Water is a fundamental human need. Humans require at least 20 to 50 liters of clean, safe water a day for drinking, cooking, and simply keeping themselves clean. Sufficient safe drinking water is vital for public welfare and an important driver of a healthy economy. According to the World Health Organization, safe drinking-water is water that "does not represent any significant risk to health” (WHO, 2017). About 2 billion people in the world lack sufficient safe drinking water. About 1 million people are estimated to die annually as a result of unsafe drinking-water (WHO, 2018). Both, access to and the quality of drinking water are important. Protecting human health from adverse effects of unsafe drinking water is a top global priority of the United Nations Sustainable Development Goals (UN, 2018).

The search for pure drinking water began in prehistoric times. Ancient civilizations established themselves around water sources. Farming and the development of settlements lead to the beginning of the problem– how to get drinkable water for humans and cattle and how to manage the waste they produce. The availability of water in large quantities has been considered an essential part of human civilizations. The importance of good quality drinking water has been known for years, but the importance of proper sanitation was not understood until the 19th century, while standards for water quality appeared only in the early 1900s. Only gradually, people recognized that their senses alone were not accurate judges of water quality (Baker, 2012; Juuti et al., 2007).

The health effects of nitrate (NO₃^-) and nitrite (NO₂^-) in drinking water have long been debated (L’Hirondel, 2001; Bryan and Van Grinsven, 2013). The 1958 WHO International Standards for Drinking-water stated that the ingestion of water containing nitrates in excess of 50–100 mg/l (as nitrate) may give rise to methaemoglobinaemia in infants under 1 year of age (Schullehner et al. 2018). In the 1963 International Standards, this value was lowered to 45 mg/l (as nitrate), which was retained in the 1971 International Standards. The current guideline values are 50 mg/l for nitrate ion and 3 mg/l for nitrite; they are meant to protect against methaemoglobinaemia in bottle-fed infants (WHO, 2017).

Nitrate in groundwater and surface waters originates primarily from nitrogen fertilizers and manure storage and spreading operations, and from sewage waste and septic systems, The global amounts of nitrate-nitrogen lost from sewage and septic systems to groundwater and rivers greatly differ between countries; averages range from 1 to 6 kg of nitrogen per person per year (Van Drecht et al., 2009). Global losses from fertilizers and manures are a factor 2 to 4 larger (Beusen et al., 2016). Nitrogen that is not taken up from soil by plants may be lost to surface waters and groundwater as nitrate via surface runoff and leaching (Burt et al., 1993). This makes the nitrogen unavailable to crops and increases the nitrate concentration in groundwater and surface waters (Sutton et al., 2011).

The pollution of groundwater and surface waters with nitrate has shifted in scale from local in the past to regional and continental dimensions currently (Burt et al., 1993). Mean nitrate concentrations in groundwater have remained relatively stable in Member States of the European Union (EU) since 1992, although there is wide variation at the scale of individual groundwater bodies. Approximately 13 % of the stations across EU in 2009, exceeded the 50 mg/l limit (EC, 2014). Pristine lakes and rivers have a nitrate concentration of about 0.1 mg NO₃^- N per liter. The mean nitrate concentration in European rivers ranged between 0.5 and 5.0 mg N per litre in 2012, suggesting a 5 to 50 times increase relative to background concentration levels (EEA, 2015). However, the average nitrate concentration in European rivers reduced 0.5 mg NO₃^-N per liter during the period 1992 to 2012, as a result of various measures.

The European Union (EU) has developed a series of directives, guidelines and policies over the last decades to decrease the pollution of drinking water sources by nitrates from agriculture, industry and households. The requirements of the EU Drinking Water Directive set an overall minimum quality for drinking water within the EU. The EU Water Framework Directive, the Groundwater Directive, and the Nitrates Directive require Member States to protect drinking water resources against nitrate pollution in order to ensure production of safe drinking water.

The aforementioned directives have as yet not achieved a consistent level of implementation and effectiveness across all Member States. As a consequence, limits for nitrate (50 mg/l) are still exceeded in some areas with vulnerable water resources. Diffuse pollution of nitrogen from agriculture is the main obstacle to meeting the Drinking Water Directive targets for nitrate and nitrite.

Various measures and good agricultural practices have been developed and implemented in practice at farm level in the EU. These measures and practices have been successful in some regions but not in all (Dalgaard et al., 2014). There is a huge diversity within the EU in farming systems, climate, geomorphology, hydrology, soils, education level of farmers, quality of extension services, and type of water supplies, which means that site-specific measures and good practices are required to decrease nitrate pollution of drinking water resources. Coherent site-specific packages of measures are needed. However, the critical success factors that determine the effectiveness of these measures on a site by site basis are not well-known. It has been recognized in several studies and working groups that environmental directives and the Common Agricultural Policy should be better integrated when focusing on the protection of drinking water resources. The possibility of an integrated risk assessment and risk management by using Water Safety Plans, which was recently included in the Drinking Water Directive, is generally welcomed as a vehicle to become more flexible and proactive. In general, there is a growing consensus that good water governance is an essential prerequisite for water management since multiple actors may contribute to pollution.

There are several excellent reviews about nitrates from agriculture in groundwater and surface waters and about measures to reduce the loss of nitrate from agriculture (e.g., Addiscott et al., 1991; Burt et al., 1993; Goulding, 2000; Kirchmann et al. 2002; Mosier et al., 2004; Osterburg et al., 2007; Hatfield and Follett, 2008; Sutton et al., 2011; Cost869, 2011). Most of our current understanding of the mechanisms of nitrate losses from agriculture and of the measures to reduce these losses has been established in 1950s to 2000s, and much of the experimental testing of measures to reduce losses has been conducted in that period. Thereafter, simulation took over much to the scientific studies on nitrate losses from agriculture (e.g., Thomassen et al., 1991). This does not mean that no testing has been done during the last 2 or 3 decades, but that the experimental testing was often done in function of model calibration and validation. As a result, there are a large number of simulation models that are able to estimate the effects of measures to reduce nitrate leaching, as function of climate, soil, hydrology, and agricultural management conditions (Table 1).

Most of these models have only been applied to the region for which they were developed. The models differ from each other with respect to:

• The aim for which they were developed (academic research, water management tool, policy advise)
• The spatial scale on which they are applied
• The type of output they can produce (nitrate fluxes and/or concentrations; groundwater and/or surface waters)
• The type of process descriptions that are implemented and the temporal simulation scale

Table 1 lists a number of simulation models for the field scale and the regional scale estimation of nitrate losses. All the models listed are able to calculated nitrogen losses from the root zone, but not all the field scale models consider transport routes to groundwater and / or surface waters. Most of the field scale models have a strong focus on the organic matter and nitrogen cycle in the root zone and how these are influenced by agricultural management.

Table 1 Overview of simulation models used to estimate nitrate leaching at the field scale and/or the regional scale.

Despite the implementation of a range of policy measures since the early 1990s, the nitrate problems still persist across EU-28, although less severe than in the 1990s-2000s (EEA, 2015). There are various reasons for explaining why policy measures have been less effective than initially thought (e.g. Sutton et al., 2011). A main reason is that nitrogen is a key input in agriculture for crop and animal productivity, and that the nitrogen cycle is a leaky cycle. Another possible reason is that measure to reduce nitrate losses from agriculture to water bodies are perhaps less effective (quantitatively) than initially thought, and/or less effective in practice than in experimental conditions.

In this section of FAIORWAYiS we review and assess of measures and practices to decrease nitrate pollution of drinking water. The work builds on insights and results gathered in EU-wide and global projects and studies. It provides an overview and assessment of the effectiveness and efficiency of measures and practices aimed at decreasing nitrate pollution of drinking water reservoirs. 

The novel aspect of this study is that the accessible literature has been screened for experimental data related to the effectiveness and efficiency of basically all measures to reduce nitrate pollution of groundwater and surface waters, in a coherent and quantitative manner, using statistical analyses. 

Note: For full references to papers quoted in this article see

» References