Empirical comparison of genomic selection to phenotypic selection for biomass yield of switchgrass

Authors: Tilhou, Neal; DOKYOUNG, LEE - University Of Illinois; RAMSTEIN, GUILLAUME - Aarhus University; POUDEL, HARI - Agriculture And Agri-Food Canada; Edme, Serge; Casler, Michael

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/17/2024
Publication Date: N/A
Citation: N/A

Interpretive Summary: Switchgrass has a long history as a valuable forage and pasture crop, but is now undergoing research and development as a biomass crop for conversion to biofuels. To accomplish this, new varieties with higher biomass production are required for many regions of the eastern USA. Molecular breeding methods are now readily available for increasing the rate of gain for biomass production of switchgrass. This study compared three generations of molecular breeding, specifically genomic selection which uses information from the entire DNA genome, to one generation of classical breeding. Two generations of genomic selection led to significant increases in biomass production while classical breeding failed to result in any significant increases. These results validate the use of advanced DNA technologies and breeding methods for accelerate the breeding of new and sustainable biomass varieties of switchgrass. These results will be of value to many switchgrass researchers.

Technical Abstract: Switchgrass (Panicum virgatum L.) is one of several grass species being bred for use as a biomass crop to support the biofuel industry. Increases in biomass yield are imperative to ensure that crops such as switchgrass can sustainably meet the needs of this industry. Genomic selection is one strategy that can accelerate breeding gains for complex traits such as biomass yield. The goal of this study was to conduct three cycles of genomic selection in a previously trained Liberty switchgrass population and to compare that to one cycle of phenotypic selection, both of which required three years to complete. The advanced lines were tested across five locations and three hardiness zones in the Central USA using a randomized complete block designs with four replicates. There were strong genotype x location interactions, but the first two generations of genomic selection were superior to Liberty at four of the five evaluation locations. Conversely, phenotypic selection failed to result in significant gains in biomass yield for any of the five evaluation locations. Based on these results from Liberty switchgrass, genomic selection methods are expected to at least double the rates of gain in biomass yield relative to previous estimates using phenotypic selection methods.