Skip to main content
eScholarship
Open Access Publications from the University of California

Diversity and population structure of northern switchgrass as revealed through exome capture sequencing.

  • Author(s): Evans, Joseph
  • Crisovan, Emily
  • Barry, Kerrie
  • Daum, Chris
  • Jenkins, Jerry
  • Kunde-Ramamoorthy, Govindarajan
  • Nandety, Aruna
  • Ngan, Chew Yee
  • Vaillancourt, Brieanne
  • Wei, Chia-Lin
  • Schmutz, Jeremy
  • Kaeppler, Shawn M
  • Casler, Michael D
  • Buell, Carol Robin
  • et al.

Published Web Location

https://doi.org/10.1111/tpj.13041
Abstract

Panicum virgatum L. (switchgrass) is a polyploid, perennial grass species that is native to North America, and is being developed as a future biofuel feedstock crop. Switchgrass is present primarily in two ecotypes: a northern upland ecotype, composed of tetraploid and octoploid accessions, and a southern lowland ecotype, composed of primarily tetraploid accessions. We employed high-coverage exome capture sequencing (~2.4 Tb) to genotype 537 individuals from 45 upland and 21 lowland populations. From these data, we identified ~27 million single-nucleotide polymorphisms (SNPs), of which 1 590 653 high-confidence SNPs were used in downstream analyses of diversity within and between the populations. From the 66 populations, we identified five primary population groups within the upland and lowland ecotypes, a result that was further supported through genetic distance analysis. We identified conserved, ecotype-restricted, non-synonymous SNPs that are predicted to affect the protein function of CONSTANS (CO) and EARLY HEADING DATE 1 (EHD1), key genes involved in flowering, which may contribute to the phenotypic differences between the two ecotypes. We also identified, relative to the near-reference Kanlow population, 17 228 genes present in more copies than in the reference genome (up-CNVs), 112 630 genes present in fewer copies than in the reference genome (down-CNVs) and 14 430 presence/absence variants (PAVs), affecting a total of 9979 genes, including two upland-specific CNV clusters. In total, 45 719 genes were affected by an SNP, CNV, or PAV across the panel, providing a firm foundation to identify functional variation associated with phenotypic traits of interest for biofuel feedstock production.

Main Content
Current View