Diversity and genome-wide association mapping of farmer preferred traits yam genetic resources in Uganda.

Although yam is a very important crop with the potential to address future food insecurity resulting from challenges due to climate change and increasing world population, its potential has largely remained unexploited in East Africa. Over the years, yam production in Uganda has been at a subsistence level, with farmers relying on unimproved landraces that are low-yielding and susceptible disease. In terms of research attention, yam is an “orphan” crop within the East African region despite its enormous potential. Currently, there is no effective yam breeding program in Uganda which hinders systematic improvement of the crop. For efficient and effective yam improvement to take place in Uganda, there is a need for a comprehensive assessment of existing local and introduced germplasm, farmers’ trait preferences and yam production challenges. In addition, crop improvement requires the utilization of existing molecular tools and genetic resources available worldwide. This study, therefore, sought to i) identify farmers’ cropping systems, production constraints and farmers' preferences for trait improvement, ii) assess phenotypic and genotypic diversity among locally cultivated and introduced yam germplasm, iii) identify genomic regions associated with useful agronomic traits using single nucleotide polymorphic markers and iv) assess the effect of genotype by environment interaction of useful agronomic traits in selected yam genotypes. The study used farmers' participatory approach and systematic prioritization of user-demanded attributes designed in 1000Minds to assess information from 540 farmers in six districts across three different agroecological zones in Uganda. The study observed that farmers' main constraints to yam cultivation were the inability to obtain good planting materials and the inaccessibility of improved varieties. However, farmers preferred high-yielding and mealy cooked tubers with high disease resistance. This study used 28 phenotypic traits and 4,957 single nucleotide polymorphic (SNP) markers to assess diversity in a collection of 291 (local and introduced) yam germplasms. Based on phenotypic traits, tuber characteristics differed between the locally cultivated and the introduced germplasm. Other phenotypic traits such as leaf and plant characteristics showed minimal diversity among the germplasm. Single nucleotide polymorphic markers used in this study were generated using diversity array technology and were informative (Polymorphic Information Content ranging from 0.238 to 0.288) and useful for diversity analysis. The study revealed a moderate molecular diversity with mean gene diversity of 0.280 across the Dioscorea rotundata yam genome. Cluster and discriminant analysis on SNP xx markers indicted that the studied accessions can be grouped into two major groups. The population structure analysis using STRUCTURE program also confirmed the presence two major populations with expected heterozygosity of 0.12 and 0.15 for POPI and POPII, respectively, with high Fst values of 0.78 and 0.75. From the study, marker-trait association analysis identified 27 significant SNPs associated with studied traits. Eight SNPs were significantly associated with inflorescence-related traits, 5 SNPs were linked to disease-related traits and 14 were SNPs related to yield traits. Four quantitative trait loci were found to be linked with dry matter content, with two of them exhibiting substantial variation among traits. The multi-environment studies using a randomized complete block design with three replications, revealed a highly significant (p ≤ 0.001) genotype by environment interaction for all traits studied. A highly significant negative correlation (r = −0.85, p < 0.001) was observed between yam mosaic virus and total yield of yam. This study generally provided insight into yam utilization, production constraints, and farmers' varietal preferences in Uganda, which will guide genetic resource conservation and plant breeding interventions for yams in the country. There was significant phenotypic variability within the yam genotypes that can be used for genetic improvement. The study, therefore, showed that yam improvement in the sub-regions is possible and the outcome will constitute a foundation for the genetic improvement of yams in Uganda. This study will be useful with regard to the identification and conservation of elite genotypes for future genetic improvement of yam. More in-depth molecular and biochemical studies to further understand the diversity are recommended.