Optimum and Decorrelated Constrained Multistage Linear Phenotypic Selection Indices Theory

Area of Science:

• Quantitative Genetics
• Animal Breeding

Background:

• Previous research focused on unconstrained multistage linear phenotypic selection indices (OMLPSI and DMLPSI).
• A need exists to extend these theories to constrained selection contexts.

Purpose of the Study:

• To extend OMLPSI and DMLPSI theories to the constrained multistage linear phenotypic selection index (OCMLPSI and DCMLPSI) context.
• To evaluate the efficiency of OCMLPSI and DCMLPSI in predicting genetic merit and gains under constraints.

Main Methods:

• Developed OCMLPSI and DCMLPSI as extensions of OMLPSI and DMLPSI for constrained selection.
• Validated indices using real and simulated datasets.
• Compared OCMLPSI and DCMLPSI based on prediction accuracy for selection response and genetic gain.

Main Results:

• OCMLPSI and DCMLPSI were shown to be the most general multistage indices, incorporating OMLPSI and DMLPSI as special cases.
• OCMLPSI accounts for index correlations across stages, while DCMLPSI enforces null correlations.
• OCMLPSI exhibited higher accuracy than DCMLPSI in predicting selection response and expected genetic gain per trait.

Conclusions:

• The OCMLPSI is a more accurate tool for predicting selection response and genetic gain compared to DCMLPSI.
• Breeders should consider utilizing OCMLPSI for phenotypic selection in constrained breeding programs.