Data-based model representations of plant-microbe interactions improve predictions of bioenergy sustainability

Joanna Ridgeway, Christopher A. Walter and Edward R. Brzostek, Department of Biology, West Virginia University, Morgantown, WV, Stephanie Juice, Rubenstein School of Environment and Natural Resources, University of Vermont, Burlington, VT

Background/Question/Methods
The viability of bioenergy relies on accurate predictions of how policy and management strategies impact
sustainability in a changing world. Enhancing soil carbon (C) storage in bioenergy systems may increase
their sustainability by compensating for C emissions from the production and use of biofuels. However,
predictions of net C dynamics remain limited by a lack of data to inform model parameters that control
how microbial physiology and plant-microbial interactions impact soil C cycling.

To meet this need, we used a coupled model-experiment approach to reduce parameter uncertainty in the
FUN-CORPSE (Fixation and Uptake of Nitrogen- Carbon, Organisms, Rhizosphere, and Protection in the
Soil Environment) which we recently adapted to bioenergy. By tracing the fate of 13C labeled bioenergy
litter in a laboratory soil incubation, we were able to constrain key parameters that control soil C
dynamics including microbial carbon use efficiency, turnover, and the protection rate of new soil C. We
then compared the ability of baseline model runs vs. model runs with empirically-based parameters to
capture our experimental results and field-based estimates of soil C dynamics from the University of
Illinois Energy Farm.
Results/Conclusions
Data-driven parameters for maize, energy sorghum, and miscanthus x giganteus (MXG) shifted modeled
soil C by up to 25% and altered the distribution between protected and unprotected forms by up to 40%.
These model results more accurately reflected measured differences across the bioenergy crops in our lab
experiment than simulations with the base model decomposition parameters. When we tested the refined
model against field data, FUN-CORPSE captured differences in soil C stocks across all three feedstocks.
These results highlight how coupled model-experiment approaches can improve our predictive
understanding of soil C dynamics and the sustainability of bioenergy.