A global assembly of landrace oat (Avena sativa L.) accessions is a discovery resource for adaptive variation, association mapping, and trait deployment

Abstract

Crop adaptation to environmental change will require genetic resources that are different from those currently deployed. The rapid global shift to both warmer temperatures and unpredictable atmospheric events must be considered in developing new breeding populations for local environments. Oats (Avena spp.) are annual grasses that represent a diversity of species and ploidy levels. The most notable, spring oat (Avena sativa L.), is a heart-healthy and gluten-free cereal crop that is grown worldwide as a source of food, feed, and cosmetics products. In the past decade, global oat production has been increasingly challenged by environmental stress and its economic value has suffered due to competition with other high-value grain crops. Although genomic resources are growing for spring oat, there is limited information about the landraces that served as founders to modern varieties. To improve knowledge of adaptive genetic variation and phenotypic diversity of spring oat founders, a set of 758 global A. sativa landrace accessions from the USDA-ARS National Small Grains Collection was investigated, herein dubbed the "Oat Landrace Diversity (OLD) Panel." High-depth genotyping-by-sequencing was conducted to assess genetic diversity, perform genome-wide association mapping for environmental variables, and provide insight into whether quantitative trait loci identified in the OLD Panel have been deployed in modern cultivar populations. Finally, we discuss the importance of leveraging genetic variation attributable to environmental adaptation to reinforce plant breeding programs from ecological instability.

Keywords: GWAS; QTL; adaptation; founder population; functional traits; landrace; oats.

Fig. 1.

Global distribution of OLD Panel accessions.

Fig. 2.

Pairwise Pearson correlation’s of environmental variables. a) A correlation heatmap of soil variables, where the number after variables represents six soil depths: 0–5 cm (1), 5–15 cm (2), 15–30 cm (3), 30–60 cm (4), 60–100 cm (5), and 100–200 cm (6). b) correlation network is depicted and c) PCA bi-plot of all environmental variables.

Fig. 3.

Genetic diversity of OLD Panel using a) DAPC clustering, b) PCA (PCs 1–3), and c) fastSTRUCTURE ($k=7$, 10, 12, 15).

Fig. 4.

Environmental GWAS in the OLD Panel. The top hit for each environmental variable (points) per 5 Mb chromosomal bin is shown. Variables are colored according to categories described in the legend. Unplaced OT3098v2 scaffold “ChrUn” (artificially merged contigs) is shown below chr1A. Full GWAS results are in Supplementary File S4, Table A.