Participatory plant breeding (PPB) has gained increasing importance in developing countries, but its value for market-oriented breeding programmes of countries with developed agriculture that are committed to pure line selection (as needed to comply with DUS requirements) is unknown. This study aimed to compare PPB vs. conventional plant breeding of pea (Pisum sativum L.) targeted to organic systems of Italy, exploring phenotypic and genome-enabled selection approaches. Priority values assigned on a 0–5 scale to 14 traits by 18 farmers from Northern and Central Italy and six breeders were used to define weights of farmer and breeder selection indexes. Farmers and breeders attributed outmost importance to a visual acceptability score assigned a few weeks before crop maturity on a 1–9 scale, followed in importance by grain yield and tolerance to lodging. However, breeders and farmers differed (P < 0.05) for trait importance in a few cases. Five phenotypic selection criteria (farmer selection index; breeder selection index; average of farmer and breeder selection indexes; grain yield; farmer acceptability score) were applied onto 306 lines evaluated in two researcher-managed experiments of Northern and Central Italy under organic crop management, selecting overall nine lines per criterion that were tested in four organically-managed environments of the same regions and one conventionally-managed site. The farmer selection index exhibited greater selection efficiency (+23% based on yield gains over elite commercial cultivars under organic farming) and farmer's acceptability of selected material than the breeder selection index. Breeding values based on the farmer selection index or the farmer acceptability score exhibited greater correlation with grain yields in independent environments than those from breeder selection criteria. Compared with grain yield-based selection, selection for the farmer acceptability score performed comparably in terms of yield gains, and somewhat better according to correlations of its breeding values with line grain yields in independent environments. The accuracy of genome-enabled predictions issued by a Bayesian Lasso model with 3443 SNP markers generated by genotyping-by-sequencing, estimated by averaging cross-environment correlations between predicted and observed values over two locations, was very high for the farmer acceptability score (rAc = 0.77), and high for grain yield (rAc = 0.59). Genomic selection for the farmer acceptability score ranked first in a preliminary comparison of eight genome-enabled or phenotypic selection criteria based on correlations of breeding values with grain yields in independent environments, suggesting its adoption for preliminary screening of genotype sets that are too numerous for field-based evaluation.
Farmer-participatory vs. conventional market-oriented breeding of inbred crops using phenotypic and genome-enabled approaches: A pea case study
Russi, L.;
2019
Abstract
Participatory plant breeding (PPB) has gained increasing importance in developing countries, but its value for market-oriented breeding programmes of countries with developed agriculture that are committed to pure line selection (as needed to comply with DUS requirements) is unknown. This study aimed to compare PPB vs. conventional plant breeding of pea (Pisum sativum L.) targeted to organic systems of Italy, exploring phenotypic and genome-enabled selection approaches. Priority values assigned on a 0–5 scale to 14 traits by 18 farmers from Northern and Central Italy and six breeders were used to define weights of farmer and breeder selection indexes. Farmers and breeders attributed outmost importance to a visual acceptability score assigned a few weeks before crop maturity on a 1–9 scale, followed in importance by grain yield and tolerance to lodging. However, breeders and farmers differed (P < 0.05) for trait importance in a few cases. Five phenotypic selection criteria (farmer selection index; breeder selection index; average of farmer and breeder selection indexes; grain yield; farmer acceptability score) were applied onto 306 lines evaluated in two researcher-managed experiments of Northern and Central Italy under organic crop management, selecting overall nine lines per criterion that were tested in four organically-managed environments of the same regions and one conventionally-managed site. The farmer selection index exhibited greater selection efficiency (+23% based on yield gains over elite commercial cultivars under organic farming) and farmer's acceptability of selected material than the breeder selection index. Breeding values based on the farmer selection index or the farmer acceptability score exhibited greater correlation with grain yields in independent environments than those from breeder selection criteria. Compared with grain yield-based selection, selection for the farmer acceptability score performed comparably in terms of yield gains, and somewhat better according to correlations of its breeding values with line grain yields in independent environments. The accuracy of genome-enabled predictions issued by a Bayesian Lasso model with 3443 SNP markers generated by genotyping-by-sequencing, estimated by averaging cross-environment correlations between predicted and observed values over two locations, was very high for the farmer acceptability score (rAc = 0.77), and high for grain yield (rAc = 0.59). Genomic selection for the farmer acceptability score ranked first in a preliminary comparison of eight genome-enabled or phenotypic selection criteria based on correlations of breeding values with grain yields in independent environments, suggesting its adoption for preliminary screening of genotype sets that are too numerous for field-based evaluation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
