Multiple-trait model through Bayesian inference applied to Jatropha curcas breeding for bioenergy

Area of Science:

• Plant breeding
• Quantitative genetics
• Bioenergy crop development

Background:

• Repeated measures analysis in plant breeding often benefits from multiple-trait models.
• These models account for genetic and residual correlations, enhancing selective accuracy.
• Jatropha curcas is a key perennial crop for bioenergy production.

Purpose of the Study:

• To propose and evaluate a multiple-trait Bayesian model for analyzing repeated measures in Jatropha curcas.
• To improve the genetic evaluation and selection accuracy for Jatropha curcas breeding programs.

Main Methods:

• Utilized a multiple-trait Bayesian model with the Markov Chain Monte Carlo algorithm.
• Analyzed grain yield data from 730 individuals across 73 half-sib families over six harvests.
• Estimated genetic parameters, genetic values, and genetic correlations between harvest measures.

Main Results:

• Observed low, moderate, and high genetic correlations between different harvest yields.
• Bayesian analysis yielded robust genetic parameter and value estimates with high selective accuracies.
• The full model, selected via the deviance information criterion, proved effective.

Conclusions:

• The multiple-trait Bayesian model facilitates reliable selection of superior Jatropha curcas progenies.
• This model is recommended for genetic evaluation of Jatropha curcas and can be generalized to other perennial crops.
• The approach enhances breeding efficiency for bioenergy crop improvement.