Recent progress in alfalfa (Medicago sativa L.) genomics and genomic selection

Abstract

Alfalfa (M. sativa L.) is a highly valuable forage crop, providing >58 Mt of hay, silage, and pasture each year in the United States. As alfalfa is an outcrossing autotetraploid crop, however, breeding for enhanced agronomic traits is challenging and progress has historically not been rapid. Methods that make use of genotypic information and statistical models to generate a genomic estimated breeding value (GEBV) for each plant at a young age hold a great deal of promise to accelerate breeding gains. An emerging genomic breeding pipeline employs SNP chips or genotyping-by-sequencing (GBS) to identify SNP markers in a training population, followed by the use of a statistical model to find associations between the discovered SNPs and traits of interest, followed by genomic selection (GS), a breeding program utilizing the trained model to predict breeding values and making selections based on the estimated breeding value (EBV). Much work has been done in recent years in all of these areas, to generate marker sets and discover SNPs associated with desirable traits, and the application of these technologies in alfalfa breeding programs is under way. However, GBS/GWAS/GS is still a new breeding paradigm, and work is ongoing to evaluate different models, software, and methods for use in such programs. In this review, we look at the progress of alfalfa genomics over the past half-decade, and review work comparing models and methods relevant to this new type of breeding strategy.