The optimization of conservation agriculture practices requires attention to location-specific performance: Evidence from large scale gridded simulations across South Asia

The ways in which farmers implement conservation agricultural (CA) practices – which entail reduced tillage, maintenance of soil cover, and crop rotations – varies considerably in different environments, farming systems, and by the intensity with which farmers administer management practices. Such variability requires an efficient tool to evaluate the cost-benefit of CA, to inform agricultural policymakers and development priorities to facilitate expanded use of CA under appropriate circumstances. Rice-wheat rotation is the principal production system in South Asia (SA). Research has shown that CA can be promising in this rotation because of improved irrigated water, energy, and labor use efficiencies, in addition to the reduction in atmospheric pollution and potentially long term improvements in soil quality. Yield responses to CA are however varying across studies and regions. With a nine-year rice-wheat CA experiment in Eastern Gangetic Plains of South Asia, this study parameterizes the Environmental Policy Climate (EPIC) model to simulate five CA and conventional managements on the RW cropping system. Information from geospatial datasets and farm surveys were used to parameterize the model at the regional scale, increasing the management flexibility and range of localities in the simulation. Yield potential of the CAs in the whole SA was thereby explored by utilizing the model with various management strategies. Our results demonstrate how geospatial and survey data, along with calibration by a long-term experiment, can supplement a regional simulation to increase the model's ability to capture yield patterns. Yield gains from CA are widespread but generally low under current management regimes, with varied yield responses among CAs and environments. Conversely, CA has considerable potential in SA to increase rice-wheat productivity by up to 38%. Our results highlight the importance of applying an adaptive definition of CA, depending on environmental circumstances, while also building the capacity of farmers interested in CA to apply optimal management practices appropriate for their environment.