Simulating long-term carbon nitrogen and phosphorus biogeochemical cycling in agricultural environments

Understanding how agricultural practices alter biogeochemical cycles is vital for maintaining land productivity, food security, and other ecosystem services such as carbon sequestration. However, these are complex, highly coupled long-term processes that are difficult to observe or explore through empirical science alone. Models are required that capture the main anthropogenic disturbances, whilst operating across regions and long timescales, simulating both natural and agricultural environments, and shifts among these. Many biogeochemical models neglect agriculture or interactions between carbon and nutrient cycles, which is surprising given the scale of intervention in nitrogen and phosphorus cycles introduced by agriculture. This gap is addressed here, using a plant-soil model that simulates integrated soil carbon, nitrogen and phosphorus (CNP) cycling across natural, semi-natural and agricultural environments. The model is rigorously tested both spatially and temporally using data from long-term agricultural experiments across temperate environments. The model proved capable of reproducing the magnitude of and trends in soil nutrient stocks, and yield responses to nutrient addition. The model has potential to simulate anthropogenic effects on biogeochemical cycles across northern Europe, for long timescales (centuries) without site-specific calibration, using easily accessible input data. The results demonstrate that weatherable P from parent material has a considerable effect on modern pools of soil C and N, despite significant perturbation of nutrient cycling from agricultural practices, highlighting the need to integrate both geological and agricultural processes to understand effects of land-use change on food security, C storage and nutrient sustainability. The results suggest that an important process or source of P is currently missing in our understanding of agricultural biogeochemical cycles. The model could not explain how yields were sustained in plots with low P fertiliser addition. We suggest that plant access to organic P is a key uncertainty warranting further research, particularly given sustainability concerns surrounding rock sources of P fertiliser.