Management effects on soil health at the European scale

According to the European Commission, 70% of European soils are degraded. Soil information is key to finding synergetic solutions between food production and delivery of other ecosystem services from agriculture, such as carbon storage, improving water quality, and conserving biodiversity. The massive increase in standardized soil sampling and monitoring in the European Union since 2009 has revealed the large-scale spatial distribution of soil properties. However, in order to develop management recommendations for farmers, the soil data has to be linked to farm management data to determine which context-specific practices lead to improved soil quality.

In this interdisciplinary project we will bridge expertise from social, computer and natural sciences to combine soil information with farm management data. We will develop machine learning algorithms that predict crop-specific farm management at soil sampling locations using data from a representative network of farms across Europe, while protecting the privacy of individual farmers. This will allow disentangling the complex interactions between soil health, pedoclimatic context, and management to inform context-specific solutions for healthier agricultural soils.