Only a few RFLP-based studies have been conducted to evaluate genetic relationships among forage grass species. Xu and Sleper (1994) used RFLPs to study the phylogenetic relationship of tall fescue with six related grass species. Charmet et al. (1997) used RFLP and chloroplast DNA (cpDNA) for phylogenetic analysis of the Festuca-Lolium complex. The RFLP probes generated from turfgrass species were employed successfully for establishment of phylogenetic relationship in turfgrasses (Yaneshita et al., 1993).
Genetic variation within forage and turf grass species has been investigated by several groups using RAPD markers (Huff et al., 1993; Vallés et al., 1993; Wu and Lin, 1994; Sun et al., 1999; Diaby and Casler, 2003; Casler et al., 2003) and AFLP markers (Roldan-Ruiz et al., 2000; Zhang et al., 1999; Mian et al., 2002; Larson et al., 2003; Vergara and Bughrara, 2003). However, the anonymous and dominant nature of RAPD and AFLP markers limit their utility in genetic studies, particularly across species. The PCR-based codominant and portable SSR markers are more useful than dominant markers for such studies (Wang et al., 2001). The SSRs were found highly effective for differentiating cultivars of perennial ryegrass (Kubik et al., 2001) and discriminating between Lolium and Festuca grasses (Pašakinskienė et al., 2000).
Variability among grass species has been assessed using tall fescue EST-SSR markers (Mian et al., 2005). A dendrogram produced from the DICE similarity coefficients based on 1666 microsatellite bands clearly separated each species and clustered all genotypes of each species together. Two of the three Festuca species (tall fescue and meadow fescue, now classified as Lolium species, as described in Chapter 2, Craven et al., 2009, this publication) were grouped together while the third species, F. rubra, was grouped with Dactylis glomerata L. (orchardgrass). All three genera (Lolium, Festuca, and Dactylis) of the tribe Poeae formed a loose cluster (Mian et al., 2005). The phylogenetic tree based on the nucleotide substitution data of DNA sequences (obtained from EST-SSR PP NFFA150) revealed results similar to those obtained from SSR marker analysis. The genetic tree placed the three genera of the tribe Poeae and three genera of tribe Triticeae in exactly the same classes, as depicted by the SSR analysis. The tree obtained from the analysis of sequences from another SSR PP (NFFA036) was quite similar except it pushed meadow fescue close to Lolium and red fescue toward tall fescue. Clustering of two Festuca species (formerly F. arundinacea and F. pratensis, now considered Lolium spp.) from a third (F. rubra) indicated separation of broad-leaf fescues (subg. Schedonorus) from fine-leaf fescues (subg. Festuca). According to the rate of nucleotide substitution in cpDNA spacers, the split of broad-leaved and fine-leaved fescues occurred some nine million years ago (Charmet et al., 1997). Internal transcribed spacer regions (ITSs) are well suited for intrageneric studies and used for the sequence analysis in grasses (Hsiao et al., 1995). Trees of the Festuca-Lolium complex obtained from cpDNA, ribosomal DNA (rDNA), and ITSs analyses showed the same differentiation of the three major groups: (i) fine-leaved fescues, (ii) broad-leaved fescues and, (iii) ryegrasses (Charmet et al., 1997). Clustering of F. rubra with D. glomerata rather than with tall fescue and meadow fescue was detected in several independent studies (Charmet et al., 1997; Leväslaiho et al., 1987; Mian et al., 2005). Tetraploid fescue was more closely related to hexaploid tall fescue than was meadow fescue; this may be because tetraploid fescue shares two genomes (G1G2), but meadow fescue shares only one (P) genome with tall fescue (PG1G2).