A population genomics approach to accelerating the domestication of the energy grass Miscanthus

Lignocellulosic biomass is expected to become the most important source of renewable energy in the EU, thereby significantly reducing dependence on fossil fuels and contributing to the mitigation of climate change. Because of their high productivity and low requirements for agricultural inputs, C4 grasses from the tropical genus Miscanthus are believed to have great potential as a bioenergy crop. However, Miscanthus species are essentially undomesticated, and their accelerated breeding is hampered by their primarily outcrossing mating systems and perennial life cycles. To help overcome these challenges, we are proposing to take advantage of a world-leading collection of Miscanthus germplasm (>1500 accessions) that is available at the Institute of Biological, Environmental and Rural Sciences and use high-density molecular marker data and state of the art population genomics approaches to complete three research objectives. First, we will use both model-based and assumption-free analytical approaches to characterise population genetic structure and genome-wide patterns of linkage disequilibrium in a broad collection of Miscanthus germplasm. This will provide the foundation for bridging the gap between phenotype and genotype using genome-wide association studies (GWAS) and genomic selection (i.e., phenotype prediction from a genome-wide set of molecular marker genotypes). Second, we will design and implement large-scale GWAS in multiple species of Miscanthus, elucidating the genomic architectures of important phenotypic traits. Finally, we will assess the feasibility of genomic selection in Miscanthus, potentially accelerating breeding cycles 2-3 times. In addition to Miscanthus biologists and breeders, this research will benefit other scientists from the fundamental fields of plant ecology, genetics and genomics, as well as applied breeders of other perennial crops, farmers and the general public.

The global demands for food and renewable energy are increasing at currently unsustainable rates that are expected to accelerate in the future. A major challenge for plant breeders is therefore to develop bioenergy crops that ideally (1) are highly productive, but carbon negative; (2) can be grown under a wide range of environmental conditions, including marginal lands, but with minimal agronomic inputs (e.g., fertilisers, pesticides, irrigation); (3) produce biomass that can efficiently be converted to biofuels; and (4) can be deployed very rapidly. However, most existing energy crops fail to meet at least one of these requirements. Furthermore, traditional breeding approaches, while certain to be effective, tend to be relatively slow. One way to accelerate breeding cycles is to use diagnostic molecular markers (DNA polymorphisms) to select superior plants at a juvenile age, instead of having to wait for years before direct evaluations can be made. However, an emerging consensus from studies that aim to identify such marker-trait correlations is that genetic variation for most phenotypic traits is underpinned by hundreds of DNA polymorphisms, making it impossible to cherry-pick superior germplasm based on a handful of markers. A more practical approach is therefore to use very large numbers of molecular markers, or even entire genome sequences, to predict phenotypes. This approach, known as genomic selection, is becoming increasingly affordable because of recent breakthroughs in sequencing technology and is believed to have great potential for accelerating crop development and optimisation. Our project will take advantage of an extensive germplasm collection and apply marker-assisted approaches to accelerate a world-leading breeding programme for the promising energy crop Miscanthus. To achieve this goal, we will first acquire prerequisite information on genome-wide patterns of DNA polymorphism. Then, we will characterise the genomic architectures of phenotypic traits targeted by breeders (i.e., determine the approximate number of DNA polymorphisms underlying genetic variation for each trait and quantify the phenotypic effect of each polymorphism) using state of the art statistical models. Finally, we will apply the genomic selection approach described above to Miscanthus, potentially accelerating breeding cycles 2-3 times and benefitting not only plant scientists and breeders, but also farmers and the general public.