Date of Graduation


Document Type


Degree Name

Master of Science in Biological Engineering (MS)

Degree Level



Biological and Agricultural Engineering


Dharmendra Saraswat

Committee Member

Marty D. Matlock

Second Committee Member

Charles P. West


Applied sciences, Health and environmental sciences, Earth sciences, Biofuel crops, Land use change, Marginal land, Swat model, Water quality, Watershed management


This study compares a novel simulation approach to the conventional Soil and Water Assessment Tool (SWAT) modeler's approach for targeting biofuel crop production on marginal lands. In conventional SWAT modeling approach, non-spatial definition of hydrological response units (HRUs) results in the simulation of biofuel crops on both marginal and non-marginal land. This study provides an alternative approach in which a marginal-land raster was integrated into the land use and land cover (LULC) raster in such a way that the land uses were divided into marginal and non-marginal components. This modified LULC was used for model setup which resulted in marginal and non-marginal HRUs. This approach was evaluated for the L'Anguille River watershed (LRW) by calibrating and validating for total flow, surface flow, base flow, sediment, total phosphorus, and nitrate-nitrogen followed by the simulation of biofuel crops only on marginal HRUs.

The results were analyzed for two cellulosic (second generation) biofuel crops: switchgrass (Panicum virgatum L.) and miscanthus (Miscanthus x giganteus). Compared to novel modeling approach, simulations using the conventional approach showed an increase in sediments by 20% and 61%, total phosphorus by 17% and 53%, and total nitrogen by 25% and 65% for the switchgrass and miscanthus, respectively. Compared to simulated pollutant losses from a mix of baseline row crops, switchgrass and miscanthus showed 94% and 78% decrease in sediment, 96% and 90% decrease in total phosphorus, and 80% and 67% decrease in total nitrogen, respectively. This study provided a novel approach to incorporate marginal land into the SWAT model and the model outputs suggest that producing perennial grass biofuel crops on marginal lands of the LRW resulted in lower sediment, total phosphorus, and total nitrogen losses than that obtained by conventional SWAT modeling. Pollutant losses from the non-targeted marginal HRUs explained the differences in the sediment, total phosphorus, and total nitrogen losses. The simulation results also suggested that substantial reduction in pollutant losses could be achieved by replacing baseline row crops with perennial grass crops on marginal lands in the LRW.