Assessment of genomic prediction models for winter survival on lowland switchgrass using multiple populations

Authors: POUDEL, HARI - University Of Wisconsin; SANCIANGCO, MILLICENT - Michigan State University; KAEPPLER, SHAWN - University Of Wisconsin; BUELL, C - Michigan State University; Casler, Michael

Submitted to: The Plant Genome
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/10/2019
Publication Date: N/A
Citation: N/A

Interpretive Summary: Switchgrass is undergoing development as a biomass crop to support the bioenergy industry. Late flowering time and increased freezing tolerance are two key traits to aid in developing high-biomass varieties for use in the northern USA. Development of freezing tolerant switchgrass is extremely difficult, because of the need for harsh winters that kill between 2 and 10% of the plants - too much or too little mortality fails to achieve the desired objective. Genomic selection would be desirable, because it would be based on field observations of mortality, but would use DNA sequence data of additional plants as the basis for selection, reducing the need to rely on questionable winter conditions. Prediction accuracies between 0.4 and 0.8 indicated that genomic selection should be sufficiently effective that switchgrass breeders can rely on this technology to make further improvements in survivorship and broaden that adaptation of this important biofuel species.

Technical Abstract: Genomic selection (GS) is emerging as a powerful tool in plant breeding. Implementation of GS could potentially enhance switchgrass breeding for winter survival by reducing generation time while eliminating the dependence on weather. Of the two ecotypes of switchgrass, the lowland ecotype has generated considerable interest because of its higher biomass yield and late flowering characteristics compared to the upland ecotype. However, lowland ecotypes planted in northern latitudes exhibit very low winter survival. The objectives of this study were to assess the potential of GS for winter survival in lowland switchgrass by combining multiple populations in the training set and applying the selected model in two independent testing datasets for validation. Validation was conducted using (1) indirect indicators of winter adaptation based on geographic and climatic variables of accessions from different source locations and (2) winter survival estimates of the phenotype. Maker data were generated using exome capture sequencing and phenotypic data were estimated from winter survival ratings in field experiments. In this study, we assessed the prediction accuracy of five GS prediction procedures namely GBLUP, Bayesian A, Bayesian B, Bayesian Lasso and random forest implemented to the different combinations of individuals within the training set. None of the other genomic prediction models resulted in any improvement over the GBLUP model. The prediction accuracies were significantly higher when the training dataset comprising all populations was used in 5-fold cross validation but its application was not useful in the independent validation dataset. Nevertheless, modeling for population heterogeneity improved the prediction accuracy to some extent but the genetic relationship between the training and validation populations was found to be more influential. The predicted winter survival of lowland switchgrass indicated latitudinal and longitudinal variability, with the northeast USA the region for most cold tolerant lowland populations. Our results suggested that GS could provide valuable opportunities for improving winter survival and accelerate the lowland switchgrass breeding programs towards the development of cold tolerant cultivars suitable for northern latitudes.