Search

In this study we investigated critical nitrogen (N) inputs and N losses and their exceedances (actual inputs or losses minus critical inputs or losses) for agricultural soils in the EU-27 region in view of adverse effects. Critical N inputs were calculated based on a critical (i) ammonia emission rate related to critical N loads in view of biodiversity protection, (ii) N concentration in runoff to surface water of 2.5 mg N l-1, related to the protection of aquatic ecosystems and (iii) nitrate (NO3-) concentration in leachate to groundwater of 50 mg NO3- l-1, based on the safe limit for drinking water. In addition, we calculated the required N inputs to achieve target crop yields, which were set at 80% of the water-limited yield potential. Calculations were performed with the INTEGRATOR model for ca 40,000 unique soil-slope-climate combinations and then aggregated at NUTS3-, country- or EU-27 level. Results show that critical N inputs at EU-27 level as compared to actual (year 2010) N inputs are 31% lower to protect biodiversity and 43% lower to protect surface water quality. Critical N inputs are most strongly exceeded in regions with high livestock densities, such as Ireland, the Netherlands, Flanders in Belgium, Brittany in France and the Po valley in Italy. Inversely, required N inputs to attain target crop yields at EU-27 level are ca. 40% higher. Especially in Eastern Europe, there is a large potential to increase yields by increasing N fertilization, but this generally requires strongly improved N use efficiency (NUE), since critical N inputs are mostly lower. Using a maximum plausible NUE, the surface water criterion cannot be achieved on 17% and 25% of all agricultural land at actual crop yield and target crop yields, respectively. Similarly, a maximum plausible reduction in NH3 emission fractions also still causes exceedances of critical N loads in view of biodiversity protection. Reducing agricultural production is needed to protect biodiversity and/or water quality in those regions.

Before the industrial revolution, nitrogen was scarce and its availability severely constrained crop production. This changed with the invention of the Haber-Bosch process, allowing humans to fix atmospheric di-nitrogen at unprecedented rates. Currently, rates of human nitrogen fixation have surpassed natural rates by a factor 2–3. This has greatly contributed to increases in crop yields, but also led to adverse impacts on the environment and human health.Sustainable nitrogen management requires a systems approach, including tools that can quantitatively assess trade-offs between benefits of nitrogen for crop and livestock production and the adverse impacts of nitrogen on the environment. This thesis describes the further development and application of spatially explicit models to assess these trade-offs at the European and global scale. More specifically, the thesis addresses the following questions:How much nitrogen can we safely apply in agriculture before crossing thresholds for air and water quality?How can we manage nitrogen to both achieve environmental targets and meet current and future feed and food demand?What is the climatic impact of anthropogenic nitrogen inputs, i.e., how much additional carbon is stored in terrestrial systems due to anthropogenic nitrogen deposition, and to what extent does this ‘carbon bonus’ offset the climate impact of N2O emissions?This thesis aims to provide information on policy-relevant nitrogen indicators that capture trade-offs between nitrogen’s adverse impacts and benefits, to support policies on sustainable nitrogen management in Europe and globally. Questions relevant to this objective are explored in seven chapters.Chapter 1 (“Introduction”) describes the challenge of sustainable nitrogen management and presents knowledge gaps that are addressed by the research described in the thesis.Chapter 2 (“Spatially explicit boundaries for agricultural nitrogen inputs in the European Union to meet air and water quality targets”) presents a method for deriving ‘critical’ agricultural nitrogen inputs related to targets for to nitrogen deposition (in view of critical limits to protect terrestrial ecosystems), nitrogen runoff to surface water (in view of water quality targets) and nitrogen leaching to groundwater (in view of drinking water norms). The method is applied to estimate spatially explicit critical nitrogen inputs in the European Union using the nitrogen balance model INTEGRATOR.Chapter 3 (“Reconciling food production and environmental boundaries for nitrogen in the European Union”) builds on the work described in Chapter 2 and assesses how crop production can be maximized while keeping nitrogen losses to air and water below critical levels. Strategies that are explored include redistributing nitrogen inputs to close yield gaps in regions where thresholds are not exceeded and increasing nitrogen use efficiency in crop and livestock production.Chapter 4 (“From planetary to regional nitrogen boundaries for targeted policy support”) presents spatially explicit critical nitrogen inputs to agriculture at the global scale, which are derived based on the IMAGE-GNM model. The assessment of critical nitrogen inputs explicitly accounts for non-agricultural nitrogen losses as well as opportunities to increase nitrogen inputs to close yield gaps in regions where environmental thresholds are not exceeded. Critical inputs are aggregated to derive a safe ‘planetary boundary’ for agricultural nitrogen inputs.Chapter 5 (“Global-scale impacts of nitrogen deposition on tree carbon sequestration in tropical, temperate, and boreal forests: A meta-analysis”) describes results from a meta-analysis of nitrogen addition experiments in forests to estimate additional carbon sequestration in forest aboveground woody biomass resulting from nitrogen addition. Average carbon-to-nitrogen response rates are derived for tropical, temperate and boreal forests, and the global nitrogen-induced carbon sink is estimated by multiplying mean responses with nitrogen deposition in each forest biome.Chapter 6 (“Experimental evidence shows minor contribution of nitrogen to global forest carbon sequestration”) further builds on Chapter 5 by exploring sources of heterogeneity in nitrogen-induced forest carbon sequestration using meta-regression. The derived regression model predicts forest carbon-to-nitrogen response based on soil, climate and tree characteristics and is used to derive global maps of nitrogen-induced carbon sequestration in forest aboveground biomass.Chapter 7 (“Synthesis”) summarizes the main findings in the context of the overall objective (providing information on policy-relevant nitrogen indicators to support policies on sustainable nitrogen management), and critically discusses the approaches chosen to answer the research questions. In addition, the chapter presents two additional applications of the developed models to policy-relevant question on sustainable nitrogen management. First, it addresses the relationship between nitrogen boundaries estimated in Chapter 4 and the climatic impact of human nitrogen use estimated in Chapter 6. Second, it presents a revised methodology to calculate the “Sustainable Nitrogen Management Index”, an indicator that is used to measure countries’ progress towards meeting the Sustainable Development Goals.

Nitrogen (N) losses to air, ground water and surface water in response to agricultural N inputs affect air and water quality. Agricultural N inputs in this article are defined as mineral N fertilisers, N manure and biosolids and biological N fixation. Using a spatially explicit N balance model, we assessed where agricultural N losses within THE EU-27 currently lead to an exceedance of critical ammonia (NH3) emissions in relation to adverse impacts on terrestrial ecosystems, critical N concentrations in runoff to surface water in relation to eutrophication impacts and critical nitrate(NO3) concentrations in groundwater in relation to drinking water quality. We then calculated the N inputs at which critical N emissions or concentrations are just not exceeded (‘critical’ N inputs). We also assessed required N inputs in order to achieve target yields, defined as 80% of the water-limited yield potential at actual N use efficiency. Actual, critical and required N inputs were calculated for c.40,000 unique soil-slope-climate combinations throughout the European Union. When actual or required N inputs exceeded critical inputs, we calculated the necessary reduction inammonia emission fractions and necessary increase in NUE to attain actual or target yield while simultaneously reaching air and water quality goals. The ammonia emission fraction referred to the ratio of the total NH3-N emissions, divided by the total N excretion by livestock. Results show that required N inputs at the EU-27 level are on average 27% higher than actual inputs. Average critical N inputs are 31% and 43% lower than actual N inputs in relation to critical NH3 emissions and critical N runoff to surface water, respectively, but 1% higher in relation to critical NO3 leaching to groundwater. The risk for surface water is, however, likely overestimated, since calculated N concentrations in runoff to surface water appear to be higher than concentrations in surface water. An overall reduction in N inputs of 30% to protect air and water quality seems a reasonable average estimate. Critical inputs are most strongly exceeded in regions with high actual N inputs, such as Ireland, the Netherlands, Belgium and Luxembourg, Brittany in France and the Po valley in Italy. The actual N use efficiency (NUE) for all agricultural land, averaged over the EU-27 is 61%. This value has to increase on average to 72% to protect surface water quality at actual crop yields and to 74% at target crop yields. Opportunities thus exist to reduce the environmental impact of agricultureby increasing the NUE, while still allowing an increase in crop production in the EU. However, in c.15-20% of the agricultural land area, it is not feasible to achieve the surface water criterion at actual crop yield and this area increases to 25% at target crop yield, because it would require an NUE over 90%. In these areas, an additional reduction of N inputs is necessary, but this comes at the expense of crop yield reductions.