Modelling the carbon cycle of Miscanthus plantations: Existing models and the potential for their improvement

The lignocellulosic perennial grass Miscanthus has received considerable attention as a potential bioenergy crop over the last 25 years, but few commercial plantations exist globally. This is partly due to the uncertainty associated with claims that land-use change (LUC) to Miscanthus will result in both commercially viable yields and net increases in carbon (C) storage. To simulate what the effects may be after LUC to Miscanthus, six process-based models have been parameterized for Miscanthus and here we review how these models operate. This review provides an overview of the key Miscanthus soil organic matter models and then highlights what measurers can do to accelerate model development. Each model (WIMOVAC, BioCro, Agro-IBIS, DAYCENT, DNDC and ECOSSE) is capable of simulating biomass production and soil C dynamics based on specific site characteristics. Understanding the design of these models is important in model selection as well as being important for field researchers to collect the most relevant data to improve model performance. The rapid increase in models parameterized for Miscanthus is promising, but refinements and improvements are still required to ensure that model predictions are reliable and can be applied to spatial scales relevant for policy. Specific improvements, needed to ensure the models are applicable for a range of environmental conditions, come under two categories: (i) increased data generation and (ii) development of frameworks and databases to allow simulations of ranging scales. Research into nonfood bioenergy crops such as Miscanthus is relatively recent and this review highlights that there are still a number of knowledge gaps regarding Miscanthus specifically. For example, the low input requirements of Miscanthus make it particularly attractive as a bioenergy crop, but it is essential that we increase our understanding of the crop's nutrient remobilization and ability to host N-fixing organisms to derive the most accurate simulations.