This study will identify domestication loci and patterns of selection imposed upon them during the domestication process. Recent research suggested that there are some shared genes and mechanisms that control domestication traits such as pod shattering across species and that the processes of domestication are remarkably similar. Crop domestication and strong selection on quality traits are often thought to have mostly resulted in “hard sweeps” that have driven a particular allele and closely linked variant sites to fixation. However, there are many scenarios where fixation is not expected, such as cases where regular introgression from wild relatives occurs, which create a “soft sweep”. Recently it has become more evident that “soft sweeps” that still segregate in populations are more difficult to discover than “hard-sweeps”. Furthermore, loci that diverged due to post-domestication selection by breeders are most likely to be region-specific or ‘soft’. The largest, hardest sweeps across the genomes of cool season legumes can be detected using well-developed approaches with existing pipelines, including combined use of transitional (D, F, HKA) and more recent tests (e.g. iHS, SweeD, H12/H21, xp-EHH). We will perform this analysis with exome capture data from our existing lentil diversity panel (LDP) and wild accessions, using the improved long-read assemblies for variant detection.

In parallel with efforts to identify selected regions, we will use a comparative demographic approach to compare the duration and intensity of bottlenecks during domestication. We have previously approached this with the G-Phocs analysis framework that handles short read and exome capture data effectively. In chickpea this approach suggests a strong bottleneck over the past ~10,000 years that has led to a loss of well over 95% of the variation present in the ancestor of cultivated chickpea, Cicer reticulatum. Collaborator P Smýkal is currently taking a complementary approach in Pisum sativum, using DaRT-seq data and both the G-Phocs and the Diffusion Approximations for Demographic Inference (DADI) demographic analysis framework. Performing this analysis for lentil will allow us to compare the demography of domestication for three similar grain legume species from the Fertile Crescent and draw larger inferences about the timing and mode of domestication and its impact on quality traits. Combined with information on regions undergoing selective sweeps, the demographic insight will enrich our understanding of the consequences of introgression breeding and portions of the genepool that might be otherwise overlooked.