Daycent model performance to simulate yield and soil carbon across diverse soil management practices in several long-term experiments

Agricultural systems are a major source of greenhouse gas emissions, contributing significantly to global climate change. As the demand for food increases, there is a need to identify sustainable soil management practices that minimize environmental impacts while maintaining or enhancing crop productivity. The Daycent model is a useful tool for simulating ecosystem responses to changes in soil management and climate change. In this study, we calibrated Daycent using yield and soil carbon data collected from five long-term ongoing experiments (PROspective and QualiAgro sites in France, V4 and V140 in Germany, FAST in Switzerland) involving various soil management practices and covering different pedoclimatic conditions across Europe, with the aim of upscaling the impacts of these practices. The treatments in these experiments include conventional to reduced tillage, addition of organic and mineral fertilizers, and the use of cover crops. Data from each experiment was split by treatment into two parts, the calibration dataset and the validation dataset. The Daycent model performance to simulate yield and soil carbon was evaluated by comparing simulated data with measured data using statistical indicators, e.g., rRMSE and R². Calibration dataset allowed for adjusting relevant parameters according to the calibration protocol. In model evaluation against the validation dataset across all sites, rRMSE ranged from 0.29 to 0.60 for yields, and from 0.03 to 0.14 for soil organic carbon stock. The R² values indicate that the Daycent model predicted 84% and 99% of the measured variability in yields and soil carbon stock, respectively across the sites, which implies that the model was able to capture the overall variability due to management and pedoclimatic conditions. This study demonstrates that the Daycent model can simulate yields and soil organic carbon in long-term field experiments with diverse soil management practices across different pedo-climatic conditions and can be used for the upscaling of these practices to the regional scale. However, the model needs to be further calibrated to effectively simulate yields across these sites to avoid under- or over-estimation. This research is a part of ClimateCropping project developed in the framework of the EJP for SOIL “Towards climate-smart sustainable management of agricultural soils” funded by the European Union Horizon 2020 research and innovation programme.