Random regression for modeling yield genetic trajectories in Jatropha curcas breeding

Area of Science:

• Plant breeding
• Quantitative genetics
• Agricultural science

Background:

• Genotypic plasticity evaluation is crucial for crop improvement.
• Random regression models (RRM) are underutilized for repeated measures in Jatropha curcas breeding.

Purpose of the Study:

• To apply and assess the potential of RRM for analyzing repeated measures in Jatropha curcas breeding.
• To evaluate grain yield (GY) over six years in 73 half-sib families.

Main Methods:

• Fitted RRM using Legendre polynomials, selected via Bayesian information criterion.
• Estimated variance components using restricted maximum likelihood (REML).
• Predicted genetic values using best linear unbiased prediction (BLUP).

Main Results:

• Significant genetic variability, plot, and permanent environmental effects were detected.
• Variance components and heritability estimates increased over time.
• Genotype × measurement interactions were identified, showing dynamic progeny rankings.

Conclusions:

• RRM is a viable and efficient tool for genetic selection in Jatropha curcas breeding programs.
• Accurate GY predictions and significant genetic correlations across harvests were achieved.
• Non-uniform genetic trajectories highlight progeny performance differences over time.