Diffuse agricultural pollution of groundwater: addressing impacts in Denmark and Eastern England

Land-use practices in aquifer recharge areas exert a major influence on groundwater recharge and especially on its quality. Every land-use practice has a direct ‘water resource footprint’ and may result in diffuse groundwater pollution whose control presents a more complex policy challenge than point-source pollution.

Diffuse groundwater pollution by agricultural land-use practices was first substantiated through widely increasing nitrate concentrations in England (Foster et al. 1982), and subsequently its scope was extended to include pesticides (Foster et al. 1993; Foster 2000). It has become widely documented in cereal-growing areas across Northern Europe with concern that the EU-Water framework Directive was not really addressing the issue and that alternative approaches are required (Barataud et al. 2014; Dolan et al. 2014; Bouleau et al. 2020; Ptak et al. 2020; Skevas 2020; Mancuso et al. 2021). The overall position has recently been concisely summarised (Foster & Custodio 2019; Foster & Chilton 2021).

Given the large storage of most aquifer systems, the groundwater response to land-use change is invariably delayed and gradual, but also long-lasting and costly or impossible to remediate. Particularly critical are excessive nutrient and pesticide leaching following the conversion of natural pasture or forest to arable farming,

Both Denmark and Eastern England are highly dependent on groundwater for public drinking water-supply and have a long history (over 50 years) of extensive cereal cultivation. Today in Denmark, arable land occupies in excess of 60% with forest and grassland subordinate (at 13 and 12%, respectively), while in Eastern England the percentage of arable land is even higher at 80–85% with only 5% residual woodland.

In Eastern England, water supply is dominated by two large regulated (but privatised) water utilities with a further three smaller water companies. In Denmark, water-supply is very decentralised with 87 municipally owned (but autonomous) water utilities, including a few large ones that have led the battle to cope with diffuse agricultural pollution. Additionally, more than 2,000 small waterworks owned by the local consumers provide drinking water in the more sparsely populated areas.

Groundwater conditions show some similarity between the two regions – but in general terms the main aquifer in Eastern England (the Chalk) is more vulnerable to pollution being covered only by very patchy, mainly permeable, drift deposits, and very thin soils. In Denmark, the aquifers (mainly Outwash Glacial Deposits and underlying Chalk) are also vulnerable to pollution, but the cover of drift deposits is generally thicker and more continuous.

Both regions experienced serious diffuse groundwater pollution from the late 1970s/early 1980s onwards, in Eastern England this was dominated by rising nitrate with some pesticides, and in Denmark the nitrate problem was less severe but pesticides were widely encountered (with, for example, 27% of VCS-Denmark waterwells exceeding 0.1 μg/l in 2020).

If groundwater degradation is to be confronted, integrated responses across the water–agriculture policy interface need to be defined, adopted, and implemented. The main obstacles to effective groundwater policy (invisible resource with delayed impacts, counterproductive agricultural subsidies, difficult measurement, and inadequate monitoring) must be confronted, and agreement reached on required measures. This may be complex as a consequence of their different scales of field research (from an aquifer system to an experimental plot), but is necessary before policy convergence between the related administrations is possible.

The diverse nature of challenges means that a combination of management measures is needed, from governance and regulatory provisions (such as groundwater conservation zones and pesticide registration) to land-user incentives for groundwater services (through agri-environmental stewardship and land leasing arrangements). In turn, the effectiveness of all these measures will depend on the prevailing legal framework and political will, the level of stakeholder awareness on groundwater sustainability issues, and the involvement of an informed and organised civil society with a clear long-term vision.

Denmark relies primarily on minimally treated groundwater for its supply of drinking water. At the same time, Danish farming is among the most intensive in the world, with major extensive use of N fertilisers and intensive livestock rearing. Since the Second World War, Denmark experienced a marked trend towards larger more intensive farms, with major growth in livestock rearing and food production now accounting for about 25% of private sector turnover.

The wide national variation of groundwater nitrate concentrations is notable (Figure 1), and excessive concentrations for drinking water use are mainly a serious problem in northern and north-eastern Jutland (the so-called ‘nitrate belt’), where abstraction is from the shallower and more vulnerable aquifers.

Pesticide pollution is much more widespread in Denmark. Although the cover of glacial deposits is generally thick and continuous, these do not appear to offer consistent protection against diffuse pollution by pesticides because of their high heterogeneity and preferential flow patterns (DEPA 2022).

The scale of the groundwater pollution problem is well known as a result of the Nationwide Monitoring & Assessment Programme for Aquatic & Terrestrial Environments (NOVANA), which has monitored groundwater quality thoroughly since 1989 (GEUS 2021a). The results show that certain pesticides (or their metabolites) are found in 51% of waterwells, with concentrations exceeding 0.1 μg/l in 15%. The statistics are worse in the younger shallower groundwater where pesticide compounds are found in 72% of monitoring wells, and exceed 0.1 μg/l in 39% (GEUS 2021a).

The most commonly detected pesticide compounds are desphenyl-chloridazone (DPC), a metabolite of chlorideazone which was banned in 1996, and dimethyl-sulfamide (DMS), a metabolite of dichlorofluanide and tolyflyanide banned in 2003 and 2010, respectively. Both compounds are a major challenge to the Danish water utilities, especially DMS since its removal presents problems with current water treatment technology, and must be considered a consequence of ‘past sins’. The fact that they are still causing problems is a testament to the long turnover times of groundwater in the aquifers used for drinking water production.

In Denmark, policy states that drinking water production should be based on clean groundwater and avoid advanced treatment. This ambitious drinking water policy has the implication of the need for land-use control and change. On the demand side, great efforts have been made to reduce gross per capita water consumption, primarily through the introduction of a ‘drinking water tax’ in 1988, and in about 30 years, household water consumption has decreased by more than 40%. Today average per capita use is only 101 l/d (DANVA 2020). This reduction in water demand has made it more feasible to seek source protection as the primary route to ensuring groundwater quality and has paved the way for a more sustainable utilisation of groundwater resources.

The threat from present pesticide use is addressed nationally through the Danish National Action Plan on Pesticides 2022–26 (DMoE 2022). The objective is to achieve a low level of pesticide use through a number of different measures including pesticide use authorisation, targeted field inspection, and provision of advice and guidance. As a part of pesticide authorization, the Danish Pesticide Leachate Assessment Programme has been established for early warning with the objectives of evaluating if approved pesticides or their metabolites leach to groundwater in troublesome concentrations and informing the authorities about the need for regulation (GEUS 2021b).

Groundwater pollution is also dealt with in Denmark on a catchment scale through national hydrogeological mapping for establishing site-specific groundwater protection zones (Thomsen et al. 2014). The mapping effort covers about 40% of the country, in essence those areas designated with particularly valuable groundwater resources. From 1999 to 2020, a total of about €320 million have been spent on national hydrogeological mapping financed by the drinking water tax.

When dealing with diffuse agricultural pollution at the catchment scale, the water utilities play a central role as the primary stakeholder for implementation of targeted measures. The present challenges of DPC and DMS are usually dealt with by abandoning affected waterwells, blending different sources, and establishing new wellfields. When the problems are especially severe, advanced drinking water treatment (with activated carbon) is gradually becoming more common, but only as an interim measure. The long-term goal is still the production of drinking water based on clean groundwater, which implies the need for land-use control and change in areas with vulnerable groundwater resources of strategic importance.

The decentralised structure of the Danish water-supply industry means that the approach to protecting vulnerable groundwater varies from water utility to water utility. Generally speaking, only a minority of water utilities have the capacity to cope with diffuse agricultural pollution through catchment-scale land-use measures. However, since they are also the largest, they represent a substantial proportion of the total national groundwater abstraction for drinking water. Smaller municipally owned water utilities and small waterworks usually only focus on groundwater protection through the establishment of wellhead protection zones, rather arbitrarily defined by a groundwater travel time of 1 year.

The impact of intensifying rainfed cereal production from the 1950s on the thin soils of Chalk downland (in the counties of Cambridge, West Norfolk, North Lincoln, and East Yorkshire), using manifold increases in applied nitrogen fertiliser, first became manifest through groundwater nitrate levels rising in excess of 50 mgNO₃/l in the 1970s (Foster & Crease 1974), and the processes involved were then further investigated (Foster et al. 1982; Lawrence et al. 1983).

Nitrate concentrations in the Chalk groundwater of Eastern England are still generally rising today, on average at about 0.4 mgNO₃/l/a, as a result of the major unsaturated zone delay. They are expected to peak by 2025 (Stuart et al. 2007; Wang et al. 2016), with the highest levels being reached in the drier areas of Eastern England with the largest proportion of arable land.

Groundwater pesticide contamination (both from arable farming practices and non-agricultural use) has also occurred in about 20% of water utility/company waterwells in excess of the stringent level of 0.1 μg/l set by the EU-Drinking Water Guideline of 1980 (Foster et al. 1993; Chilton et al. 2005; Lapworth & Gooddy 2006; Stuart 2011)). Arable farming under minimal tillage requires relatively large applications of herbicides, of which isoproturon, chloroturon, mecoprop, and glyphosate have figured prominently over the past 20–30 years. These compounds degrade only very slowly in groundwater systems compared to agricultural soils, with half-lives of 100s rather than 10s of days (Chilton et al. 2005). National monitoring data for 2004–05, which analysed for 20 pesticide compounds with a large percentage of samples from the aquifers of Eastern England, showed only a small percentage (up to 2%) of the 1,000–2,000 samples annually with individual compounds exceeding 0.1 μg/l (DEFRA 2007).

However, the most common and persistent pesticides occurring in Chalk groundwater are atrazine and simazine (and their metabolites), which have also been heavily used in non-agricultural weed control, and these were found in up to 5% of the 1,000–2,000 samples taken in the national monitoring survey of 2004–05 (DEFRA 2007).

The vulnerability of the Chalk aquifer recharge area to pollution from the land surface was studied in increasing detail – and is everywhere elevated for persistent contaminants with a very slow (decadal) rate of groundwater system response due to volumetrically high water retention in the unsaturated zone and generally deep water-table. In addition, fast preferential rates of downward flow can occur almost everywhere through fissures, but these are only activated under high man-made hydraulic loading and during occasional major rainfall events. In some cases this vulnerability reaches extreme levels due to rock solution and micro-karstification, notably along the edge of the Tertiary cover. Here high vulnerability to microbiological pollution exists and Cryptosporidium from animal excreta has been detected in water-supply sources, including a major Hertfordshire incident in 1998 (Morris & Foster 2000).

In Eastern England, the water utilities/companies (with just a few notable exceptions) left the management of diffuse agricultural pollution of groundwater to the Environment Agency (the national and local regulatory body), although it took more than a decade of joint field research by environmental and agricultural centres before the widespread threat of diffuse nitrate pollution was generally accepted (because agricultural scientists insisted that soil denitrification would predominate over nitrification and aqueous leaching in all soil types), and policies to combat it agreed (Figure 5).

The Environment Agency, after consultation with agricultural research centres, strongly advocated the elimination of autumn fertiliser applications (previously at levels of 50 kg/ha), with the exception of land under oil-seed rape (about 15% of the total) where applications should not exceed 20 kg/ha. This had the benefit of substantially reducing leaching losses from land intensively cultivated for cereal production but not to below 50mgNO₃/l, and led to widespread major investment by water utilities/companies on advanced water treatment facilities (Environment Agency 2014).

In 1990, the so-called NSA (Nitrate-Sensitive Area) Scheme was introduced with incentives for the conversion of limited land areas with shallow water-table to rough pasture or coppice woodland, but with the advent of the EU-Nitrates Directive (1991) this was superceded in 1996 by the declaration of NVZs (Nitrate Vulnerable Zones) over about 60% of Eastern England. In these zones, ‘best agricultural practice’ was to be followed to reduce nitrate including eliminating the application of N fertilisers to autumn-sown crops, reducing soil oxidation by direct drilling for seed-sowing, promoting autumn-sown (over spring-sown) crops or using winter cover-crops to reduce leaching losses during the wettest months. However, over extensive areas groundwater recharge with 60–80 mgNO₃/l continued to be experienced and the concentration in many water utility/company waterwells exceeded 50 mgNO₃/l. The response of water utilities/companies has been to invest in mixing different sources and in very costly nitrate removal plant, passing the additional cost to the water consumer. While gross per capita water demand has decreased somewhat in the last 20 years it still stands at 140–150 lpd per capita.

The regulation of pesticide use in the United Kingdom has been considerably strengthened, with the sale of atrazine and simazine being banned since 1993 for non-agricultural use and from 2005 for agricultural use, and more recently groundwater mobility and pollution risk have been introduced as criteria for the approval of new pesticide compounds (HSE 2021). The water utilities/companies in Eastern England have chosen to cope with groundwater pesticide pollution variously on an ad hoc basis by blending different sources, abandoning some waterwells, and advanced water treatment technology, and this has involved major capital investment and increased water-supply production costs (Environment Agency 2014).

Both water utilities and regulatory agencies across Europe, and indeed around the world, are faced with the major challenge of how to cope with the challenges presented by diffuse groundwater pollution from agricultural land-use practices. In the case of water utilities, in deciding whether to direct attention to controlling land-use practices or to rely on treatment solutions, they must take into account the institutional feasibility of land-use management, the relative cost of the options and the implications of using advanced treatment technology in terms of the major implied increase in the carbon footprint of water use.

A key conceptual question is whether water utilities should regard the groundwater catchment area to their waterwell sources as part of their supply infrastructure, and therefore as an integral part of their asset management, or whether they should leave land-use management issues to the corresponding regulatory agency. It seems to the authors that the latter approach is something of an excuse for not taking full responsibility for their assets!

In reality, to date, the response of water utilities to the considerable challenge of diffuse groundwater pollution by agricultural land-use practices has been diverse, and involved mixing different waterwell sources, taking polluted waterwells out-of-service, advance water treatment technology, developing new wellfields in areas of more favourable groundwater quality, and promoting local partnerships to reduce soil leaching losses. The lattermost approach requires a water utility with a strong commitment to protecting its groundwater assets and to promoting the necessary local partnerships and agreements to achieve this objective. Sometimes investment in advanced water treatment can reduce the willingness to get involved with promoting land-use control to protect groundwater quality. However, going down the route of advanced treatment in isolation will have major long-term implications in terms of greatly increasing the carbon footprint of drinking water supply.

The authors of this paper gathered together by participating in an IWA Webinar on Groundwater Management organised by International Water Association headquarters staff in advance of the IWA Congress in Copenhagen in September 2022.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.