Search Filter
- Publications
- Abstract of Theses and Dissertations
- Database
- Single- and Two-stage Index Selection Schemes in Cotton (Gossyplum sp...
Single- and Two-stage Index Selection Schemes in Cotton (Gossyplum sp. L.)
Dissertation Abstract:
The study used genetic information obtained from the analysis of a complete diallel cross of five cotton varieties, namely: UPL Ct-2, MK 73, Mc Nair 1032B. Z441-2A495-5, and Acala 1517-10. Smith's, WiIIiams'. and Elston's selection indices with varying component traits were constructed and evaluated under a restricted single- and two-stage selection scheme for their efficiencies in improving lint yield and its components and fiber properties.
The index components showed considerable differences in heritability. Seedcotton and lint yield appeared to be genetically more variable than the prolificacy, boll size, and ginning outturn. However, the fiber properties, together with maturity, were more heritable than lint yield or its components. In terms of genetic relationships, yield was positively correlated with prolificacy, boll size, and ginning outturn. However, the fiber properties and maturity were more heritable than lint yield or its components. In terms of genetic relationships, yield was positively correlated with prolificacy, boll size, and ginning outturn but negatively associated with maturity, fiber length micronaire, and strength.
In all circumtances, Smith's index was generally more superior than Williams' and Elston's indices, but its efficiency was more enhanced relative to direct se lection for yield when seedcotton or lint yield itself was included. (with maturity excluded) in the index.
Under a single-stage scheme where fiber properties were not index components, the index that incorporated profificacy. Iint yield, and ginning outturn was the most satisfactory in predicting higher response in yield. When fiber properties were added. efficiency was increased. But when prolificacy was included. seedcotton and lint yield were the most effective.
Even greater expected yield gains were predicted under a two-stage selection scheme, with the plan to simultaneously select for prolificacy, seedcotton, and lint yield in the first stage and the fiber properties in the second stage as the best. However, just as in all evaluated selection indices under the single-stage selection strategy, correlated responses for fiber properties under this scheme were all negative.
Restricted indices that im posed zero genetic change in fiber properties predicted lower yield gains than their parallel unrestricted Smith's indices under the single- and two-stage se lection schemes. Nevertheless, the expected improvements were still better than direct selection for seedcotton or lint yield. On the other hand, index selection for tiber properties were generally no better than direct selection for the individual characters.