Optimal Contribution Selection Improves the Rate of Genetic Gain in Grain Yield and Yield Stability in Spring Canola in Australia and Canada

Abstract

Crop breeding must achieve higher rates of genetic gain in grain yield (GY) and yield stability to meet future food demands in a changing climate. Optimal contributions selection (OCS) based on an index of key economic traits should increase the rate of genetic gain while minimising population inbreeding. Here we apply OCS in a global spring oilseed rape (canola) breeding program during three cycles of S0,1 family selection in 2016, 2018, and 2020, with several field trials per cycle in Australia and Canada. Economic weights in the index promoted high GY, seed oil, protein in meal, and Phoma stem canker (blackleg) disease resistance while maintaining plant height, flowering time, oleic acid, and seed size and decreasing glucosinolate content. After factor analytic modelling of the genotype-by-environment interaction for the additive effects, the linear rate of genetic gain in GY across cycles was 0.059 or 0.087 t ha−1 y−1 (2.9% or 4.3% y−1) based on genotype scores for the first factor (f1) expressed in trait units or average predicted breeding values across environments, respectively. Both GY and yield stability, defined as the root-mean-square deviation from the regression line associated with f1, were predicted to improve in the next cycle with a low achieved mean parental coancestry (0.087). These methods achieved rapid genetic gain in GY and other traits and are predicted to improve yield stability across global spring canola environments.

Accuracy of Selection in Early Generations of Field Pea Breeding Increases by Exploiting the Information Contained in Correlated Traits

“…In this study, we achieved a predicted genetic gain in the next cycle of −3.8% in ABS, +5.0% in CST, +5.1% in SD, and +7.8% in EAngle, while controlling DTF and IL and achieving low parental coancestry of 0.12 (Table 4). Rapid and sustained genetic gain for several traits was predicted in stochastic models of S 0,1 family selection [53,54], and validated in the field over several cycles of recurrent selection in spring canola for grain yield and several other low to moderate heritability traits [55].…”

“…S 0 recurrent selection [3] may be augmented by including self progeny of parent plants at S 2 or higher selfing generations (S 2+ ). Augmenting S 0 recurrent selection with S 2+ selfs of parent plants improved connectivity of genetic relationships between and within cycles, and improved the accuracy of PBV [55]. It is also of practical and commercial value to include S 2+ selfs of parent plants in augmented S 0 recurrent selection, as this provides inbred lines ready for commercial evaluation after two cycles.…”

“…As a result, OCS was implemented in animal breeding to maximize genetic gain whilst controlling the rate of inbreeding [67]. Recently, crop breeders have adopted OCS or similar optimized selection frameworks that retain genetic variability of the population under selection to improve the long-term genetic response [1,55,58,65,[68][69][70][71]. OCS was valuable in this study to predict the outcome of MLMM on genetic response for several traits in the next generation based on an optimised mating design.…”

“…The challenge remains to implement MLMM in small plot trials grown at many sites per cycle across multiple cycles of selection, with measurements of GY, lodging, and other commercial traits. So far, this has only been achieved with univariate analyses, one trait at a time, across all sites and cycles [55].…”

AI-assisted selection of mating pairs through simulation-based optimized progeny allocation strategies in plant breeding

“…OCS aims to optimize the expected contribution vector, i.e., the number of times each individual is used as a parent for the next generation, by maximizing the genetic gains while constraining the inbreeding in the next generation. Although the efficiency of the OCS-based GS was evaluated in several simulation studies (Gorjanc et al, 2018; Cowling et al, 2023), OCS still cannot assist in identifying optimal mating pairs even for the next generation.…”