Modeling soil organic matter: theory, development, and applications in bioenergy cropping systems
Date
2015
Authors
Campbell, Eleanor Elizabeth, author
Paustian, Keith, advisor
Parton, William J., committee member
Cotrufo, M. Francesca, committee member
Reardon, Kenneth F., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Soil organic matter (SOM) is a complex, dynamic, and highly variable soil constituent that is of fundamental importance to many soil functions, terrestrial ecosystem processes, and biogeochemical cycles. Its importance extends across scales, ranging from site-specific impacts on soil fertility to the global net exchange of carbon between terrestrial systems and the atmosphere. Soil organic matter is impacted by human activities, as seen most directly in agricultural systems. In this context, SOM models play an important role in integrating the understanding of complex, interacting soil processes across temporal and spatial scales, contributing to land use decision making by providing comparative evaluation of soil impacts associated with different management practices. Crop-based bioenergy feedstock productions systems are an emerging area for these types of SOM model applications. However, model evaluations are dependent on the theoretical basis of a given SOM model, as well as the quality of data used to drive the model for a given system or management scenario. This study therefore explores linkages between advances in the theoretical understanding of SOM dynamics, the development of SOM models to reflect these advances, and the application of SOM models to assess crop-based bioenergy production systems. First, five emerging areas in SOM research were reviewed in the context of SOM models, including SOM stabilization mechanisms, saturation kinetics, temperature sensitivity, dynamics in deep soils, and incorporation into earth system models. These reviews demonstrated the importance of identifying where SOM model development and applications are most limited, whether in theoretical understanding, in model implementation, or in data availability. For example, SOM saturation kinetics is theoretically well understood but remains difficult to implement in SOM models, only yielding improvements in a narrow set of ecological conditions. SOM temperature sensitivity and deep soil dynamics, however, are more limited by poor data availability in addition to poor theoretical understanding of interacting processes. A selection of shortfalls in SOM modeling were then addressed and explored with the Litter Decomposition and Leaching (LIDEL) model, a litter decomposition model that incorporates dynamic microbial carbon use efficiency (CUE) and yields dissolved organic carbon (DOC) as one of the byproducts of litter decomposition. In this analysis a hierarchical Bayesian statistical approach was used to test model performance and estimate unknown model parameters using experimental data. While this analysis showed the LIDEL model successfully integrates hypotheses for litter nitrogen and lignin controls on dynamic microbial CUE and the generation of DOC from litter decomposition, there remains a great deal of uncertainty in the rate of microbial biomass turnover as well as the proportioning of biomass from microbial turnover between solid versus soluble microbial products. Targeted experimental evaluation of the generation of DOC from microbes versus litter would support greater certainty in these model parameters and further model development for more general applications. Finally, the performance of the DAYCENT ecosystem model was evaluated in simulating US corn residue removal and Brazilian sugarcane production, two types of crop-based bioenergy feedstocks. DAYCENT is a process-based ecosystem model that integrates a soil organic carbon model to simulate carbon and nitrogen cycling processes through plant-soil interactions. The results of DAYCENT corn residue removal simulations highlighted several DAYCENT model biases, such as low corn yield estimates in dry regions and an overestimation of soil carbon loss with conventional tillage. Despite these biases, the results showed the importance of considering interactive effects between corn residue removal and other crop management practices in this type of bioenergy feedstock production system. The results suggest corn residue removal is ideally paired with management practices—such as reduced tillage—to maintain or improve soil carbon stocks. The analysis of Brazilian sugarcane management practices also highlighted management practices poorly simulated by DAYCENT, in particular identifying the need to improve DAYCENT simulations of high N₂O emission conditions observed in mechanically-harvested sugarcane, perhaps by adding simulation of DOC movement across the soil profile. However, this analysis also identified a need for more accurate and consistent daily precipitation data to drive DAYCENT simulations of N₂O emissions from Brazilian sugarcane management practices, particularly as there is interest in regionally-scaled analyses of direct greenhouse gas emissions from sugarcane production in Brazil. Taken together, the results of this study show the importance of a close connection between emerging areas in SOM theory, SOM model developments, and SOM model applications in crop-based bioenergy feedstock production systems. This connection allows for the identification of specific areas in need of further research, whether developing new modeling approaches or gathering additional data to parameterize, drive, and evaluate model simulations. This connection should remain a central emphasis as SOM models are increasingly incorporated into crop-based bioenergy policy and land management decision making.
Description
Rights Access
Subject
crop-based bioenergy
soil organic matter
sugarcane
DayCent
corn stover
soil organic matter modeling