Ag-PIE: A GIS-based screening model for assessing agricultural pressures and impacts on water quality on a European scale

https://doi.org/10.1016/j.scitotenv.2005.07.013Get rights and content

Abstract

Diffuse pollution of water resources from agricultural sources is a major environmental issue in the European Union, and has been dealt with by specific legislation: the Nitrate Directive of 1991 and the Water Framework Directive of 2000. These attempts to provide a coordinated approach to solving environmental problems require methods and tools for spatial analysis and modelling on a continental scale, with river basins being used as spatial units. This paper presents a screening model (Ag-PIE), developed in a GIS environment, for the assessment of pressures from agricultural land use and the consequent impacts on surface and groundwater. Ag-PIE has been applied at the European scale (EU15), with focus on nitrogen pollution from chemical fertilisers and manure. The model adopts a multi-criteria evaluation procedure applied to spatial data layers which represent the variety of factors affecting the pollution process. The DPSIR (Driving forces, Pressures, State, Impact, Responses) approach is applied to provide the modelling approach with a conceptual framework and to further analyse and communicate results. Ag-PIE is ultimately aimed at providing a tool making use of state-of-the-art geographical databases to support policy-makers at the European level. The scale of reference adopted is the river basin, in particular those that extend across national boundaries. The quality of the results obtained has been assessed against existing related studies and monitoring reports and by means of sensitivity analysis. Conclusions are driven by considering the potential of Ag-PIE in devising policy support and its strengths and weaknesses in view of identifying future research needs.

Introduction

Diffuse pollution of water resources from agricultural sources is a major environmental issue both in the European Union (EU) and abroad. Nutrients released in surface and ground-waters from cultivated fields and livestock production are the main source of concern, together with pesticides. The main European legislative act dealing with the protection of water resources from agricultural pollution, nitrogen in particular, is the Nitrate Directive (ND) (EC, 1991). More recently, the Water Framework Directive (WFD) (EC, 2000) has elaborated the contents of the ND into a broader concept of the sustainable management of water resources. These directives set common policy measures, giving member countries a free hand to implement them in the most adequate and efficient manner, by taking into account the specific local conditions and, in particular, the physical characteristics of the land.

Both the directives provide ambitious and rather demanding implementation plans. Unfortunately, the enforcement of the provisions of the ND have been met with delays and the measures taken are not always efficient (Grossman, 2000). Some of the difficulties in implementing the directive are associated with delineation of the “vulnerable zones” (areas of land which drain into the waters affected by pollution and which contribute to pollution) (Goodchild, 1998). Further, the approaches and implementation criteria of the ND differ from state to state. This makes it difficult to aggregate and compare local information and thus to provide an overview on an EU level. Trans-boundary river basin management, as required by the WFD, has become extremely challenging.

Nitrogen pollution from agricultural sources is characterised by remarkable temporal and spatial variability, depending on the interplay of the effects of human driving forces with environmental variables (climate, soil and topography). These phenomena differ in their scale and spatial features, which is a factor that should be accounted for when designing and implementing policy measures. Latacz-Lohmann (2001) clearly pointed out the need for targeting (i.e. tailoring) agri-environmental policies so that they apply to heterogeneous situations as a means to achieve better environmental and economic performances. Therefore, it is necessary to provide homogeneous and systematic information in support of such tailoring.

The interaction between the spatial distribution of pollution sources (due to economic activities of agricultural production implemented by the farmers) and the geographical diversity of the physical characteristics of agricultural land require the adoption of spatially distributed assessment methods. Such methods should account for the causal relationships between the pressures exerted by agricultural activities in terms of the generated pollution loads and the state of the environments under pressure, or rather their vulnerability.

As regards the assessment of pressures, the methodologies that are currently used for quantifying diffuse nitrogen losses differ profoundly in (i) their level of complexity, (ii) their representation of system processes and pathways, and (iii) resource (data and time) requirements (EUROHARP project, 2001). Deterministic computer models have been found useful at a local level where more data are available, whereas screening models require less data, and are more appropriate for large-scale analysis or where simplified evaluation is needed to guide further steps. Some studies (e.g. Navulur and Engel, 1996, McLay et al., 2001) indicate that screening procedures are an appropriate tool for making preliminary large-scale assessments of pollution. These screening procedures use data that are commonly available or that can be estimated and their application produces results that are relatively easy to interpret and to incorporate in policy or decision-making (Margane, 2003).

Regarding the assessment of vulnerability, several approaches have been proposed and implemented in contexts such as the designation of the ND vulnerable zones. Consider, for example, groundwater vulnerability screening models such as DRASTIC (Aller et al., 1985 cited in Navulur and Engel, 1996), SEEPAGE (Richert et al., 1992, cited in Navulur and Engel, 1996), GLA-Method (Hoelting et al., 1995 cited in Margane, 2003), PI-Method (Goldscheider et al., 2000 cited in Margane, 2003), etc. Examples of an application of the concepts of vulnerability and risk in environment management can be found in OECD, 1999, Nemeth et al., 1998, Meinardi et al., 1995, Lake et al., 2003, Dunn et al., 2003, EnRisk project, 2001.

The present study has adopted the concept of vulnerability as an index to assess the state of the environment, and the concept of impact as the result of a specific combination of pressures and vulnerability in a given location. The need to account for spatial aspects means that Geographic Information Systems (GIS) become fundamental tools not only for implementing and presenting results simulated by computer models, but also for manipulating spatial data for screening purposes (Hartkamp et al., 1999). GIS are increasingly used to process spatial data representing pollution factors and to study how vulnerable a given area is to groundwater pollution (see for instance Lake et al., 2003, Al-Adamat et al., 2003, McLay et al., 2001, Burkart and Feher, 1996) or the potential risk of surface water pollution (Kellogg et al., 1999, Dunn et al., 2003).

The complexity of the phenomena and the challenges associated with policy implementation require that the entire human and environmental system be framed within a common conceptual framework. Such a framework can be found in the approach proposed by the European Environmental Agency for environmental assessment and reporting (EEA, 1999), commonly known as the DPSIR framework (Driving Force – Pressure – State – Impact – Response). All the stages of policy-making and management processes can be formalised with the DPSIR approach, adopted also by the WFD, by identifying cause-effect links between the elements in the chain of human–environmental interaction. The method uses indicators to represent the elements of the chain, thus simplyfing the information which is conveyed to broad groups of stakeholders and the general public in short, clear messages, thus enhancing the transparency of decision-making (OECD, 2002). The indicators may become map layers in a spatial context. This approach has also been operationally implemented in decision support system (DSS) tools in the field of water resources management (Giupponi et al., 2004), also in a spatial context (Fassio et al., 2005), and is increasingly used by EU institutions because it complies with the planning and management processes of the WFD (EC, 2003). The DPSIR approach has also been presented as a fundamental component for the assessment of the effectiveness of environmental policies (EEA, 2001), such as the Nitrate Directive.

The great potential of the DPSIR approach in terms of general applicability and simplicity is counterbalanced by the fact that the definition of its components is rather generic and the identification of indicators and their attribution to the five nodes may be affected by subjectivity. Therefore, the present work makes use of the definitions and the approach officially adopted by the WFD Guidance for the analysis of Pressures and Impacts (EC, 2003). According to the cited document, Driving forces (or Drivers) are those anthropogenic activities that may have environmental effects (in our case agriculture), by exerting Pressures, which cause a change in the State of the environment (here chemical characteristics of surface and ground-waters). Impacts are identified as the environmental effect of the pressure (here water ecosystem modifications) and Responses are the measures taken to improve the state of the water body (e.g. limiting pollution discharges, developing best practices, etc.).

The degree of Impact on the environment of the agricultural Drivers is determined by the specific spatial combinations of P indicators (agricultural production systems) and S indicators (vulnerability of natural resources). The assessment of the changes of the State and of the Impacts (in terms of alterations of indicators) could then be used by the decision-maker to design adequate Responses targeted at the Driving Force, Pressures or the State.

The research presented herein is aimed at developing and analysing the applicability of a spatial screening model for policy support at the European level for the geographical assessment of diffuse pollution from agricultural sources: Ag-PIE (Agricultural Pressures and Impacts on European waters). The following section illustrates the methods developed for (i) assessing the effects of agricultural activities on surface and ground-water quality, and (ii) the model structure and procedure. We then present the results of the application of the Ag-PIE model on a European scale (EU151 area) and discuss them by analysing the possibility of their validation and of the management of uncertainties. The critical factors of environmental impact are then identified as a means to support the design of spatially targeted policies, and finally, conclusions are offered regarding the potential of the proposed tool in view of the current data available and the need for improvements in further research.

Section snippets

The assessment of agricultural impacts

The factors determining the phenomenon of nitrogen pollution of water from agriculture are diverse in their nature and origin — some result from human activities and some stem from the physical processes taking place in the natural environment. The interaction between these factors determines the environmental impacts. According to the EC Guidance on the analysis of Pressures and Impacts (EC, 2003, p. 36–37) nitrate concentrations in water bodies and eutrophication status were identified as the

Impacts on surface and ground-water quality

The maps of vulnerability, pressures and the resulting impacts on the river basin scale are presented in Fig. 2. The impact modelled allows to differentiate the magnitude of the phenomena and problems on a continental scale, and to distinguish between the watersheds where agriculture has a high impact on water quality and those where agricultural activities are not a major source of pollution. This information can be useful for policy-making on a large scale (EU, country, region) as it allows

Conclusions

The present work has drawn from previous research conducted in the same area (Fassio et al., 2005), by building a structured and reproducible model, Ag-PIE, which may be used in various contexts in support of agri-environmental analyses and policies for the EU15 area.

The main potentials and limitations of the proposed approach in the field of nitrate pollution have been presented in the previous sections. Those relevant for future utilisation of the model and for identifying research needs are

Acknowledgements

Research supported by the European Commission (EESD EVK1-CT-2002-57007). The authors gratefully acknowledge the suggestions by A.Fassio (FEEM, Venice) and C.Simota (Research Institute for Soil Science and Agrochemistry, Bucharest). The provision of spatial climate data by the ATEAM Project (Mitchell et al., in press) is also acknowledged with gratitude. C.Giupponi provided overall supervision on the work and guided the development of the screening model methodology. All computations have been

References (40)

  • L. Aller et al.

    Drastic: a standardized system for evaluation of groundwater pollution potential using hydrogeologic settings

    US EPA, Robert S. Kerr Environment Research Laboratory, Ada, OK

    (1985)
  • S.M. Dunn et al.

    Modelling nitrate losses from agricultural activities on a national scale

  • J.R. Eastman

    IDRISI Kilimanjaro GIS

    (2004)
  • EC (European Communities)

    Council directive concerning the protection of waters against pollution caused by nitrates from agricultural sources (91/676/EEC), 12 December 1991

    Off J Eur Communities

    (1991)
  • EC (European Communities)

    Directive of the European Parliament and of the Council of 23 October 2000 establishing a framework of Community action in the field of water policy (2000/60/EC), 22 December 2000

    Off J Eur Communities

    (2000)
  • EC (European Communities)

    Implementation of Council Directive 91/676/EEC concerning the protection of waters against pollution caused by nitrates from agricultural sources: synthesis from year 2000 Member State reports

    (2002)
  • EC (European Communities)

    Common strategy on the implementation of the water framework directive (2000/60/EC), guidance document N 3, analysis of pressures and impacts

    EC, Luxembourg

    (2003)
  • EC-JRC (European Commission Joint Research Centre)

    On line dataset: SeaWiFS images archive

    Institute of Environment and Sustainability, Inland and Marine Waters Unit

    (2000)
  • EEA (European Environmental Agency)

    Environmental indicators: typology and overview

    Tech Rep

    (1999)
  • EEA (European Environmental Agency)

    Reporting on environmental measures. Are we being effective?

    Environ Issues Rep

    (2001)
  • Cited by (48)

    • Risk and vulnerability assessment in coastal environments applied to heritage buildings in Havana (Cuba) and Cadiz (Spain)

      2021, Science of the Total Environment
      Citation Excerpt :

      As a result, the first risk studies were based on collections, archives or museums (Ashley-Smith, 1999; Lyall, 1988; Waller, 1994), while studies on monuments or historic cities were based on the assessment of the main risks, and only occasionally on a multi-hazard approach (Indirli et al., 2011; Mazzolani et al., 2009). Today, there are many applications and projects using integrated tools for hazard assessment and risk evaluation (Aronoff, 1989; Burrough and McDonnell, 1998; Giupponi and Vladimirova, 2006; Indirli et al., 2013; Indirli and Castellano, 2008; Mazzolani et al., 2009; Posen et al., 2006), and the use of Geographic Information Systems (GIS) offers efficient storage, management and analysis of spatial and non-spatial hazards (Accardo et al., 2003). Ramalhinho and Macedo (2019) made a review of main methodologies developed between 1999 and 2016 to analyze risks and vulnerability related to Cultural Heritage, from qualitative to quantitative models.

    • Assessing the potential impacts of a revised set of on-farm nutrient and sediment ‘basic’ control measures for reducing agricultural diffuse pollution across England

      2018, Science of the Total Environment
      Citation Excerpt :

      The pollutant sources included in SEPARATE comprise agriculture, urban areas, channel banks, direct atmospheric deposition, sewage treatment works, septic tanks, combined storm overflows, storm tanks and groundwater. Different screening tools are reported in the literature for both the UK (e.g. Comber et al., 2013) and elsewhere (e.g. Giupponi and Vladimirova, 2006; Brouwer and De Blois, 2008; OECD, 2012). SEPARATE was selected since it includes river bank erosion which can be an important source of sediment which was one of the water pollutants considered in this work.

    View all citing articles on Scopus
    View full text