| The upland cotton cultivars and breeding lines are the most effective and efficient resources for cotton genetic improvement among cotton germplasm, for their abundant genetic diversity in breeding target traits pyramided through natural evolution and breeding process. Genetic variations are the prerequisite and basis for selection, so evaluating genetic diversity and relatedness of these materials is very important to germplasm collection, conservation and utilization.In this study,356 upland cotton cultivars and breeding lines,349 released in China and 7 introduced from America, were collected as our experiment panel.Chinese cultivars could be divided into 4 groups as Huanghe River (179), Yangtze River (114), North-west China Inland (29) and North China Early Maturity group (27) according to eco-geographical growing region; and into 6 groups as 1930-1960 (1,26),1961-1970 (Ⅱ,26),1971-1980 (Ⅲ,40),1981-1990 (IV,83),1991-2000 (V,125) and 2000-2005 (VI,49) according to their released age. All the accessions could be divided into 6 groups: Deltapine (110), King (60), Stoneville (51), Foster (28), Uganda (26) and other/anonymous group according to their maternal pedigree. The cultivars introduced from abroad, DPL15, DPL16, STV2B, King, Foster6 and Uganda3, were used as check group in diversity evaluation and comparision, for they had been used as main founder parents in upland breeding programme and derived a lot of cultivars in China. We selected 381 pairs of SSR markers evenly distributed in 26 tetraploid cotton chromosomes to genotype the accessions and planted them in three locations in two years to investigate 21 phenotypic traits.Genetic diversity analysis was carried out at phenotype and genotype level respectively. The results showed that cultivars released in China had greater diversity than check cultivars both in phenotype and genotype. Among historically released cultivar groups,15 of 21 phenotypic traits showed significant difference. The seed cotton yield (SL), lint yield (LY), lint percentage (LP), lint index (LI), bolls per plant (BN), fiber strength (FS), spinning consistency index (FSCI) increased and short fiber index (SFI), fiber elongation index (FS) decreased significantly with the step of breeding period, but fiber length (FL) and micronaire reading (FMic) did not. Fifteen of 21 trait means showed significant difference among 4 eco-geographical groups. The plant height (PH), fruit branch number (FBN), first fruit branch height (FBH) of the Yangtse River group were significantly higher than that of other ecological groups. The PH, FBH values of North China group were lower than that of the other groups. The North-west China group has the highest FS, SFI, SFCI, and fiber uniformity ratio (FU) value. Twelve of 21 traits showed significant difference among 5 pedigree derived cultivar groups. The Deltapine group had the highest PH, FBH, FBN value and those of the King and Foster groups were low. The Uganda group had the highest BN, SY and LY, but their LP and LI were small. The Stoneville and Foster group had the highest LP and LI. The King and Foster group had high FS and FSCI value, meanwhile, their SFI are the smallest. Two hundred and fifty four alleles were detected at 145 loci in 6 CK varieties, while the total allele number in cultivars released in China reached 415, and with the advance of breeding periods, the allele number of each historic group was increasing. That is to say, after the original introduction of upland cotton from the USA, plenty of genetic variation had been created and accumulated in domestic cultivars during the more than one century breeding history. The genetic richness of the Hanghe river group was the highest (2.69), followed by the Yangtze river group (2.46), and that of the Northwest China and North China groups were the lowest (both 2.27). The average Nei’s genetic distance among cultivars released in the North China, Northwest China, Huanghe river and Yangtze river growing region was 0.4143,0.3931,0.3723 and 0.315, respectively, so cultivars from the Yangtze river and the Huanghe river regions had more similarity. Based on the Nei’s distance, most of the same pedigree derived varieties were clustered together, but ecological source were more consistent with the cluster dendrogram. After generations of hybridization and selection, light or (and) temperature sensitive genes might be accumulated and form the genetic background of eco-geographic adaption.The pattern of Linkage disequilibrium in Upland cotton genome was estimated using TASSEL 2.1 software in the variety population with the allele matrix of 145 SSR loci. The result indicate that LD in the upland cotton genome was moderately high, but not even. Relative higher levels of LD (D’> 0.5) mainly distributed in chromosome D5, D6, A9, D9 and A11. The average genetic distance of LD blocks extended was 19.76 cM. To eliminate spurious association, marker-trait association analysis was carried out through TASSEL 2.1 software with the mixed linear model (MLM), the population structure (Q) and kinship (K) were estimated by STRUCTURE 2.3 and by SPAGeDi software, respectively. Four hundred and six marker-trait associations were detected between 145 loci and 19 yield and fiber quality related traits. Fourty six QTL detected by association ananlysis were consistent with the results of previous QTL mapping studies. A lot of significant associations of different traits with the same locus were detected, which might be caused by tightly linked genes or/and pleiotropy.To map QTL conferring Fusarium Wilt resistance in upland cotton, both linkage and association analysis were performed in this study. A composite-cross segregating population, as Xuzhou 142/Yumian 21//Xuzhou 142/9901, was made and used to perform linkage mapping. An intraspecific linkage map with 183 SSR loci and 40 linkage groups was constructed, which covered a total length of 1378.8 cM with an average interval of 7.5 cM, and seven QTL were detected by linkage analysis in the family based population. Meanwhile, marker-trait association mapping was carried out in the cultivar population and 35 significant (P<0.05) associations were detected by association analysis (MLM). Five of the seven QTL detected by linkage mapping were consistent with the results of association mapping.To mine elite alleles of the target traits, trait means of accessions sharing the same allele of stably associated loci were analyzed by ANOVA and LSD multiple comparisons. Sixty alleles with significant phenotypic effects were recommended and the population frequency of each allele in originally introduced and historically released groups were calculated and compared. Among the 60 elite alleles,37 alleles existed in originally introduced cultivars, while the other 23 ones newly emerged in domestic cultivars. In other words, elite alleles were not only selected but also innovated in the breeding process. The elite allele number and their population frequency of historically released groups increased stably with the advance of breeding periods, just as elite cultivars were developed by gradually pyramiding elite alleles in breeding history. The average number of elite alleles possessed in 356 individuals was 25.58, and 53 domestic cultivars had more than 30 alleles. Most alleles presented in originally introduced varieties and emerged in domestic ones were transmitted in Chinese modern breeding process, but some of them were lost. Seven cultivars with more than 32 elite alleles were selected as maternal parents, and optimal donors with single-, three-way-, and four-way-cross combinations to extremely pyramid elite alleles was recommended by molecular design permutation. So, locally adapted cultivars with maximum elite alleles should be recommended as maternal parents, and resources with more elite and complementary alleles be chosen as donor parents. Meanwhile, molecular design and MAS must be performed in cross combination and progeny selection due to multiple loci involved. |
PDF Full Text Request