Winter Hardiness of Miscanthus (III): Genome‐Wide Association and Genomic Prediction for Overwintering Ability in Miscanthus sinensis


CABBI Theme: Feedstock Production

Keywords: Biomass Analytics, Genomics



Dong, H., Clark, L.V., Lipka, A.E., Brummer, J.E., Głowacka, K., Hall, M.C., Heo, K., Jin, X., Peng, J., Yamada, T., Ghimire, B.K., Yoo, J.H., Yu, C.Y., Zhao, H., Long, S.P., Sacks, E.J. March 22, 2019. “Winter Hardiness of Miscanthus (III): Genome‐Wide Association and Genomic Prediction for Overwintering Ability in Miscanthus sinensis.” GCB Bioenergy. DOI: 10.1111/gcbb.12615.


Box plots of overwintering ability during the first winter (establishment year) and second winter for a Miscanthus sinensis diversity panel at each of five field trial locations.


Overwintering ability is an important selection criterion for Miscanthus breeding in temperate regions. Insufficient overwintering ability of the currently leading Miscanthus biomass cultivar, M. × giganteus (M×g) ‘1993–1780′, in regions where average annual minimum temperatures are −26.1°C (USDA hardiness zone 5) or lower poses a pressing need to develop new cultivars with superior cold tolerance. To facilitate breeding of Miscanthus, this study characterized phenotypic and genetic variation of overwintering ability in an M. sinensis germplasm panel consisting of 564 accessions, evaluated in field trials at three locations in North America and two in Asia. Genome‐wide association (GWA) and genomic prediction analyses were performed. The Korea/N China M. sinensis genetic group is a valuable gene pool for cold tolerance. The Yangtze‐Qinling, Southern Japan, and Northern Japan genetic groups were also potential sources of cold tolerance. A total of 73 marker–trait associations were detected for overwintering ability. Estimated breeding value for overwintering ability based on these 73 markers could explain 55% of the variation for first winter overwintering ability among M. sinensis. Average genomic prediction ability for overwintering ability across 50 fivefold cross‐validations was high (~0.73) after accounting for population structure. Common genomic regions for overwintering ability were detected by GWA analyses and a previous parallel QTL mapping study using three interconnected biparental F1 populations. One QTL on Miscanthus LG 8 encompassed five GWA hits and a known cold‐responsive gene, COR47. The other overwintering ability QTL on Miscanthus LG 11 contained two GWA hits and three known cold stress‐related genes, carboxylesterase 13 (CEX13), WRKY2 transcription factor, and cold shock domain (CSDP1). Miscanthus accessions collected from high latitude locations with cold winters had higher rates of overwintering, and more alleles for overwintering, than accessions collected from southern locations with mild winters.



Plant Materials and Trial Locations (396 KB)

Genotypic Effects and Predicted OWA Values (68 KB)

Variance from the Statistical Model for each Genomic Prediction Analysis (13 KB)

Significance Testing Results (1.5 MB)