Using cross-country datasets for association mapping in Arachis hypogaea L.

Groundnut (Arachis hypogaea L.) is one of the most important climate-resilient oil crops in sub-Saharan Africa. There is a significant yield gap for groundnut in Africa because of poor soil fertility, low agricultural inputs, biotic and abiotic stresses. Cross-country evaluations of promising breeding lines can facilitate the varietal development process. The objective of our study was to characterize popular test environments in Uganda (Serere and Nakabango) and Malawi (Chitala and Chitedze) and identify genotypes with stable superior yields for potential future release. Phenotypic data were generated for 192 breeding lines for yield-related traits, while genotypic data were generated using skim-sequencing. We observed significant variation (p < 0.001; p < 0.01; p < 0.05) across genotypes for all yield-related traits: days to flowering (DTF), pod yield (PY), shelling percentage, 100-seed weight, and grain yield within and across locations. Nakabango, Chitedze, and Serere were clustered as one mega-environment with the top five most stable genotypes being ICGV-SM 01709, ICGV-SM 15575, ICGV-SM 90704, ICGV-SM 15576, and ICGV-SM 03710, all Virginia types. Population structure analysis clustered the genotypes in three distinct groups based on market classes. Eight and four marker-trait associations (MTAs) were recorded for DTF and PY, respectively. One of the MTAs for DTF was co-localized within an uncharacterized protein on chromosome 13, while another one (TRv2Chr.11_3476885) was consistent across the two countries. Future studies will need to further characterize the candidate genes as well as confirm the stability of superior genotypes across seasons before recommending them for release. Plain Language Summary Most countries in eastern and southern Africa derive their groundnut breeding lines from International Crops Research Institute for the Semi-Arid Tropics breeding program based in Malawi. In some cases, the same genotype is released in several countries under different names. However, the evaluation of the genotypes is often taken independently in each of the countries, leading to duplication of work. A more cost-effective method is to identify similar environments across different countries and evaluate the same genotypes across such environments. In this study, we evaluated 192 groundnut genotypes across four environments, two each from Uganda and Malawi. Additive main effects and multiplicative interaction analysis clustered the Ugandan sites and Chitedze in one mega-environment, implying that future evaluations could take advantage of such environments towards varietal releases. We also used the same data to detect marker-trait associations across the different locations for agronomic traits. Our results revealed more consistent results within Uganda than Malawi.