Genomic prediction of landscape-scale genotype-by-environment variance for plant biomass and yield surrogate traits in switchgrass (Panicum virgatum)

Authors: Tilhou, Neal; BONNETTE, JASON - University Of Texas At Austin; BOE, ARVID - South Dakota State University; Fay, Philip; FRITSCHI, FELIX - University Of Missouri; Mitchell, Robert - Rob; ROUQUETTE, FRANCIS - Texas A&M University; WU, YANQUI - Oklahoma State University; MATAMALA, ROSER - Argonne National Laboratory; MAHER, SHELLEY - Natural Resources Conservation Service (NRCS, USDA); JASTROW, JULIE - Argonne National Laboratory; JUENGER, THOMAS - University Of Texas At Austin; LOWRY, DAVID - Michigan State University

Submitted to: G3, Genes/Genomes/Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/24/2024
Publication Date: N/A
Citation: N/A

Interpretive Summary: Switchgrass is a potential source for bioenergy or carbon sequestration. Industry adoption of this new crop could be improved with higher yielding cultivars, but switchgrass breeding is slow and complex since different populations are adapted to different growing conditions. This study evaluated 630 switchgrass individuals across ten different environments for three years. This study was able to accurately predict performance of unevaluated individuals in new environments. This was done using a combination of: integrating genetic data (genomic prediction), sharing information among similar growing environments (mega-environments) and using multiple traits which are inexpensive replacements for biomass yield measurements (plant height and plant flowering time). This work has the potential to reduce the cost of creating new switchgrass cultivars, but further work is needed to validate these observations at a commercial scale.

Technical Abstract: Switchgrass is a potential source for bioenergy or carbon sequestration. Industry adoption of this new crop could be improved with higher yielding cultivars, but currently new switchgrass cultivars require 5-10 years to develop. Also, different switchgrass populations are adapted to a wide range of growing conditions, requiring expensive multi-site evaluations of yield. Genomic prediction models applied to multisite studies could generate new cultivars in <5 years. Application of genomic prediction models to genetically diverse plantings at ten sites spanning the latitudinal range of switchgrass revealed that a few simple traits like plant height and flowering time measured at a few sites can successfully predict biomass observations across the latitudinal gradient of switchgrass. This work has the potential to reduce the cost of creating new switchgrass cultivars, but further work is needed to validate these observations at a commercial scale. These results are valuable for producers and professionals interested in new opportunities to create low input biomass feedstocks.