Most policies for limiting global climate change to less than 2 °C incorporate significant biofuel usage to meet society’s energy needs. However, traditional corn-based biofuel production fails to achieve carbon neutrality, in part, due to soil carbon (C) losses. By contrast, alternative feedstocks (e.g., miscanthus, sorghum) can potentially sequester C in both biomass and soils. Given initiatives to scale-up biofuel production, there is a critical need to improve predictive understanding of how trait differences between feedstocks impact soil C cycling. A key knowledge gap lies in how models represent differences between feedstocks in the strength of interactions between roots and microbes in the rhizosphere. To address this gap, we adapted the Fixation and Uptake of Nitrogen – Carbon, Organisms, and Rhizosphere Processes in the Soil Environment (FUN-CORPSE) model to bioenergy systems. FUN-CORPSE dynamically predicts the C investment by plants to gain nutrients with the cascading impacts on rhizosphere microbes, soil organic C (SOC) cycling and the distribution of C in protected and unprotected pools. We validated the model against field trials of natural prairie, corn, sorghum, and miscanthus from the University of Illinois Energy Farm. After validation, we performed model experiments to assess the impacts of rhizosphere processes and management practices.
Results/Conclusions
The model accurately captured differences between feedstocks in soil C pools in the field. Following prairie spin up, soil C storage in corn systems was lower than in perennial bioenergy crop systems, in which a significant fraction of soil C lost from the transition of prairie to row-crop agriculture was restored. This resulted from the larger residue remaining following harvest, differences in belowground C allocation, and litter chemistry of corn compared to perennials. SOC predictions were highly sensitive to rhizosphere processes and crop management. Shutting down the flow of plant C to the rhizosphere resulted in greater soil C increases in plants with relatively greater root biomass. Moreover, soil C storage responded more to residue removal and rhizosphere size as compared to tillage and fertilization, such that decreasing residue removal and increasing rhizosphere size both increased soil C storage. Collectively, our model experiments show that dynamically modeling plant-microbial interactions leads to divergent trajectories in predictions of SOC in bioenergy systems.