Spatial Differentiation of Arable Land and Permanent Grassland to Improve a Land Management Model for Nutrient Balancing

Abstract

Agroecosystems play an important role in providing economic and ecosystem services, which directly impact society. Inappropriate land use and unsustainable agricultural management with associated nutrient cycles can jeopardize important soil functions such as food production, livestock feeding, and conservation of biodiversity. The objective of this study was to integrate remotely sensed land cover information into a regional land management model (LMM) to improve the assessment of spatially explicit nutrient balances for agroecosystems. Remotely sensed data and an optimized parameter set contributed to an improved LMM output, allowing for a better land allocation within the model. The best input parameter combination was based on two different land cover classifications with overall accuracies of 98%, improving the land allocation performance compared with using nonspatially explicit input. We conclude that the combined use of remote sensing data and the LMM has the potential to provide valuable guidance for farm practices. It further helps to generate a spatial description of farm-level nutrient balance, a crucial ability when choosing policy options related to sustainable management of agricultural soils.