Cassava Breeding I: The Value of Breeding Value

Area of Science:

• Plant breeding and genetics
• Agricultural science
• Crop improvement

Background:

• Cassava breeding involves multiple selection stages, from single row trials (SRT) to uniform yield trials (UYT).
• Evaluating over 20,000 genotypes over 14 years, only 114 reached the final UYT stage.
• Predicting progeny success in later stages from early-stage data is crucial for efficient breeding programs.

Purpose of the Study:

• To assess the predictive power of single row trial (SRT) data for identifying cassava genotypes likely to reach the uniform yield trial (UYT) stage.
• To evaluate the effectiveness of the selection index (SIN) and progenitor breeding values in predicting progeny advancement.
• To explore strategies for improving the accuracy of early-stage selection in cassava breeding.

Main Methods:

• Phenotypic data from SRT were integrated into the cassava breeding program's selection index (SIN).
• Average SIN of progenies per progenitor was calculated as an approximation of breeding value.
• Regression analysis was used to correlate progenitor breeding values and family size with progeny success in UYT.

Main Results:

• Average breeding value (SIN) at SRT was a poor predictor of genotypes reaching UYT (r² = 0.05).
• The number of families and progenies per progenitor more effectively predicted progeny advancement to UYT.
• Large within-family genetic variation masked true progenitor breeding values.

Conclusions:

• Early-stage selection based on average breeding value in cassava is not efficient for predicting later-stage success.
• Progenitor family size is a better predictor than estimated breeding value at the SRT stage.
• Partial inbreeding (e.g., S1 or S2) of progenitors can reduce within-family variation, enhance breeding value accuracy, and accelerate genetic gains.