Impact of early genomic prediction for recurrent selection in an upland rice synthetic population

Abstract

Population breeding through recurrent selection is based on the repetition of evaluation and recombination among best-selected individuals. In this type of breeding strategy, early evaluation of selection candidates combined with genomic prediction could substantially shorten the breeding cycle length, thus increasing the rate of genetic gain. The objective of the present study was to optimize early genomic prediction in an upland rice (Oryza sativa L.) synthetic population improved through recurrent selection via shuttle breeding in two sites. To this end, we used genomic prediction on 334 S0 genotypes evaluated with early generation progeny testing (S0:2 and S0:3) across two sites. Four traits were measured (plant height, days to flowering, grain yield and grain zinc concentration) and the predictive ability was assessed for the target site. For days to flowering and plant height, which correlate well among sites (0.51–0.62), an increase of up to 0.4 in predictive ability was observed when the model was trained using the two sites. For grain zinc concentration, adding the phenotype of the predicted lines in the non-target site to the model improved the predictive ability (0.51 with two-site and 0.31 with single-site model), while for grain yield the gain was less (0.42 with two-site and 0.35 with single-site calibration). Through these results, we found a good opportunity to optimize the genomic recurrent selection scheme and maximize the use of resources by performing early progeny testing in two sites for traits with best expression and/or relevance in each specific environment.

Genomic prediction of zinc-biofortification potential in rice gene bank accessions

“…If successful, this would allow for a very efficient search of potential new donors for high grain Zn concentrations. The prediction accuracy of 0.51 achieved in this study is similar to the PA ranging from 0.33 to 0.69 in wheat (Velu et al 2016 ), of 0.43–0.73 reported for maize (Mageto et al 2020 ) and of 0.51 in upland rice (Baertschi et al 2021 ). The present study was conducted in two low-input farmers’ fields rather than under the more controlled conditions one may encounter on research farms and that the PA achieved here is comparable to PAs reported from research farms is further suggestive of GP being a suitable approach for identifying potential donors from gene banks.…”

“…In addition, the heritability for grain Zn concentrations (H 2 = 0.79) was larger than for GY (H 2 = 0.40). High heritability (> 0.70) for grain Zn has also been reported elsewhere (Swamy et al 2016 , 2018 ; Baertschi et al 2021 ).…”

“…Such robustness across environments would offer options to further economize resources through sparse testing of only part of the entire training set at each site or environment. Baertschi et al ( 2021 ) suggested optimizing the GP scheme by evaluating small training sets and using phenotypic correlation between sites to calibrate the model, and in their case the phenotypic correlation for grain zinc concentration between sites ( r 2 > 0.41) was similar to what was achieved here.…”

“…Genomic Prediction (GP; Meuwissen et al 2001 ) for mineral content has already proven efficient in maize and wheat where the predictive ability (PA) for grain Zn content was between 0.43 and 0.73 in maize (Mageto et al 2020 ) and between 0.33 and 0.69 for wheat (Velu et al 2016 ) depending on the population and environment chosen. A similar PA of 0.51 was achieved for grain Zn concentration improvements in a rice synthetic population managed through recurrent selection when multi-site data were considered for the calibration model (Baertschi et al 2021 ).…”

“…Whereas several QTL and GWAS studies have been reported in the literature this is only the second report applying GP for grain Zn concentrations in rice. In contrast to the work by Baertschi et al ( 2021 ) that focused on assessing the potential of GP models for early selection of families to improve upland rice synthetic populations, the aim of the GP approach taken here is to predict grain Zn concentrations of gene bank accessions in lowland rice. If successful, this would allow for a very efficient search of potential new donors for high grain Zn concentrations.…”