Among the fertile germplasm, accessions should be selected as parents to be used in crosses for breeding purposes or to create mapping populations for genetic studies. As a Miscanthus crop takes several years to mature, genetic markers and phenotypic methods need to be developed to speed up the breeding process.

4.3.2.1 Tools for Genetic Studies and Breeding

The polyploid nature and the relative large size of the Miscanthus genome complicate genetic analyses. Using flow cytometry and stomatal cell analyses, Rayburn et al. [131] found M. x giganteus had a genome size of 7.0 pg (6.8 Gb, the number of DNA base pairs per nucleus being assumed 0.965 x 109 bp per pg by the author) while Miscanthus sinensis and Miscanthus sacchariflorus had genome sizes of 5.5 pg (5.3 Gb) and 4.5 pg (4.3 Gb) respectively (Table 4.5). It is clear that there are many gaps that require further

investigation. However, further studies will be facilitated by the use of plants such as sugarcane, sorghum and maize, which are likely to be good models for genomics and breeding issues in Miscanthus.

Detailed DNA mapping and sequencing studies in plants related to Miscanthus will pro­vide relevant genetic tools and information. For example, Sorghum Bicolor is diploid and, with a relatively small genome of about 730 Mb, has been completely sequenced [132]. Although sugarcane is related to Miscanthus [11], its genome is much more complicated due to its very high degree of polyploidy (about 12x for modern cultivars, Le Cunff et al. [133]). The monoploid genome size for S. officinarum (x = 10) is about 926 Mb while that of S. spontaneum (x = 8) is about 760 Mb (Butterfield et al. [132]). Maize is less related to Miscanthus than sugarcane and sorghum but its genome has been fully sequenced [134]. Syntenic regions or candidate gene sequences can be expected and exploited for compar­ative genetic studies. From the conserved syntenic regions, markers can be developed in Miscanthus and related to traits of interest for marker-assisted selection.

A wide diversity of molecular markers are available from plants related to Miscanthus but their transferability for use in Miscanthus needs to be determined. First comparisons are promising, however, with Hernandez et al. [136] showing that 75% of the maize microsatel­lites tested gave highly reproducible amplification with Miscanthus DNA. More recently, Swaminathan et al. [137] showed that sorghum could be used as a reference genome sequence for Andropogoneae grasses. In a survey of the complex Miscanthus x giganteus genome using 454 pyrosequencing of genomic DNA and Illumina sequencing-by-synthesis of small RNA, Swaminathan et al. [137] found that the coding fraction of the Miscant — hus x giganteus genome had a high level of sequence identity to that of other grasses (sorghum, maize and rice). In addition, Kim et al. [138] designed SSRs from sugarcane expressed sequence tags (ESTs) and in applying these to a Miscanthus mapping population succeeded in generating EST-SSR-based genetic maps of Miscanthus.

Genetic linkage maps offer an efficient tool in the study of the inheritance of quanti­tative traits. Most Miscanthus species are self-incompatible, resulting in a high level of heterozygosity from outcrossing. Grattapaglia and Sederoff [139] proposed a two-way pseudo-testcross model for the genetic mapping of highly heterozygous organisms.

Several maps are available for marker-assisted studies in Miscanthus. The first genetic map of Miscanthus was constructed with this pseudo-testcross strategy using intraspecific hybrids from a cross between two Miscanthus sinensis clones [10]. 383 RAPD markers were developed for this map but a higher density of molecular markers was needed due to the high number of linkage groups (28) relative to the basic chromosome number (x = 19). This map had a total length of 1074.5 cM. A decade later, Kim et al. [138] developed a genetic map with highly heterozygous individuals being interspecific hybrids from a controlled cross between heterozygous single plants of M. sacchariflorus Robustus (2n = 2x = 38) and M. sinensis (2n = 2x = 38). Their map used cDNA-derived SSR loci and comprised 23 linkage groups with 303 markers and was 2238.3 cM in total length. Ma et al. [140] created a high-resolution genetic map of Miscanthus sinensis using genome sequencing and comprising 3745 SNP markers spanning cM on 19 linkage groups with a 0.64 cM average resolution.

Miscanthus linkage groups of the map developed by Kim et al. [138] were aligned suc­cessfully to the Sorghum chromosomes. A duplication of the whole genome was produced and corresponds to the Miscanthus lineage after the divergence of subtribes Sorghinae and Saccharinae [138]. Comparative genomics analyses of their map to the genomes of sorghum, maize, rice and Brachypodium distachyon [140] indicated that sorghum had the closest syntenic relationship to Miscanthus. This validates the use of sorghum as a model for the genomics of Miscanthus.

Breeding programs will be directly guided in the future by the genome sequencing of Miscanthus x giganteus and its close relatives, to capture, for example, the genes of interest present in these species. It is noticeable that sequencing efforts of four Miscanthus species (M. x giganteus, M. sinensis, M. sacchariflorus, and M. floridulus) are ongoing along with the creation of genomic resources by the Energy Biosciences Institute (http://www. energybiosciencesinstitute. org/) and by the Joint Genome Institute (http://www. jgi. doe. gov/).