There is an increased demand on agricultural systems in the United States and the world to provide food, fiber, and feedstock for the emerging bioenergy industry. The agricultural intensification that this requires could have positive and negative feedbacks in productivity and the environment. In this paper we used the simulation model EPIC to evaluate the impact of alternative tillage and management systems on grain sorghum (Sorghum bicolor L. Moench) production in central and south Texas and to provide long-term insights into the sustainability of the proposed systems as avenues to increase agricultural output. Three tillage systems were tested: conventional (CT), reduced (RT), and no-tillage (NT). These tillage systems were tested on irrigated and rainfed conditions, and in soils with varying levels of structural erosion control practices (no practice, contour tillage, and contours + terraces). Grain yield differed only slightly across the three tillage systems with an average grain yield of 5.7 Mg ha(-1). Over the course of 100-year simulations, NT and RT systems had higher soil organic carbon (SOC) storage (100 and 91 Mg ha(-1), respectively) than CT (85 Mg ha(-1)), with most of the difference originating in the first 25 years of the simulations. As a result, NT and RT systems showed lower net global warming potentials (GWPs) (0.20 and 0.50 Mg C ha(-1) year(-1)) than CT (0.60 Mg C ha(-1) year(-1)). Irrigated systems had 26% higher grain yields than rainfed systems; yet the high energy needed to pump irrigation water (0.10 Mg C ha(-1) year(-1)) resulted in a higher net GWP for irrigated systems (0.50 vs. 0.40 Mg C ha(-1) year(-1)). Contours and contours + terraces had minimal impact on grain yields, SOC storage and GWP. No-till was the single technology with the largest positive impact on GWP and preservation or enhancement of SOC. Overall, the impact of individual tillage cropping systems on GWP seems to be decoupled from the productivity of a given location as determined by weather or soil type. When expressed per unit of output, high yield locations have a much lower GWP than low yield locations and would be therefore prime targets for production intensification. Published by Elsevier Ltd.
