Studies On Genetic Effects And Heterosis Of Main Economic Characters In Cotton

Nine upland cotton parents, which were collected from the Yangze River Valley and the Yellow River Valley, were crossed in half diallel design. The 9 parents and 21 F₁ hybrids were arranged in random block with three repetitions to evaluate the genetic effects of agronomic traits and the relationship between callus growth rates and yields. The data processing was conducted by using additive-dominant genetic model (Zhu, 1997). The pairwise similarity coefficients and genetic distance between any two parents were calculated. A dendrogram was constructed using the unweighted pair group method of arithmetic average (UPGMA) on the basis of the similarity coefficient matrices. The main results were as follows:1. Performances of F₁ were between in two parents, or surpassing the better parent in the most combinations. The mean heterosis of yield components over the check variety were all negative, but heterosis over the mid-parent and the higher were found positive in almost all yield components except lint percentage. The female parents contributed more heterosis of yields than male parents.2. The result of genetic effects were showed that boll weight was completely controlled by dominant effects; the seed index, lint index and boll number were mainly controlled by dominant effects. So we can take advantage of the high heterosis in selecting the combinations. The additive effects were most important in seed index, lint index and lint percent. Fiber quality parameters were mainly controlled by the additive effects which would affect the selection effects in early generations, but the additive variance ratios were lower. Except fiber length, the additive and environment interaction or dominance and environment interaction existed in other fiber quality parameters. Among yield traits, heritability in the broad-sense of lint yield was the highest, on the contrast seed index was lowest. But heritability in the narrow-sense of all economic traits was low.3. The GCA effects of female parents were higher than male'. Most economic traits of female still had positive GCA effects, although the male parents had negative GCA effects. 'Ekang 3' had the highest positive GCA effects for yield traits and fiber quality parameters. 'Ejing 1' had positive GCA effects for yield traits and 'Simian 3' had positive GCA effects for fiber uniformity, fiber strength and micronaire value. 'Zhongmiansuo10' had the negative GCA effects for yield traits and fiber quality parameters. Cross of 2×3 had the highest positive SCA effects for seed cotton and lint cotton yields which were not significant. Only 2×8 had the most significant positive SCA effects for lint yield among all crosses. Crosses 1×6, 1×8, 3×9 and 4×6 were considered promising for yield traits which had significant positive SCA effects. 4×9 showed low SCA effects for yield traitsbut could improve fiber length and elongation for their significant positive SCA effects. The lowest SCA effects for fiber quality parameters belonged to 2x6 which had the significant negative SCA effects for fiber length and fiber elongation.4. For estimation of covariance for genetic effects of yield traits and fiber quality, it is showed that the varieties with more boll numbers might have low seed index, but the other yield traits would be improved because of the significant positive covariance between them. The significant positive genetic covariance between boll weight and seed cotton yield x lint yield which could help breeders to improve these traits simultaneously. Higher lint percentage usually accompanied with lower seed cotton yield and seed index. The selection of higher lint index would improve seed cotton yield because there was the significant positive genetic covariance between them. The negative dominant genetic covariance existed between uniformity ration and fiber strength > uniformity ration and elongation. The most significant negative dominance-environment interaction genetic covariance was found between fiber length and fiber elongation, fiber strength and fiber elongation. There were significant positive dominance-environment interaction between fiber uniformity and fiber strengths fiber uniformity and micronaire value. Therefore, the affects of environment were strong on agronomic traits of upland cotton.5. SRAP and RAPD markers have been used to analyze genetic diversity of the 9 parents. The polymorphism revealed by SRAP and RAPD markers among upland cotton cultivars was not rich. Nine parents could be divided into three groups at the point where the similarity coefficient was 0.65. The parents between 'Yumian 12' and 'Jinmian 6' had the lowest genetic distance value of 0.200, and the highest similarity coefficient 0.793. And the two parents of 'Simian 3' and 'Jinmian 6' had the highest genetic distance value of 0.826. No significant relationships were found between the genetic distance and Fi yield-, or yield heterosis.6. Callus growth rates of the most Fi were between two parents, some crosses even surpass their better parents. The mean mid-parent heterosis in many crosses were positive but the high-parental heterosis were usually negative. There were no significant correlations between callus growth rates and yield performances such as seed cotton yield, lint cotton yield, and yield heterosis. Therefore, it was not possible to predict yield heterosis by callus growth rates.