Modeling the effects of mixed farming systems on soil carbon and crop-livestock productivity in central-west NSW

Simulation models are increasingly used for improvement of agricultural productivity through studying the interactions among biophysical processes in an agricultural farming system. In this study, we used AusFarm (agricultural systems analysis model) to simulate crop and livestock performance including long-term soil organic carbon (SOC) dynamics under different management options such as conventional and reduced tillage, non-tillage with continuous cropping and perennial pasture in a grazing system. The AusFarm model was able to explain 70% of the observed variability in wheat grain yield with a RMSD of 554 kg/ha and showed good agreement between observed and simulated livestock performance. Simulated crop yield and gross margins were higher in conventional and reduced tillage with 2-years annual pastures cropping rotations than continuous cropping. In the long-term field experiment, the observed SOC increased initially, and were comparatively higher in reduced tillage than conventional practices. However, observed SOC decreased by 2012 in all other management options though SOC are higher in perennial pasture. The simulated trends of SOC under different farming system treatments were generally comparable with the observed field trial data. Although crops, pastures and livestock performance is related to timing of rainfall and amount of rainfall, equally, soil fertility decline can influence crop-livestock productivity. Long-term simulations suggest, cropping frequencies including crop-livestock rotation adjustments and soil-nutrient management are needed for productive mixed farming with environmental gains.