Alfalfa (Medicago sativa L.), a perennial and outcrossing species, is a widely planted forage legume for hay, pasture and silage throughout the world. Currently, alfalfa breeding relies on recurrent phenotypic selection, but alternatives incorporating molecular marker assisted breeding could enhance genetic gain per unit time and per unit cost, and accelerate alfalfa improvement. Many major quantitative trait loci (QTL) related to agronomic traits have been identified by family-based QTL mapping, but in relatively large genomic regions. Candidate genes elucidated from model species have helped to identify some potential causal loci in alfalfa mapping and breeding population for specific traits. Recently, high throughput sequencing technologies, coupled with advanced bioinformatics tools, have been used to identify large numbers of single nucleotide polymorphisms (SNP) in alfalfa, which are being developed into markers. These markers will facilitate fine mapping of quantitative traits and genome wide association mapping of agronomic traits and further advanced breeding strategies for alfalfa, such as marker-assisted selection and genomic selection. Based on ideas from the literature, we suggest several ways to improve selection in alfalfa including (1) diversity selection and paternity testing, (2) introgression of QTL and (3) genomic selection.

Background: Alfalfa, a perennial, outcrossing species, is a widely planted forage legume producing highly nutritious biomass. Currently, improvement of cultivated alfalfa mainly relies on recurrent phenotypic selection. Marker assisted breeding strategies can enhance alfalfa improvement efforts, particularly if many genome-wide markers are available. Transcriptome sequencing enables efficient high-throughput discovery of single nucleotide polymorphism (SNP) markers for a complex polyploid species.Result: The transcriptomes of 27 alfalfa genotypes, including elite breeding genotypes, parents of mapping populations, and unimproved wild genotypes, were sequenced using an Illumina Genome Analyzer IIx.De novoassembly of quality-filtered 72-bp reads generated 25,183 contigs with a total length of 26.8 Mbp and an average length of 1,065 bp, with an average read depth of 55.9-fold for each genotype. Overall, 21,954 (87.2%) of the 25,183 contigs represented 14,878 unique protein accessions. Gene ontology (GO) analysis suggested that a broad diversity of genes was represented in the resulting sequences. The realignment of individual reads to the contigs enabled the detection of 872,384 SNPs and 31,760 InDels. High resolution melting (HRM) analysis was used to validate 91% of 192 putative SNPs identified by sequencing. Both allelic variants at about 95% of SNP sites identified among five wild, unimproved genotypes are still present in cultivated alfalfa, and all four US breeding programs also contain a high proportion of these SNPs. Thus, little evidence exists among this dataset for loss of significant DNA sequence diversity from either domestication or breeding of alfalfa. Structure analysis indicated that individuals from the subspeciesfalcata, the diploid subspeciescaerulea, and the tetraploid subspeciessativa(cultivated tetraploid alfalfa) were clearly separated.Conclusion: We used transcriptome sequencing to discover large numbers of SNPs segregating in elite breeding populations of alfalfa. Little loss of SNP diversity was evident between unimproved and elite alfalfa germplasm. The EST and SNP markers generated from this study are publicly available at the Legume Information System (http://medsa.comparative-legumes.org/) and can contribute to future alfalfa research and breeding applications.