval mapping, seven QTL was identified for lint percent, explaining 6.2-29.3% of the lint percent variance, collectively explaining 74.3% of the trait variance. Nine QTL for 2.5% fiber length were identified, explaining 3.8-20.9% of the fiber length variance, collectively explaining 64.0% of the trait variance. Six QTL for fiber length uniformity were identified,explaining 5.3-16.0% of the fiber length uniformity variance, collectively explaining 50.7% of the trait variance. Three QTL for fiber strength were identified, explaining 9.0, 11.7 and 11.3% of the fiber strength variance, respectively, collectively explaining 32.0% of the trait variance. Six QTL for fiber elongation were identified, explaining 7.3-16.2% of the fiber elongation variance, collectively explaining 59.4% of the trait variance. Four QTL for fiber fineness were identified, explaining 5.5-18.5% of the fiber fineness variance, collectively explaining 49.4% of the trait variance.Based on multiple interval mapping, six QTL was identified for lint percent, explaining 2.9-26.4% of the lint percent variance, collectively explaining 64.0% of the trait variance. Seven QTL for 2.5% fiber length were identified, explaining 2.1-26.6% of the fiber length variance, collectively explaining 53.6% of the trait variance. Three QTL for fiber length uniformity were identified, explaining 8.5-21.8% of the fiber length uniformity variance, collectively explaining 39.3% of the trait variance. Five QTL for fiber strength were identified, explaining 2.9-17.2% of the fiber strength variance, collectively explaining 44.0% of the trait variance. Six QTL for fiber elongation were identified, explaining 6.7-11.7% of the fiber elongation variance, collectively explaining 58.7% of the trait variance. Six QTL for fiber fineness were identified, explaining 0.5-29.9% of the fiber fineness variance, collectively explaining 49.4% of the trait variance.As is shown above, the composite mapping and multiple interval mapping detected more QTL than the interval mapping. Most of QTL had effects controlling 5-20% of the total phenotypic variation, and only a few QTL had effects controlling less than 5% of the phenotypic variation. However, all three methods can detect most of the QTL explaining more than 15% of the phenotypic variation, and these QTL may be major QTL, which show that the fiber-related traits are controlled by major genes and minor genes.For all fiber-related traits, the alleles increasing trait phenotypic value, which originated from Yumianl, account for 45.0-60.6% of the detected QTL, and the alleles from T586, account for 39.4-55.0% of the detected QTL. The results show that both parents provide QTL increasing the trait phenotype values, which explained the transgressive segregations among the fiber-related traits of F2:3 family lines.QTL action modesThe action modes of detected QTL were complex, 40.0-66.7% of the mainly effects present from additive to dominant effects, 22.0-31.0% of the mainly effects present from dominant to overdominant effects, and 8.0-30.0% of the mainly effects present overdominant effects. The dominant effects, which increased phenotype value, account for 33.3-51.4% of the detected QTL, and the ressive effects, which increased phenotype value, account for 48.6-66.7% of the detected QTL. The results show that the QTL controlling fiber-related traits were mainly additive, secondly dominant, and meanwhile including overdominant, and that most of QTL appeared to be recessive.Except for mainly QTL, 18 epistatic interactions were detected by multiple interval mapping, which collectively explained 6.9-39.3% of the traits variances. Among the interactions, additive X additive (AA) accounted for 34.5% of the interactions, additiveX dominant (AD) and dominantX additive (DA) collectively accounted for 21.7% of the interactions, and dominant X dominant (DD) accounted for 43.5% of the interactions? The results show that the epistatic effects play an important role in controlling the fiber traits, and additive X additive (AA) and dominant X dominant (DD) are major epistatic components. However, the epistatic effects and the epistatic components are different from the fiber-related traits, e.g., the epistatic effects play a larger roles in affecting fiber strength and fiber fineness, and the lint percent mainly has additive X additive (AA) component, and the fiber fineness mainly has dominantX dominant (DD) component.QTL distributionsMore than 60% of QTL distributed on five to six linkage group, e.g., the LG19 has QTLaffecting lint percent, fiber strength, fiber elongation and fiber fineness, and the LG20 (chromosome 6) has QTL affecting lint percent, fiber length, fiber strength, fiber elongation and fiber fineness, which explained the phenotype relationship among fiber-related traits, but the present study cannot distinguish the relationship between linkage and pleiotropy.Mapping QTL by selective genotypingThe linkage map constructed by selective genotyping population contained 149-170 markers, distributing on 23-23 linkage groups, covering 991.4-1083cM, which accounts for 20.5-23.4% of the...
|
PDF Full Text Request