|Oene Oenema, Meindert Commelin, Piet Groenendijk, John Williams, Susanne Klages, Isobel Wright, Morten Graversgaard, Irina Calciu, António Ferreira, Tommy Dalgaard, Nicolas Surdyk, Marina Pintar, Christophoros Christophoridis, Peter Schipper, Donnacha Doody
|»Oenema, O. et al. 2018. Review of measures to decrease nitrate pollution of drinking water sources. FAIRWAY Project Deliverable 4.1, 125 pp
Most of the drinking water used in EU originates from groundwater (66%) followed by surface waters (30%) (Figure 12). The use of groundwater is dominant in Germany, France, Spain, Italy, Denmark, Belgium, The Netherlands. The use of surface water is dominant in United Kingdom, Portugal, Czech Republic, Finland, Estonia, and Ireland. The use of groundwater and surface waters greatly depends on the availability of fresh and clean groundwater and surface waters.
The pollution of groundwater and surface waters with nitrate from agriculture depends on the nitrate sources in agriculture, the hydrological pathways and the nitrate removal/retention processes during transport (Klages et al., 2018). Here we briefly discuss the hydrologic cycle, hydrological pathways and the factors that contribute to groundwater recharge and nitrate removal/retention processes during transport.
|1. The hydrological cycle
|2. Surface runoff and nitrate leaching
|3. The potential pollution of groundwater and surface waters with nitrates
|4. Monitoring of the pollution of groundwater and surface waters with nitrates
Solar radiation is the basic driver of the hydrological cycle (Figure 13). It ‘fuels’ evapotranspiration from plants, soil and water surfaces. The moist air moves up but once in cold air layers it condenses to form clouds, and thereafter returns to the surface as precipitation. Some of the rain evaporates back into the atmosphere, some enters surface waters through surface runoff, and some infiltrates the soil and percolates into groundwater and may ultimately seeps its way to rivers, lakes and oceans, and then is released back into the atmosphere through evaporation (Figure 13).
The geomorphology determines the drainage system that is formed by the pattern of streams, rivers, and lakes in a particular drainage basin. The drainage basin is the topographic region from which a stream receives runoff, through flow, and groundwater flow. The number, size, and shape of the drainage basins varies from region to region. The geomorphology influences also the partitioning of water between groundwater and surface waters, depending also on the balance of precipitation and evapotranspiration. Factors affecting the water balance are plant-atmosphere interactions, surface runoff, infiltration, flow in the unsaturated-saturated zone and subsurface runoff. The dynamics of the groundwater itself also influences the partitioning between groundwater and surface waters; at high groundwater levels infiltration decrease and surface runoff increase.
Groundwater is often divided in two subsystems
- the shallow groundwater with the (partly) unsaturated zone with rapid transport of solutes through shallow groundwater to local water courses (subsurface runoff) and
- the deep groundwater saturated zone with slow transport towards larger streams and rivers.
Shallow groundwater flow is assumed to occur in the top layer of the soil, and is characterised by short residence times before water enters local surface water (small rivers) or deeper groundwater. Deep groundwater flow occurs in unconsolidated aquifers of ~50 m thickness and has often a much longer residence times before water enters large rivers.
The infiltration capacity of the soil depends on its porosity, which depends on its texture and structure, as well as on the soil moisture content before the rainfall started. The initial infiltration capacity of a dry soil may be high but, as the rain continues, it decreases until it reaches a steady state infiltration rate. When the rate of rainfall (intensity) exceeds the infiltration capacity of the soil, runoff will be generated and continues as long as the rainfall intensity exceeds the actual infiltration capacity of the soil. The vegetation exerts influence on the infiltration capacity of the soil; a dense vegetation cover shields the soil from the raindrop impact and reduces crusting effects.
Meteorological factors that affect runoff are type of precipitation (rain, snow, etc.) and rainfall intensity, amount and duration. Biophysical factors affecting runoff are land use and vegetation, soil type and depth, type of underlying bedrock, drainage area, geomorphology (slope of the land), basin type, and drainage network patterns, including ponds, lakes, reservoirs, sinks, which prevent or delay runoff from continuing downstream. Human activities that may affect runoff are the removal of vegetation and soil, grading the land surface, including terracing, and constructing drainage networks. These activities increase runoff volumes and shorten runoff time into streams from rainfall and snowmelt. Also, soil sealing in urban and infrastructural areas, and soil compaction by heavy machinery decrease the infiltration of water into the soil and thereby surface runoff. As a result, the peak discharge, volume, and frequency of floods may increase in nearby streams.
The residence time of water in a groundwater systems is important for the prognosis of the long-term behaviour of groundwater systems in response to nitrate inputs. The longer the residence time, the older the water, the greater the chance that the groundwater has been influenced by anthropogenic influence, and the greater the chance that natural remediation can improve the quality of polluted groundwater.
There are two loss pathways at the soil surface that are causing nitrogen losses to surface waters, namely surface runoff and erosion. Nitrogen losses through surface runoff are in general much larger than N losses via erosion. Losses of N via runoff and erosion are related to the factors controlling runoff and erosion, slope and geomorphology, vegetation cover, and to the amounts of soil mineral N and particulate N in the soil surface layers. The potential of N loss via surface runoff is much higher directly following applications of fertilizer N and manure than following crop harvest when mineral N has moved from the soil surface into the (sub)soil and/or has been taken up by the crop. In contrast, losses of particulate N via erosion may by higher following crop harvest when the soil surface is exposed to the impacts of rain than during the growing season when the soil surface is shielded by vegetation.
The N losses via downward leaching are related to
- the factors controlling infiltration,
- the amounts of mineral N in the soil profile, and
- the removal of nitrate via uptake by the crop and denitrification.
The amounts of mineral N in the soil profile depend on the balance of total N inputs to the soil and total N output via harvested crops and soil surface losses (NH3 and N2O emissions, surface runoff and erosion), corrected for net N mineralisation of organically bound N and denitrification in the soil. Factors controlling denitrification are
- the presence of an energy source for denitrifying bacteria, mostly easily decomposable organic carbon,
- near anoxic (anaerobic) conditions, and
- the availability of nitrate in soil.
If any of these three conditions is not fulfilled, denitrification is unlikely.
The leaching of nitrate to below the rooting zone moves further to either subsoil lateral leaching to surface waters or to groundwater (Figure 13). Groundwater transport of nitrate may take place over long distances and time-scales, and the groundwater system may act as a temporary sink, depending on denitrification, i.e. the reduction of nitrate (NO3-) and nitrite (NO2-) to N2O, NO and N2. The importance of denitrification in groundwater reservoirs itself is uncertain (Van Drecht et al., 2003; Rivett et al., 2008; Bouwman et al., 2014).
Increases in precipitation will generally lead to an increase in N leaching. As a result, there is often a good relationship between precipitation amount and nitrate loads to a river and nitrate concentrations in a river. The relationship between nitrate concentration and river flow result from the leaching of nitrate from the soil during periods of high rainfall. Hence, the leaching of nitrate is affected by dry and wet climatic cycles and by variation in precipitation, both between and within years. This has consequences for the interpretation of the results of the monitoring of groundwater quality, for example for assessing the effectiveness of measures to reduce nitrate leaching. Hence, measured nitrate concentration in monitoring programs may be corrected for dilution associated with differences in actual precipitation and average precipitation (Fraters et al. 2015).
Soil texture influences soil porosity and drainage, which in turn influences nitrate leaching and the aeration of the subsoil, which subsequently control mineralisation and (de)nitrification processes. In general, denitrification losses will increase in the order: sandy soil < loamy soils < clay soils < peat soil (Rivett et al., 2008; Fraters et al., 2015). The risk of nitrate leaching decreases when the rooting depth increases, as deeply rooting crops can remove NO-3 from the subsoil (Kristensen & Thorup-Kristensen, 2004). Organic matter-rich soil may mineralize NO3- from the soil organic matter, which may leach via subsurface tile drainage, especially in wet years follow dry years (Randall & Mulla, 2001; Hatfield, 1996). However, a high organic C content of the soil may also increase the denitrification capacity (Bijay-Singh et al., 1988; Munch & Velthof, 2006). Conversely, denitrification is low and nitrate leaching risk high when the organic C content of the soil is low, because the denitrification capacity is low when the total degradable C content of the soil is low (Bijay-Singh et al., 1988; Kronvang et al., 2005). Since the organic matter content of grassland is generally higher than that of arable land, grasslands have a higher denitrification capacity and a lower risk for nitrate than arable land per unit N surplus.
Drainage of poorly drained soils will lead to a lowering of the groundwater level, which may result in increased mineralisation, especially in organic matter-rich soils, which in turn may result in an increase in nitrate leaching. Sub-surface irrigation, i.e., water delivered from below soil surface, may also lower nitrate leaching losses because of less downward water flows and less mineralisation (due to dryer topsoil) (Elmi et al., 2002; Randall & Mulla, 2001).
Summarizing, the source-pathway-receptor linkage is complex and greatly varies between landscapes and regions, also because the traveling pathway and the traveling time of the groundwater may vary before it seeps into surface waters. Commonly, a distinction is made between
- surface runoff of N, which leads to nitrate pollution and eutrophication of surface waters, and
- downward leaching of nitrate, which leads to groundwater pollution by nitrates, and upon seepage of this groundwater may lead to nitrate pollution and eutrophication of surface waters.
The risk of nitrate leaching and runoff is related to a combination of
- incidence of occurrence, i.e., frequency of surface runoff and erosion, and
- the presences of nitrate in the soil.
Risks are termed high when both the incidence of occurrence and the amounts of nitrate in soils are high.
The demand for nutrients and water, and the demand for pest and disease control depend on the crop growth potential and management. Crop growth potential is an important determinant for the demand of nitrogen, and indirectly also for the leaching of nitrate to groundwater and surface waters. The spatial patterns of the potential crop biomass yields resemble similar spatial patterns as for climate, geomorphology and soil types. Areas with a high potential biomass yield demand more nutrients than areas with a low potential biomass yield. A large difference between potential biomass yield and water-limited biomass yield indicates the areas where irrigation may be important and hence, where irrigation induced nutrient losses may occur.
The potential risk of runoff and leaching of nitrate to surface waters is determined by a combination of pedo-climatic factors and the amounts of nitrate and in the top soil. The important pedo-climatic factors are:
- rainfall amount and distribution, especially heavy rainfall events, and
- water infiltration rate into the soil.
The latter is determined by slope, soil texture, soil structure, including soil cracking, slaking and preferential flow characteristics, soil depth to underlying rock, including karst formations and impermeable soil layers, vegetation cover, which determines evapotranspiration and affect surface roughness, snow and frost and freeze-thaw cycles, and the presence of terraces, tree-lines, buffer zones, riparian zones, which all contribute to intercepting overland flows.
The potential risk of downward nitrate leaching to groundwater is also determined by a combination of the amounts of nitrate in soil and pedo-climatic factors. The amounts of nitrate in soil are mainly determined by fertilization practices and the uptake capacity of the growing crop(s). Important pedo-climatic factors are rainfall surplus (i.e., rainfall minus evapo-transpiration), rainfall distribution, water infiltration rate into the soil and the hydrological conductivity of the soil, which is determined by soil texture, soil structure, including soil cracking, slaking and preferential flow characteristics, soil depth to underlying rock, slope, soil cover, and denitrification capacity of the soil. Soils with a high nitrate leaching vulnerability have a high infiltration rate and a high hydrological conductivity, such as coarse-sandy soils and shallow soils overlying karst formations. This vulnerability is increased in case of crops with short growing periods in a climate with high rainfall.
The potential risks of surface runoff in EU-27 is shown in Figure 15. Three classes of risk have been distinguished, namely low, medium and high. Most areas in Europe have a low to medium high risk of surface runoff. The potential risks of downward leaching in EU-27 is shown in Figure 16. Again, three classes of risk have been distinguished, namely low, medium and high. Most areas in Europe have a medium to high risk of leaching.
In summary, the surface runoff and leaching risk maps depicted in Figures 15 and 16 provide only a general overview, based on pedo-climatic factors. The maps show that the potential risks of surface run off to surface waters and leaching to groundwater are wide-spread across EU-27. The actual surface runoff and leaching also depend on the presence of nitrate sources. The maps are too course to derive the risks for individual drinking water resources, because the influencing pedo-climatic factors and N use in agriculture greatly vary at small spatial scales.
The EU-Nitrates Directive demands Member States to monitor the nitrate concentrations in groundwater and surface waters, and to report to the European Commission the results of the monitoring programs every four years. The most recent synthesis report provides a detailed overview of the monitoring network and of the results of the monitoring. In the reporting period 2012-2015, the total number of groundwater monitoring stations in the EU-28 was 34,091 which is an average of 7.8 stations per 1,000 km2 of land. The station density varied from 0.6 in Finland to 130 stations per 1,000 km2 of land in Malta. The average sampling frequency of groundwater was nearly twice a year, and varied from less than once a year in Denmark, Latvia, Poland and Sweden to around 5 times a year in Belgium and Croatia. The total number of fresh surface water monitoring stations in the EU-28 was 33,042 which is an average of 7.6 stations per 1,000 km2 of land. The station density varied from 0.5 per 1,000 km2 in Finland to 34 stations per 1,000 km2 of land in the United Kingdom. The average sampling frequency was around four times a year, and varied from almost once a year in Sweden to 20 times a year in Ireland (EC, 2018).
The average annual nitrate concentration exceeded 50 mg/L in 13% of groundwater monitoring stations in the EU-28 during 2012-2015. This varied from no exceeding stations in Ireland, to more than 20% in Spain, Germany and Malta. At EU-28 level, there was a slight improvement compared to the previous reporting period, when 14% of the groundwater monitoring stations exceeded an average annual nitrate concentration of 50 mg/L. Compared to the previous reporting period 2008-2011, 26% of all stations in the EU-28 showed an increasing trend and 32% a decreasing trend. Figure 17 shows a maps with the location of the groundwater monitoring station and their average nitrate concentration. Stations with nitrate concentrations exceeding 50 mg/L are diffusively spread across EU-28, with the exception of Sweden, Finland and Ireland. Yet, there are also a few hot spot regions.
The average annual nitrate concentration exceeded 50 mg/L in 1.8% of the fresh water monitoring station in the EU-28 during 2011-2015. Another 2.0% of the stations had average annual nitrate concentrations between 40 and 50 mg/L and 8.8% between 25 and 40 mg/L. Low average nitrate concentrations in fresh surface water were found in Sweden, Ireland and Greece, and relatively high in the United Kingdom, Belgium and Malta. High nitrate concentrations are generally observed in rivers. There was a slight improvement compared to the previous reporting period, in which 2.4% of the monitoring stations had annual average nitrate concentrations exceeding 50 mg/L and 2.4% showed concentrations between 40 and 50 mg/L. Compared to the reporting period 2008-2011, a decreasing trend in annual average nitrates concentrations was observed in 31% of all freshwaters monitoring stations, and an increasing trend was observed in 19% of freshwaters monitoring stations. Figure 18 shows a maps with the location of the freshwater monitoring station and their average nitrate concentration. Stations with nitrate concentrations exceeding 50 mg/L are again diffusively spread across EU-28, with the exception of Sweden, Finland and Ireland. Yet, there are also a few hot spot regions.
Note: For full references to papers quoted in this article see