From gene banks to farmer’s fields: using genomic selection to identify donors for a breeding program in rice to close the yield gap on smallholder farms

Theoretical and Applied Genetics https://doi.org/10.1007/s00122-021-03909-9 ORIGINAL ARTICLE From gene banks to farmer’s fields: using genomic selection to identify donors for a breeding program in rice to close the yield gap on smallholder farms Ryokei Tanaka1 · Sarah Tojo Mandaharisoa2 · Mbolatantely Rakotondramanana2 · Harisoa Nicole Ranaivo2 · Juan Pariasca‑Tanaka3 · Hiromi Kajiya‑Kanegae1 · Hiroyoshi Iwata1 · Matthias Wissuwa3 Received: 20 January 2021 / Accepted: 6 July 2021 © The Author(s) 2021 Abstract Key message Despite phenotyping the training set under unfavorable conditions on smallholder farms in Madagascar, we were able to successfully apply genomic prediction to select donors among gene bank accessions. Abstract Poor soil fertility and low fertilizer application rates are main reasons for the large yield gap observed for rice produced in sub-Saharan Africa. Traditional varieties that are preserved in gene banks were shown to possess traits and alleles that would improve the performance of modern variety under such low-input conditions. How to accelerate the utilization of gene bank resources in crop improvement is an unresolved question and here our objective was to test whether genomic prediction could aid in the selection of promising donors. A subset of the 3,024 sequenced accessions from the IRRI rice gene bank was phenotyped for yield and agronomic traits for two years in unfertilized farmers’ fields in Madagascar, and based on these data, a genomic prediction model was developed. This model was applied to predict the performance of the entire set of 3024 accessions, and the top predicted performers were sent to Madagascar for confirmatory trials. The prediction accuracies ranged from 0.10 to 0.30 for grain yield, from 0.25 to 0.63 for straw biomass, to 0.71 for heading date. Two accessions have subsequently been utilized as donors in rice breeding programs in Madagascar. Despite having conducted phenotypic evaluations under challenging conditions on smallholder farms, our results are encouraging as the prediction accuracy realized in on-farm experiments was in the range of accuracies achieved in on-station studies. Thus, we could provide clear empirical evidence on the value of genomic selection in identifying suitable genetic resources for crop improvement, if genotypic data are available. Introduction Communicated by Andreas Graner. * Matthias Wissuwa 1 Department of Agricultural and Environmental Biology, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1‑1‑1 Yayoi, Bunkyo, Tokyo 113‑8657, Japan 2 Rice Research Department, The National Center for Applied Research On Rural Development (FOFIFA), 101 Antananarivo, Madagascar 3 Crop, Livestock and Environment Division, Japan International Research Center for Agricultural Sciences (JIRCAS), 1‑1 Ohwashi, Tsukuba, Ibaraki 305‑8686, Japan The demand for rice in sub-Saharan Africa (SSA) is increasing steadily, outpacing local supply, and forcing many Africa countries to import increasing amounts of rice from Asia (USDA 2018). This growing shortage in local supply is due to the much lower average yields (2.3 t ha−1) achieved across Africa compared to Asia (4.8 t ha−1) (FAOSTAT 2018). The low grain yields in rice in SSA are caused by a combination of low fertilizer application rates with generally low soil fertility of the highly weathered soils which are typical throughout the region (Saito et al. 2019). Soils like the commonly encountered Oxisols are known to bind phosphorous (P) in forms that are not plant available, causing P to be the most frequent limiting nutrient in rice production in SSA (Saito et al. 2019). VanDamme et al. (2015) highlighted that a cost-efficient partial solution to the soil fertility problem in 13 Vol.:(0123456789) Theoretical and Applied Genetics SSA would be the development of varieties with improved P acquisition and utilization efficiencies. The abovementioned general trends for rice production and consumption also apply to Madagascar, the second biggest rice producer in SSA. Fertilizer applications have remained very low (Tsujimoto et al. 2019) despite standard recommendations for NPK fertilizers being in existence for decades. As a result, the national average yield remains below 3 t ha−1, whereas achievable on-farm yields can exceed 7 t ha−1 (Saito et al. 2017) and are therefore comparable to tropical regions elsewhere. This large yield gap highlights that conventional on-station breeding approaches that seek to select breeding lines with high yield potential under “ideal” high-input conditions may not produce desired results. The prevalence of traditional rice varieties throughout Madagascar (Minten and Barrett 2008) is a further sign that plant breeding has not properly addressed the needs of the mostly resource-poor smallholder farmers. It is furthermore indicative of specific adaptations to lower soil fertility being present in such traditional varieties, which were found to be more efficient in P acquisition (Mori et al. 2016) and internal P utilization (Wissuwa et al. 2015) or may even show a combination of both desirable traits (Rose et al. 2015). A future breeding program targeting to close this yield gap may move the selection process from highly fertilized breeding stations to fields representing conditions a crop may experience in farmers’ fields and should attempt to utilize any adaptive traits of traditional varieties. That traditional varieties may contain useful genes and alleles for certain traits and may therefore serve as donors to improve such traits in modern breeding populations which have been a chief reason to collect and preserve such varieties in crop gene banks. The largest collection of rice genetic resources with more than 130,000 accessions is stored in the gene bank of the International Rice Research Institute (IRRI). One potential problem of exploiting this resource is its sheer size. Phenotyping thousands of lines will require resources in terms of land and labor that few projects and institutes can manage. It is therefore of importance for the utility of such resources to have enough information associated with accessions to allow for targeted selections of smaller sub-sets of accessions for more detailed phenotypic evaluations. One invaluable step in this regard was the establishment of the publicly available SNP-Seek database (https://snp-seek.irri.org) providing sequence variants and passport data of more than 3000 rice accessions (3 K accessions) of diverse origin and genetic background (Mansueto et al. 2017). Such genotyping efforts allow for the implementation of genomic selection (Meuwissen et al. 2001) as a tool to identify promising accession from gene banks. Based on phenotypic values available for a subset of the 3 K accessions, their genotypic values can be predicted applying a statistical 13 model that establishes relations between SNP genotype and phenotype. Using this genomic s (...truncated)