The advantages of using field- and farm-scale data to target agri-environmental measures—an example of afforestation

Abstract Requirements for substantial reductions in environmental pollution such as nitrogen (N) loading to aquatic environments, and in greenhouse gase (GHG) emissions impose challenges for agriculturally-intensive regions in Europe. Here we use afforestation to illustrate how high-spatial resolution data can be used to improve the efficiency of implementation of an environmental measure. Since afforestation of agricultural land has the potential to reduce both aquatic N load and GHG emissions, targeting the reduction of one pollutant will also affect the non-targeted pollutant. We developed a method to use nationally-available, high-resolution data to minimise the agricultural area selected for potential afforestation, for a given reduction of N load or GHG emissions, and assess the co-reduction in the non-targeted pollutant. To illustrate the effect of imposing policy restrictions on the implementation of measures, two restrictions were investigated; limitations on the maximum proportion of each farm that could be afforested and threshold proportions of the farm area, above which the whole farm must be afforested. For N load, both the N leaching below the root zone and the efficiency of denitrification between the bottom of the root zone and the recipient aquatic ecosystem were significant factors determining the selection of fields for afforestation. Since N leaching was the only location-dependent GHG emission source and these were only a minor contributor to the total GHG emissions, the selection of land for afforestation to reduce GHG emissions depended more on farm-scale than field-scale characteristics. In a case study area, the availability of high-resolution data allowed the use of a targeted afforestation selection method that significantly reduced the land area required, relative to a non-targeted approach. With the targeted approach, reducing N load or GHG emissions by 25 % of the maximum potential reduction required the afforestation of 14 % and 18 %, respectively, of the case study area. This represented reductions in area of 42 % and 24 % compared to the untargeted approach. Measures targeting one pollutant also substantially reduced the non-targeted pollutant. We conclude that the implementation efficiency of some environmental policy interventions in agriculture depends on the availability of high-resolution agricultural and landscape data, and adequate methods to utilize these data.