Patterns of genetic variation in a prairie wildflower, silphium integrifolium, suggest a non-prairie origin and untapped variation available for improved breeding

Authors: RADUSKI, ANDREW - University Of Minnesota; HERMAN, ADAM - University Of Minnesota; POGODA, CLOE - University Of Colorado; Dorn, Kevin; VAN TASSEL, DAVID - The Land Institute; KANE, NOLAN - University Of Colorado; BRANDVAIN, YANIV - University Of Minnesota

Submitted to: American Journal of Botany
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/17/2020
Publication Date: 1/29/2021
Citation: Raduski, A., Herman, A., Pogoda, C., Dorn, K.M., Van Tassel, D., Kane, N., Brandvain, Y. 2021. Patterns of genetic variation in a prairie wildflower, silphium integrifolium, suggest a non-prairie origin and untapped variation available for improved breeding. American Journal of Botany. 108(1):145-158. https://doi.org/10.1002/ajb2.1603.
DOI: https://doi.org/10.1002/ajb2.1603

Interpretive Summary: Silphium integreifolium is a perennial flower commonly found in American prairies and is a relative of sunflower that is being developed as a new oilseed crop. The breeding programs developing Silphium are based upon seed collected in a limited geographic region. In this study, we applied population genetic studies to this species to improve the breeding efforts by characterizing the genetic structure and genetic diversity of the existing germplasm base, as well as to determine geographic origins of the species. We determined that the base germplasm used in the breeding programs came from the least genetically diverse geographic region. This finding highlights the need for additional collections and integration of germplasm into Silphium breeding programs from outside of the current geographic range to increase the genetic diversity available to breeders. The genetic structure of the germplasm evaluated in this study point to specific genetic targets that have driven the adaptation on the spread of Silphium across its geographic range, as well as guides future seed collection efforts.

Technical Abstract: Understanding the relationship between genetic structure and geography provides information about a species' evolutionary history and can be useful to breeders interested in de novo domestication. Silphium integrifolium is an iconic perennial American prairie wildflower that is targeted for domestication as an oilseed crop. Germplasm in the existing breeding program is derived from accessions collected in restricted geographic regions. We present the first application of population genetic data in this species to address the following goals: (1) improve existing breeding programs by characterizing genetic structure and (2) identify the species geographic origin and potential targets and drivers of natural selection as the range has expanded. We developed a reference transcriptome to use as a genotyping reference for dozens of samples from throughout the species range. Population genetic analyses were used to describe the distribution of genetic variation and demographic inference modeling was used to characterize potential processes that have shaped variation. Divergence outlier scans for selection and associations with environmental variables were used to identify loci linked to putative targets and drivers of natural selection. Genetic variation partitions samples into three geographic clusters. Patterns of variation and demographic modeling suggests that the geographic origin of the species is in the American southeast. Breeding program accessions are from the region with lowest observed genetic variation. This iconic prairie species did not originate within the modern prairie. Breeding programs can be improved by including accessions from outside of the germplasm founding region, which has relatively little genetic variation compared to the south and east. The observed geographic structuring of variation coupled with the identified targets and environmental drivers of adaptation during range expansion can guide future collecting efforts towards wild populations with beneficial agronomic traits.