Background and aimsWheat-based diversified crop sequences with varied crop residue amounts can result in considerable variation in the amount and type of crop residues generated, which in turn can alter the water-use efficiency and yield of conservation agriculture practices in dryland Mediterranean regions. Simulation analysis may generalize experimental findings over time and location, examine the impacts of management alterations in greater depth, and yield specific outcomes. The Agricultural Production Systems SIMulator (APSIM) is well-suited for this research; nevertheless, there are questions over its accuracy in simulating varying quantities and kinds of crop wastes in dryland conservation agriculture.MethodsData generated from the 9-year field experiment conducted at Cunderdin, Western Australia using factorial combinations of 10 different crop sequences (monoculture wheat vs. wheat rotated with legumes and canola) and two crop residue levels (high-spread & standing-HR and low-only standing-LR), were parameterized and validated through APSIM version 7.10 based on measured and simulated yield and water-use data in different crop sequences. We used 10 built-in modules for this purpose: Wheat, Barley, Canola, Lupin, Fieldpea, Chickpea, Oats, SOILWAT, SURFACEOM, and Fertilizer. The calibrated model runs were also implemented at the study site to understand the long-term (1980-2015) impacts of different crop sequences managed with HR & LR conditions. Two indicators, coefficient of variation (CV) of wheat equivalent yield (WEY) for all simulated years, and mean WEY were used to find-out the risk management role of different crop sequences managed with HR & LR condition.ResultsAPSIM performed 'satisfactorily' (p <= 0.05) in case of predicting grain yields of wheat, barley and canola with the model efficiency or relative RMSE of 0.20, 0.26 and 0.36, respectively, under LR condition. APSIM also performed 'satisfactorily' (p <= 0.05) in case of predicting water-use by wheat (0.12 & 0.13), barley (0.08 & 0.09) and canola (0.14 & 0.15) in LR & HR conditions, respectively. Long-term simulations revealed 1) significantly (p <= 0.05) higher WEY were recorded under HR than LR condition. 2) Within HR & LR, the highest WEY was recorded in monoculture wheat rotation with the CV of 36% (less risk) followed by cereal rotation including barley and oat cover crop with the CV of 51% (high risk).ConclusionOur study provided a critical evaluation of the accuracy with which APSIM could simulate grain yields and water-use for diverse crop sequences managed under a long term no-tillage situation in a rainfed Mediterranean Western Australian environment or in other areas with similar ecology. These results will give confidence to the modellers to run this validated model in long-term for predicting the water-use efficiency and yield of diverse crop sequences for climate projections in rainfed Mediterranean environments.