Assessment and Reporting on Soil Erosion, Slides of Environmental Science

Download Assessment and Reporting on Soil Erosion and more Slides Environmental Science in PDF only on Docsity! Technical Report No.84 ASSESSMENT AND REPORTING ON SOIL EROSION Background and workshop report Prepared by: Anne Gobin, Gerard Govers, Katholieke Universiteit Leuven Robert Jones, Joint Research Centre Mike Kirkby, University of Leeds Costas Kosmas, Agricultural University of Athens July 2002 Project manager: Anna Rita Gentile European Environment Agency European Environment Agency 2 5 ANNEXES ..............................................................................................................................77 ANNEX I LIST OF PARTICIPANTS .................................................................................77 ANNEX II AGENDA.............................................................................................................79 ANNEX III BACKGROUND PAPERS PRESENTED AT THE WORKSHOP..........81 STATE-PLAY OF EEA WORK ON SOIL EROSION INDICATORS ...........................................................81 SOIL EROSION HOT SPOTS MAP FOR EUROPE ....................................................................................88 BACKGROUND .....................................................................................................................................88 DATA QUALITY ISSUES..........................................................................................................................90 THE DESIGN AND USE OF THIS MAP .......................................................................................................91 INTERPRETATION OF THE MAP ..............................................................................................................91 FUTURE WORK.......................................................................................................................................93 QUALITATIVE SMALL SCALE SOIL DEGRADATION ASSESSMENT DATABASES - THE GLASOD MAP ..........................................................................................................................................93 THE GLASOD MAP (1990) ...................................................................................................................93 FOLLOW-UP OF GLASOD / DERIVED INITIATIVES................................................................................96 METHODOLOGICAL DETAILS.................................................................................................................98 RESULTS OF THE ASSESSMENT ............................................................................................................100 INDICATORS OF SOIL EROSION AT THE ETC/SOIL ..........................................................................101 INTRODUCTION....................................................................................................................................101 THE INDICATOR CONCEPT...................................................................................................................101 DPSIR APPLIED TO SOIL EROSION.......................................................................................................102 SUMMARY ...........................................................................................................................................105 GISCO DATABASES AND TOOLS TO DERIVE DRIVING FORCE/PRESSURE INDICATORS FOR SOIL EROSION.................................................................................................................................106 INTRODUCTION TO GISCO DATABASES AND TOOLS ..........................................................................106 OVERVIEW OF DRIVING FORCE/PRESSURE INDICATORS PROPOSED BY EEA ......................................110 GISCO AND DRIVING FORCE/PRESSURE INDICATORS.........................................................................110 PROPOSED INDICATOR FRAMEWORK MODEL ......................................................................................111 REMARKS AND CONCLUSIONS.............................................................................................................111 REGIONAL ASSESSMENT OF THE IMPACT OF SOIL EROSION BY WATER ........................................112 SOIL EROSION INDICATORS OF STATE .................................................................................................112 THE REVISED DPSIR FRAMEWORK ....................................................................................................113 PROCESSES OF SOIL EROSION BY WATER ............................................................................................113 REGIONAL ASSESSMENT METHODS OF SOIL EROSION .........................................................................114 PROCESS MODELLING TO ASSESS REGIONAL SOIL EROSION: PESERA...............................................116 CONCLUSIONS .....................................................................................................................................119 DATA AVAILABILITY FOR SOIL EROSION INDICATORS AT EUROPEAN LEVEL ..............................119 DETERMINING THE CAUSES OF SOIL EROSION.....................................................................................119 MODELLING SOIL EROSION .................................................................................................................120 SOIL EROSION RISK ASSESSMENTS ......................................................................................................122 ENVIRONMENTAL INDICATORS FOR SOIL EROSION .............................................................................123 ANNEX IV SOIL EROSION GLOSSARY.......................................................................126 ANNEX V PROCESSES OF SOIL EROSION.................................................................128 SOIL EROSION BY WATER ....................................................................................................................129 6 ABBREVIATIONS CAP CORINE DPSIR DSR EEA EFMA EIONET ETC/S ETC/TE GLASOD NDVI OECD PESERA RUSLE UN UNCED USLE Common Agricultural Policy Co-ordination of Information on the Environment Driving forces-State-Impact-Responses Driving forces-State-Responses European Environment Agency European Fertiliser Manufacturer’s Association European Environment Information and Observation Network European Topic Centre on Soil European Topic Centre on Terrestrial Environment Global Assessment of Human-Induced Soil Degradation Normalized Difference Vegetation Index Organisation for Economic Co-operation and Development Pan-European Soil Erosion Risk Assessment Revised Universal Soil Loss Equation United Nations United Nations Conference on Environment and Development (Rio 1992) Universal Soil Loss Equation ACKNOWLEDGEMENTS Special thanks to the national experts who participated in the EEA technical workshop on indicators for soil erosion held in Copenhagen in March 2001; to Paul Campling at the Katholieke Universiteit Leuven for his help in the organisation of the workshop; and to Robert Evans, University of East Anglia and Jaume Fons, Autonomous University of Barcelona for their useful comments. Executive Summary This report has been prepared by the Katholieke Universiteit Leuven under contract to the EEA and is the final result of a working group on indicators for soil erosion. The working group was established by the EEA in order to progress with the work on soil in the interim period before the new ETC on Terrestrial Environment (ETC/TE) started in July 2001. In 2001 EEA carried out a peer review of its work on soil, with particular reference to the development of policy-relevant indicators and the identification of probable problem areas for soil degradation (the so-called ‘hot-spots’). The review was in particular focused on work on indicators for soil erosion and soil sealing and two associated technical workshops were held in March 2001 to facilitate this review. This report provides the background, analyses the work done by EEA on soil erosion and summarises the conclusions of the workshop on indicators for soil erosion, held in Copenhagen on 27-28 March 2001. The purpose of the workshop was to identify a set of recommendations concerning reporting on soil erosion (as part of the wider theme of soil degradation) that could then be considered for inclusion in the work programme for the new ETC on Terrestrial Environment. Soil erosion is a natural process, occurring over geological time. Most concerns about erosion are related to accelerated erosion, where the natural rate has been significantly increased by human activities such as changes in land cover and management. This report focuses on accelerated erosion caused by water. Runoff is the most important direct driver of severe soil erosion. Processes, which influence runoff, must therefore play an important role in any analysis of soil erosion intensity, and measures, which reduce runoff, are critical to effective soil conservation. In Europe, soil erosion is caused mainly by water and, to a lesser extent, by wind. In the Mediterranean region, water erosion results from intense seasonal rainfall on often fragile soils located on steep slopes. The area affected by erosion in Northern Europe is more restricted and moderate rates of water erosion result from less intense rainfalls falling on saturated, easily erodible soils. According to the GLASOD assessment, in Europe, excluding the Russian Federation, about 114 million ha or more than 17 % of the total land area is affected by soil erosion, of which more than 24 million ha or approximately 4 % show high or extreme degradation and nearly 70 million ha or 11 % are affected by moderate degradation. The various regions of Europe show different patterns, for example in the EU and EFTA countries the area subjected to soil erosion is about 9 % of the total land area. It increases to 26 % in the Candidate countries and to 32 % in the rest of Europe (excluding the Russian Federation). However, these findings are based on fragmented and non-standardised information. Soil erosion: a priority at the European Level In April 2002, the Commission adopted a Communication on soil protection, endorsed by the European Council in June 2002. The Communication considers soil erosion as one of the major threats to Europe’s soils and a priority for action. Increasing the awareness amongst scientists and policy makers about the problem of soil 10 1 Introduction 1.1. Scope of the report This report has been prepared by the Katholieke Universiteit Leuven (Catholic University of Leuven) under contract to the EEA and is the final result of a working group on indicators for soil erosion. The working group was established by the EEA in order to progress with the work on soil in the interim period before the new ETC on Terrestrial Environment (ETC/TE) started in July 2001. In 2001 EEA carried out a peer review of its work on soil, with particular reference to the development of policy-relevant indicators and the identification of probable problem areas for soil degradation (the so-called ‘hot-spots’). The review was in particular focused on work on indicators for soil erosion and soil sealing and two associated technical workshops were held in March 2001 to facilitate this review. A separate document was prepared for the workshop on soil sealing and the ‘hot-spots’ review (EEA 2002b). Soil erosion is a natural process, occurring over geological time and may be caused by water or wind. Most concerns about erosion are related to accelerated erosion, where the natural rate has been significantly increased by human activities such as changes in land cover and management. This document focuses on accelerated erosion by water. A workshop on assessment and reporting on soil erosion was held in Copenhagen on 27-28 March 2001. The purpose of the workshop was to identify a set of recommendations concerning reporting on soil erosion (as part of the wider theme of soil degradation) that could then be considered for inclusion in the work programme for ETC/TE. The report provides the background, analyses the work done by EEA on soil erosion (Part I) and summarises the conclusions of the workshop on indicators for soil erosion (Part II). 1.2. Background The EEA was established in Council Regulation 1210/90 in May 1990 and started its operations in Copenhagen in July 1994. The EEA mission is to contribute to the improvement of the environment in Europe and to support sustainable development through the provision of relevant, reliable, targeted and timely information to policy makers and the general public. This should enable the Community and Member States to take the necessary measures to protect the environment, to assess the results of such measures and to be supported with the necessary technical and scientific issues. The EEA mandate is to provide information to Community Institutions and member countries required to frame, identify, prepare, implement and evaluate sound and effective policies on the environment and to ensure that the public is properly informed. EEA main tasks are: 1. To report on the state and trends of the environment; 2. To establish, develop and make use of the European Environmental Information and Observation Network (EIONET); 3. To facilitate access to data and information supplied to, maintained and emanating from EEA and EIONET, together with access to other relevant environmental information 11 developed by other national and international sources. The role of the EEA, as defined by its mission and mandate, is therefore to provide policy makers and the public with quality information, and to do so through a range of products and services. The Agency works as a facilitator or bridge between member countries1, the Community Institutions (in particular Commission, Parliament and Council) and other environmental organisations and programmes to bring together, use, make available and thereby improve the quality of information on the environment relevant at the European level for policy making and assessment. This is done through basic activities, including the support to national monitoring, the gathering and storage of existing information and currently accessible and reliable data, the analysis and assessment of data to produce policy relevant information and indicators, the reporting of results to the policy makers and the dissemination of information to the general public (ENVISION model, monitor to reporting -MDIAR- core activities) (Gentile 1999a). The European Topic Centre on Soil (ETC/S)2 was established by EEA in 1996 with the objective to provide and develop information and data on soil aspects, covering all EEA member countries, in order to increase the understanding of soil as a natural resource, document soil degradation processes and improve the level of reliable and comparable information about contaminated sites, thus contributing to the development of the EEA Work Programme. ETC/S operated until December 1999. A new Topic Centre on Terrestrial Environment (ETC/TE) started operations in July 2001. ETC/TE is carrying out the work initiated by the ETCs on Soil, Land Cover and Marine and Costal Environment (terrestrial part of costal environment). On the basis of the results of the First EIONET workshop on soil (EEA 2001a,b) and a wider review of the EEA work on soil (October 1999), in the period 2000-mid 2001 the implementation of the work programme progressed through three working groups on indicators for: • Soil contamination (from local and diffuse sources); • Soil sealing; and • Soil erosion. This report is the final product of the working group on soil erosion. 1.3. Policy developments Since 2001 important progress took place at the policy level. In fact, the 6th Environmental Action Programme (6EAP) has introduced a new strategy on soil protection for the European Union. The programme, proposed by the European Commission in 2001, lays down the Community action programme for the period 2001-2010 in the field of the environment. The 6EAP recognises that ‘… Little attention has so far been given to soils in terms of data collection and research. Yet, the growing concerns on soil erosion and loss to development as well as soil pollution illustrate the need for a systematic approach to soil protection...’ 1 Since January 2002 EEA membership counts 29 countries, comprising the EU15, three EFTA countries (Iceland, Liechtenstein, Norway) and 11 of the 13 candidate countries (Poland and Turkey are expected to join shortly). 2 ETCs are consortia of organisations that are assigned to carry out specific tasks concerning an environmental theme. They help the EEA develop its Multi-Annual and annual working programmes. 12 Moreover, ‘… Given the complex nature of the pressures weighing on soils and the need to build a soil policy on a sound basis of data and assessment, a thematic strategy for soil protection is proposed...’ (European Commission 2001) In April 2002, the Commission adopted a Communication on soil protection, endorsed by the European Council in June 2002. The Communication considers soil erosion as one of the major threats to Europe’s soils and a priority for action. A communication on soil erosion, soil organic matter decline and soil contamination, containing detailed recommendations for future measures and action, has been planned. To facilitate this process, a conference on soil erosion and organic matter decline in the Mediterranean with the participation of the major stakeholders is being organised by the Commission and expected to take place in 2003 (European Commission 2002). In a long-term perspective, the implementation of the work on indicators discussed in this report, would certainly contribute to improve the information basis needed to prepare, implement and monitor a sound European strategy on soil, in line with the priorities set down in the 6EAP and the Communication on soil protection. 1.4. Objectives and methodology of the review The specific objectives of this report are the following: • provide a summary overview of EEA work on soil erosion indicators; • review the EEA European framework for the assessment and monitoring of soil and the proposed soil erosion indicators in relation to data availability and analytical soundness; • discuss the link between soil erosion indicators and land use or land use intensity; • review methods for assessing soil erosion at a regional scale; • present options for future development with particular reference to existing European data sources; and • present the results of the workshop on indicators on soil erosion. The methodology adopted in the review process consisted first of all in the evaluation of EEA work carried out by a group of experts and the preparation of a background report (included in part I). An analysis of existing approaches for a regional assessment of the extent of soil erosion in Europe was also carried out (see section 4). A selection of national experts was asked to evaluate the results of EEA work on soil erosion and invited to discuss the results of the evaluation at the workshop. Questions to guide the review were provided (see Annex II). The main items of the discussion and the conclusion of the workshop are summarised in Part II. 1.5. Soil erosion in Europe The main problems for soils in the European Union are irreversible losses due to increasing soil sealing and soil erosion, and continuing deterioration due to local and diffuse contamination. It is envisaged that Europe’s soil resource will continue to deteriorate, probably as a result of changes in climate, land use and other human activities. A policy framework is needed which recognises the environmental importance of soil, takes account of problems arising from the competition among its concurrent uses, both ecological and socio- economic, and is aimed at maintaining its multiple functions (EEA, 2000). 15 26 % in the Candidates countries and to 32 % in the rest of Europe (excluding the Russian Federation). 16 PART I ASSESSMENT AND REPORTING ON SOIL EROSION 2 A European framework for the assessment and monitoring of soil The degradation of the environment through soil erosion is an important concern for policy makers. Objective and measurable criteria with potential to compare between areas and monitor changes over time are needed to describe the condition and management of land resources and the pressures exerted upon the land. There is now a requirement for environmental protection agencies to periodically report on the state of the environment and particularly whether this is deteriorating, stable or improving. Agencies are dealing more commonly with a degrading environment, hence the search for ‘indicators’ that can quantify this degradation in some way. International organisations such as the EEA, OECD and UN have initiated programmes on developing measurable and policy-relevant agri-environmental indicators to assess and monitor progress in reaching sustainable development, as defined in Agenda 21 by the UNCED. 2.1. The assessment framework An update of the state-of-progress of the EEA soil work programme and the relevance of indicator development including the reporting system were presented at the EIONET Workshop on Indicators for Soil Contamination in Vienna, 18 – 19 January 2001 (EEA 2002a,b). The concept of multiple soil functions and competition is crucial in understanding current soil-protection problems and their multiple impacts on the environment. The EEA considers soil with its multiple ecological and socio-economic functions and multiple impacts as having a fundamental role in Europe’s environment (EEA 1999a). The ecological functions comprise production of biomass; filtering, buffering and transforming; gene reserve and protection of flora and fauna. The socio-economic functions include support to human settlements; source of raw materials, including water; and protection and preservation of cultural heritage. Soil degradation means loss or deterioration of its functions (Blum 1990). Soil losses due to erosion can be considered as irreversible in relation to the time needed for soil to form or regenerate itself. The OECD DSR-framework (Driving Force-State-Response) has established a holistic systems approach to include cause-effect relationships (OECD 1993). The OECD model has been extended by EEA to cover the causes (pressures) and the impacts on the environment (EEA 1999b; 2000). The DPSIR assessment framework shows a chain of causes-effects from Driving forces (activities) to Pressures, to changes on the State of environment, to Impacts and Responses (EEA 1999; 2000). DPSIR is based on the assumption that economic activities and society’s 17 behaviour affect environmental quality. The relationships between these phenomena can be complex. DPSIR highlights the connection between the causes of environmental problems, their impacts and the society’s response to them, in an integrated way. The DPSIR applied to soil resources is shown in Figure 2-1. Figure 2-1: The DPSIR framework applied to soil (EEA 2000). In addition to the DPSIR, EEA has defined the multi-function and multi-impact approach (MF/MI), based on the recognition of the role played by the soil multiple functions and the problems arising from the competition between these functions. (see Figure 2-2) Both DPSIR and MF/MI approaches are analytical tools for the definition of policy-relevant indicators to describe pressures placed upon soil resources, changes in the state of soil and impacts or responses by society to these changes, within the context of policy and soil resource management (Gentile 1999a). These approaches also provide a framework for the subsequent interpretation and assessment of the indicators. In environmental monitoring, indicators have been defined as ‘parameters, or values derived from parameters, which point to/provide information about/describe the state of a phenomenon/environment/area with significance extending beyond that directly associated with a parameter value’ (OECD 1993). 20 expressed in terms of on-site and off-site effects, respectively (Figure 3-4). The responses at the European level include CAP-reform, soil conservation measures and land use practices in accordance with sustainable development. However, a European policy framework on soil protection, similar to those already in place for air and water, does not exist. Moreover, there is no reporting mechanism in place to assess whether existing measures are leading to improvement of soil conditions or to gauge the level of implementation of existing legislation (EEA 2000).3 The assessment carried out through the DPSIR framework does not aim at understanding or analysing soil erosion as a process, but provides information to support policy-makers’ actions so that the necessary measures can be defined and the effect of current measures can be assessed. (Human population) Land development Agriculture* *Intensification Natural events Off-site Driving Forces Pressures State Responses Impact Desertification Convention Development of a European soil protection policy ? „Good agricultural practice“ - Land use practices in accordance with sustainable development - Local programmes on soil erosion consulting - ... SOIL LOSS On-site: SOIL DEGRADATION Physical deterioration Off-site: emission to air, water and land (De-forestation) (Forest fires) Land use practices Changes in crop yields On-site Loss of soil fertility Desertification Changes in soil functions Effects on other media: e.g. - water stress - eutrophication Economic aspects e.g. - impediment of traffik - disturbance of drainage Changes in soil functions Changes in crop yields Figure 2-4: The DPSIR Framework applied to Soil Erosion (EEA-ETC/S 1999). Note for publication department, please correct ‘traffik’ with ‘traffic’ 2.3. Is the proposed DPSIR framework adequate to comprehend soil erosion? The result of the application of the DPSIR and MF/MI assessment tools to soil erosion is the identification of a set of policy-relevant indicators. However, it has to be recognised that there is a huge difference between actual and potential soil erosion, which is not adequately reflected in the present framework (EEA-ETC/S 1999). Indicators describing the driving 3 In April 2002, the Commission adopted a Communication on soil protection, later endorsed by the European Council in June 2002. The communication considers soil erosion as one of the major threats to Europe’s soil and a priority for action (European Commission 2002; see also section 1.4). 21 forces and pressures may affect the risk of soil erosion, but they may not affect soil erosion in itself, which is also depending on physical parameters such as climate and relief. A mechanism is therefore needed to jointly estimate the potential and actual risk, based on links between the identified driving force and pressure indicators, and on an estimation or measurement of what is actually happening. Agricultural intensification is seen as the most important driving force (EEA-ETC/S 1999; EEA 2000). However, tourism and transport could be added to the list of driving forces. The effect they have in common is that they change the land cover, which is the major pressure indicator for soil erosion. This would lead to a revised DPSIR scheme, presented in Figure 2- 5. Figure 2-5: The DPSIR framework applied to soil erosion modified from EEA 2000 and EEA-ETC/S 1999. Note for publication department, please correct ‘Mass wasting’ with ‘Mass movements’ The general DPSIR framework lends itself to systems analysis and as such is very useful in describing the relationships between the origins and consequences of environmental problems. Obviously, the real world is more complex than can be expressed in simple causal relationships. Linkages between the different types of indicators are explored through the DPSIR chain. However, the linkages deserve further attention, not least to capture the dynamics of the system. Moreover, linkages within one type of indicators (e.g. pressures) are not explored, despite their repeatedly reported importance. The emphasis of the DPSIR assessment framework is on socio-economic related indicators, while physical indicators of pressure are not fully explored, nor explicitly mentioned. Climate 22 change is considered as a driving force but only in the sense that it relates to human activities. Important physical factors that influence soil erosion are topography, soil type, soil vulnerability and climatic factors (particularly rainfall). These factors cannot be separated from the identified pressure indicators. On the other hand, they are implicitly incorporated into indicators of state. A major problem with soil erosion is the temporal and spatial scale of reporting and the spatial extent to which the phenomenon occurs. Although problems of both spatial and temporal patchiness are well recognised in the various reports (EEA 2000; EEA 2001a), a more integrated approach of reporting seems recommendable. One solution could be to develop a regional model that allows for estimating the potential soil erosion risk, combined with periodical monitoring of actual soil erosion in selected test areas. The regional soil erosion model should express the links between the different biophysical and socio-economic factors, i.e. be process-based; establish various spatial and temporal resolution linkages; and, provide a nested strategy of focussing on environmentally sensitive areas which may require remedial measures to be taken. Sections 3 and 5 provide more details on the requirements for future regional soil erosion reporting in order to develop sound indicators of state. In the different reports made by EEA, it is recognised that a distinction ought to be made between on-site and off-site impacts of soil erosion. This distinction, however, already applies at an earlier stage in the DPSIR chain, namely at the stage of state indicators. Soil erosion can be measured in terms of actual sediment loss per unit area (on-site) or in terms of sediment delivery into streams or rivers (off-site). The DPSIR framework does not enable the identification of actors related to the perceived environmental problem. The identification of several actors related to the environmental problem requires a stakeholder analysis. Environmental problems can be identified and discussed by each group of stakeholders using participatory methods for eliciting the various aspects of the perceived problem. A general stakeholder analysis ultimately helps formulating policies for remediation and mitigation strategies. In conclusion, the DPSIR framework is an excellent approach onto which further extensions and strategies of reporting can be built. The framework sets a good basis for identifying the different factors influencing soil erosion, but doesn’t explicitly allow for the identification of actors in the DPSIR chain. 2.4. EEA typology of indicators applied to soil erosion The EEA identifies four different types of indicators (EEA 1999b): • Descriptive indicators, describing the actual situation in the DPSIR framework; • Performance indicators, comparing the actual situation with a specific set of desirable conditions in terms of a ‘distance to target’ assessment; • Efficiency indicators, expressing the relation between separate elements of the causal chain such as between environmental pressures and human activities; • Total Welfare indicators, measuring ‘sustainability’ in the form of an index (Green GDP or Index of Sustainable Economic Welfare), currently not within EEA’s mandate. Efforts related to soil erosion have concentrated on descriptive indicators within the DPSIR philosophy. Without a European policy framework on soil protection, however, little progress can be expected on the other three types of indicators. Sound advice on how to develop performance indicators on soil protection will be one of the challenges of the European Topic 25 2.6. Review Indicators for soil erosion should incorporate the following characteristics: • The indicators will be a measure of soil loss due to erosion as a result of climate, topography, soil properties, land cover and land management, • The extent and severity of both potential and actual soil erosion risk will have to be quantified and related to land cover changes, • The nature of soil erosion has to be assessed in order to evaluate the on-site loss and the possible off-site impacts. As accelerated erosion is a complex process, it is necessary to develop indicators that identify the causes. Physical factors that influence erosion rates include topography, soils, climate and land cover. Land cover is in turn influenced by the socio-economic environment and as such by anthropogenic activities, notably land use and management. Table 2-2 lists the EEA indicators for soil erosion with brief comments on the OECD criteria listed in Section 2.1. The first six indicators relate to pressures as a result of agricultural intensification. These pressure indicators all have in common that they are complex and not directly linked to the phenomenon of soil erosion. The identified indicators of state and impact are difficult or expensive to measure and the data are usually not readily available. Indicators of response are prevention and control measures, which are rarely in existence at present. A more comprehensive discussion follows in the next sections. 2.6.1. Indicators of driving forces and pressures According to the EEA (EEA-ETC/S 1999), the main driving force on soil that causes erosion in regions with potential and actual soil erosion risks is the intensification of agriculture. This is a complex indicator and it is related to different pressure indicators. The corresponding pressures are cost-effective but unsustainable land use practices, the use of machinery for the cultivation of enlarged fields, the overgrazing and other instruments of intensive land use practices (EEA-ETC/S 1999). Average field sizes (and increase of field sizes), combined with average farm size per region as well as the consumption of fertilisers and the number of grazing animals, give an indication of the intensification of agriculture. The intensification of agriculture is not necessarily directly related to soil erosion. The higher the degree of intensity of agricultural land use the higher may be the soil loss by water and wind erosion in potentially high erosion risk areas, but the reverse could equally be true. For example an intensive farming system employing soil conservation measures, such as terracing and cover crops, may result in less soil erosion than a more extensive system that does not involve conservation techniques. Intensive land use can be combined with efficient soil conservation measures. Section 4 concentrates on other aspects related to pressure indicators. One major remark is that the intensity of agriculture should never be evaluated alone in relation to erosion. Soil loss due to erosion is a result of climate, topography, soil properties, land cover and land management. Land cover also includes the natural vegetation. 26 Table 2-2: EEA indicators for soil erosion tested according to the OECD criteria. EEA Indicator Policy relevant Utility Analytical soundness Measurability Effect Comments Representa tive Easy to interpret Compara ble Scientific /Technic ally Data available Documen ted Updated Fertiliser use & trend Yes No Probably ??? Eurostat, OECD Yes Yes Complex Economic criterion, link variable Farm size & trend Yes Yes Sometime s Probably Nationally Yes periodical ly Complex Not linked directly Field size & trend Yes In part Yes Probably National/ regional Yes Yes Complex Data partially available Crop yield & trend Yes No Yes Yes National/ EU Yes Yes Complex Data for actual and estimated (CGMS) yields Net profit & trend Yes No Yes Probably National Yes Yes Not relevant Stocking rate & trend Yes No Yes No National/ EU Yes Yes Complex Dichotomy between intensive indoor and outdoor stocking Actual soil erosion Yes No ??? Yes Rarely available In part No Direct Extent not known, expensive to measure Delivery of sediment Yes No ??? Yes Difficult to measure In part No Direct Measurement difficult, source difficult to establish Removal of sediment No No No No ?? No No Direct Comprehensive measurements not possible Prevention (agriculture) Yes No Yes Yes Probably not In part No Direct Usually piecemeal Prevention (forest, natural) Yes No Yes Yes No In part No Direct Usually piecemeal Erosion control Yes Yes Yes Yes Rarely available In part No Direct Usually piecemeal 27 2.6.1.1. Consumption of fertilisers The proposed indicator is ‘the consumption of fertilisers per defined region (e.g. member state)’, measured in tonnes/ha. The consumption of fertilisers can give an indication of the intensification of agriculture (EEA-ETC/S 1999). Another positive aspect is that data on estimated consumption of fertilisers are available at national level from the European Fertiliser Manufacturer’s Association (EFMA) or via Eurostat/OECD. The reliability of the data used to calculate this indicator may be seriously questioned. The main source of information on fertilisers in Europe is the European Fertiliser Manufacturer’s Association (EFMA), see http://www.efma.org/. The data from EFMA are the production of fertiliser from the associated members. Then the EFMA uses data on imports and exports to calculate fertiliser use or consumption at the national level. For example, the current approach is: [Fertiliser Consumption {in a member state}] = [production] – [exports] + [imports] [2.1] To determine the actual fertiliser use by equation [2.1], certain adjustments should be applied to take account of losses (e.g. 10-15 %) and use outside general agriculture e.g. in market and domestic gardens (e.g. 10 %). However, fertiliser applications vary for different crops so it is not possible to predict the consumption of fertilisers using this approach without knowing precisely the spatial distribution of crops and local agricultural practices. The main conclusion is the higher the degree of intensity of agricultural land use, the higher the likely loss of soil through water and wind erosion in potentially high erosion risk areas (EEA-ETC/S 1999). However, fertiliser consumption data cannot be determined accurately enough to be used as an indicator for soil erosion at the scale required. Moreover, fertiliser applications may increase when using soil conservation measures so that soil erosion decreases. Together with consumption of fertilisers, average farm size [per defined region (e.g. member state) and its increase], average field size (and its increase), average crop yield [per area and its increase], average net profit [per area] and number of grazing animals give an indication of the intensification of agriculture (EEA-ETC/S 1999). This does not necessarily imply an increase in soil erosion. The factors that relate directly to erosion are soil type, topography, crop cover and precipitation. However, knowing the contribution of the agro-economic sectors to soil erosion is essential for the policy makers to be able to take the requisite measures and monitor their implementation, but the lack of good quality data hinders the development of suitable indicators in the short-term. 2.6.1.2. Average farm and field size The proposed indicators are ‘average farm size per defined region (e.g. member state) (and its increase)’ and ‘average field size (and its increase)’, both measured in ha. Data on farm and field size are available at national and European level. These data are periodically updated, with full farm surveys every 10 years and sample surveys of farm structure every 2- 4 years. However, these data are only averages on a large area basis e.g. member state and there is no demonstrable direct link between actual soil erosion and either farm or field size. As stated above, average farm size and average field size can give an indication of the intensification of agriculture. Furthermore, monitoring an increase in farm and field size 30 (EuroWaterNet), which could be a possible source for data on river sediments. A difficulty to consider is that data on sediment transport for selected rivers do not relate to the exact source of the sediment. The sediment loads in rivers can only give an indication of the erosion taking place over large areas. As an indicator for soil erosion, sediment delivery data are rarely accurate enough to be an independent indicator. EEA-ETC/S (1999), in fact, consider the transport of sediments as an indicator of impact. 2.6.3. Indicators of impact Indicators of impact could be divided into on-site and off-site impacts. On-site impacts in terms of loss of soil fertility are mostly compensated for by technical advances. On the other hand, off-site impacts are more easily measured and could be expressed in economic terms. 2.6.3.1. Removal of sediment deposits The proposed indicator relates to ‘expenditures for removals of sediment deposits in built up areas (traffic routes, houses)’. Data on remedial measures are rarely available at the national level, let alone at the European level. However, there are subsidies provided by the EU for remedial works via the CAP. Remedial measures usually follow major floods and should be linked to flood forecasting systems. 2.6.4. Indicators of response The comparison of soil erosion rates with, yet to be defined, soil loss tolerances for different regions would provide estimates of the impacts and the required response. 2.6.4.1. Conservation practices An important indicator of response is the expenditure for ‘Local agricultural programmes to enforce sustainable farming management systems (including the set-aside of arable land)’. These practices include contouring, terracing, strip cultivation, and subsurface drainage (Renard et al 1997). Other measures involve adoption of minimum tillage systems, planting cover crops (to reduce the duration of bare ground), and changing fundamentally the land use system (for example conversion form arable to pasture). Conservation practices have been demonstrated to considerably reduce soil loss through erosion in other parts of the world. Many of these practices increase plant cover and therefore directly reduce erosion. Many are also recognised as ‘good agricultural practice’. However, data and information on conservation practices are rarely collected systematically and stored centrally in Europe. Conservation practices are important in reducing or eliminating soil erosion but they are usually only adopted after soil erosion has been identified as a significant problem. 2.6.4.2. Mitigation strategies The indicator proposed is the ‘expenditures for special soil erosion prevention programmes, including forest fire protection’. Measures involve implementation of fire prevention systems and building of holding 31 reservoirs. Conservation practices are important in reducing or eliminating soil erosion but they are usually only adopted after soil erosion has been identified as a significant problem. Data and information on conservation practices are rarely collected systematically and stored centrally in Europe. 2.7. Options for the future: determining the risk of soil erosion From the review of the current indicators for soil selected by the EEA, it is concluded that, from a scientific and technical standpoint, the most appropriate state indicator is the area affected by erosion. However, because there is a serious lack of direct measurements of soil loss, by water and by wind, a surrogate parameter or indicator is needed. Conventional wisdom suggests that the area actually affected by erosion should be directly related to the area at risk from erosion, provided that the area at risk has been determined using an appropriate model of soil erosion, together with the necessary spatial data sets. Soil erosion takes place at the field scale, and the main problem is that the digital data sets used to quantify the factors causing erosion are usually too coarse (in terms of spatial resolution) to enable accurate estimation of soil losses at this scale. An important surrogate indicator of actual erosion is its risk. A risk is the chance that some undesirable event may occur. Risk assessment involves the identification of the risk, and the measurement of the exposure to that risk. The response to risk assessment may be to initiate categorisation of the risk and/or to introduce measures to manage the risk. In some cases, the risk may simply be accepted. In other cases, the priority will be to adopt a mitigation strategy. Such risk management, traditionally a significant activity in the commercial sector (e.g. the insurance industry) has now been adopted in the environmental protection field. Various approaches can be adopted for assessing soil erosion risk. A distinction can be made here between expert-based and model-based approaches. 2.7.1. Expert-based methods An example of an expert-based approach is the soil erosion risk map of Western Europe by De Ploey (1989). The map was produced by various experts who delineated areas where, according to their judgement, erosion processes are important. A limitation of this approach is that the author does not give a clear-cut definition of the criteria according to which areas were delineated (Yassoglou et al. 1998). Factorial scoring is another approach that can be used to assess erosion risk (Morgan 1995). The CORINE soil erosion methodology produced soil erosion risk maps with a resolution of 1 km² for Southern Europe (CORINE 1992), excluding Northern Europe. A relative ranking of soil erosion risk per area was obtained through the summation of individual erosion risk scores for each of the parameters rainfall, soil susceptibility, slope angle, slope distance, land use and prevention measures. The CORINE approach relies heavily on risk assessment by experts, and it remains difficult to assess the effect of changes in land use and/or climate on the erosion risk as no quantitative estimate of soil erosion is made. For the same reasons, it is not feasible to incorporate more detailed data, nor is it possible to evaluate the accuracy of the final result. More details are provided in section 6.3. Montier et al. (1998) developed an expert-based method for the whole of France. As with CORINE, the method is based on scores that are assigned to factors related to land cover (9 classes), the soil’s susceptibility to surface crusting (4 classes), slope angle (8 classes) and 32 erodibility (3 classes). An interesting feature of their method is that it takes into account the different types of erosion that occur on cultivated areas, vineyards, mountainous areas and the Mediterranean. This way, the interaction between soil, vegetation, slope and climate is accounted for to some extent. A problem with most methods based on scoring is that the results are affected by the way the scores are defined. In addition to this, classifying the source data in e.g. slope classes results in information loss, and the results of the analyses may depend strongly on the class limits and the number of classes used. Moreover, unless some kind of weighting is used each factor is given equal weight, which is not realistic. If one decides to use some weighting, choosing realistic values for the weights may be difficult. The way in which the various factors are combined into classes that are functional with respect to erosion risk (addition, multiplication) may also pose problems (Morgan 1995). Finally, as factorial scoring produces qualitative erosion classes, the interpretation of these classes can be difficult. 2.7.2. Model-based methods A wide variety of models are available for assessing soil erosion risk. Erosion models can be classified in a number of ways. One may make a subdivision based on the time scale for which a model can be used: some models are designed to predict long-term annual soil losses, while others predict single storm losses (event-based). Alternatively, a distinction can be made between lumped models that predict erosion at a single point, and spatially distributed models. Another useful division is the one between empirical and physical-based models. The choice for a particular model largely depends on the purpose for which it is intended and the available data, time and money. Jäger (1994) used the empirical Universal Soil Loss Equation (USLE) to assess soil erosion risk in Baden-Württemberg (Germany). De Jong (1994) used the Morgan, Morgan and Finney model (Morgan et al. 1984) as a basis for his SEMMED model. Input variables are derived from standard meteorological data, soil maps, multi-temporal satellite imagery, digital elevation models and a limited amount of field data. This way, erosion risk can be assessed over large, spatially diverse areas without the need for extensive field surveys. So far, the SEMMED model has been used to produce regional erosion risk maps of parts of the Ardêche region and the Peyne catchment in Southern France (De Jong 1994, De Jong et al. 1998). Kirkby and King (1998) assessed soil erosion risk for the whole of France using a model- based approach. Their model provides a simplified representation of erosion in an individual storm. The model contains terms for soil erodibility, topography and climate. All storm rainfall above a critical threshold (whose value depends on soil properties and land cover) is assumed to contribute to runoff, and erosion is assumed to be proportional to runoff. Monthly and annual erosion estimates are obtained by integrating over the frequency distribution of rainstorms. Several problems arise when applying quantitative models at regional or smaller scale. First, most erosion models were developed on a plot or field scale, which means that they are designed to provide point estimates of soil loss. When these models are applied over large areas the model output has to be interpreted carefully. One cannot expect that a model that was designed to predict soil loss on a single agricultural field produces accurate erosion estimates when applied to the regional scale on a grid of say 50 km pixels or coarser. One should also be aware of which processes are actually being modelled. For example, the well- known Universal Soil Loss Equation (USLE) was developed to predict rill- and inter-rill 35 provide temporal information and are proposed here as an important source of information for vegetation cover. The normalised difference vegetation index (NVDI) is often used as an indicator of vegetation growth determined by optical sensors. This index when compared during different periods of the year can indicate the vegetation cover change during the growing period of crops and natural vegetation. However, the introduction of the NDVI- index will only make sense if it is combined with regularly updated land use data, such as established in the CORINE land use map. 3.2. Review of the proposed indicators in relation to land use intensity The proposed indicators in relation to land use and intensity of land use partially satisfy the needs for assessing the soil erosion risk across different agro-ecological regions. As it is mentioned in the EEA-ETC/S working report (EEA-ETC/S 1999), data on soil erosion made available by the EU countries are highly variable. Furthermore, the application of the Universal Soil Loss Equation has the disadvantage of requiring data such as vegetation cover at a high temporal and spatial resolution. The proposed indicators of intensification of agriculture can be considered as a good basis for assessing soil erosion risk but they require further expansion with other factors such as other human activities that affect land cover, existing policies for the protection of soils and the degree of enforcement of such policies. It has to be considered that in hilly cultivated areas, tillage erosion is usually much more important than wind and water erosion. In the last decades there is an increasing awareness that the erosion processes which are primarily responsible for the severe degradation occurring in topographically complex landscapes, cannot be attributed to wind or water erosion only, but is caused mainly by tillage erosion. Tillage erosion is a progressively down slope translocation of soil caused mechanically by tillage implements, and it is considered as a main cause of land degradation and land abandonment in hilly cultivated areas throughout the European Union countries. Areas that have been introduced to cultivation during this century are being abandoned at an increasing rate in the last decades due to dramatic decrease of the land productivity resulting mainly from tillage erosion. The availability of heavy powerful machinery has favoured deep soil ploughing with high speeds, and in directions usually perpendicular to the contour lines, causing displacement of huge amounts of soil from upper landscape positions and deposition to lower landscape positions. Tillage erosion exposes subsoil, which may be highly erodible by wind or water, and fills in ephemeral flow areas, acting as a delivery mechanism for water erosion. Data from various sources shows that tillage erosion can account for up to 70 % of the total loss in cultivated areas (Van Muysen et al. 1999). 3.3. Options for the future on relating land use and land use intensity to soil erosion The rate of soil degradation is dependent upon the rate of land cover degradation, which in turn is influenced by both adverse climatic conditions and land use management changes. Vegetation cover, type of land use, and intensity of land use are clearly important factors controlling the intensity and the frequency of overland flow and surface wash erosion. Vegetation cover may be altered radically by man within a short time, but physical and biological changes within the soil, affecting erosion rates, may take longer periods. Type of 36 land use and land use intensity is affected by various environmental and socio-economic factors, therefore, indicators for soil erosion risk assessment should be related to these factors. 3.3.1. Climate characteristics affecting vegetation The characteristics of the climate of an area that can affect vegetation growth and vegetation cover and therefore soil erosion are rainfall, both amount and intensity, and aridity. These climate characteristics are easily available for all regions of the EU. Erosion data collected in various sites along the Mediterranean region shows that the amount of rainfall has a crucial effect on soil erosion. Generally, there is a tendency of increasing run-off and sediment loss with decreasing rainfall in hilly Mediterranean shrublands, especially in the region where rainfall is greater than 300 mm/year. Below the 300 mm annual rainfall limit, run-off and sediment loss decreases with decreasing rainfall. Rainfall amount and distribution are the major determinants of biomass production on hilly lands. Decreasing amounts of rainfall combined with high rates of evapotranspiration drastically reduce the soil moisture content available for plant growth. In areas with annual precipitation less than 300 mm and high evapotranspiration rate, the soil water available to the plants is reduced drastically and the soil remains relatively bare favouring overland water flow whenever rainfall events happen. Aridity is a critical environmental factor in determining the evolution of natural vegetation by considering the water stress, which may occur and cause reduced vegetation cover. In the Mediterranean region, vegetation presents a great capacity of adaptation and resistance to dry conditions, and many species can survive many months through prolonged droughts with soil moisture content below the theoretical wilting point. Aridity can greatly affect plant growth and vegetation cover, particularly annual plants. Under dry climatic conditions in areas cultivated with rainfed cereals, the soil remains bare favouring high erosion rates under heavy rainfalls following a long dry period. Closely related to climatic characteristics is the topographic attribute, slope aspect. Slope aspect is considered an important factor for land degradation processes. Aspect affects the microclimate by regulating the angle and the duration at which sunrays strike the surface of the soil. In the Mediterranean region slopes with southern and western facing aspects are warmer, and have higher evaporation rates and lower water storage capacity than northern and eastern aspects. Therefore, a slower recovery of vegetation and higher erosion rates are expected in southern and western aspects than in northern and eastern aspects. As a consequence, southern exposed slopes usually have a persistently lower vegetation cover than northern exposed slopes. The degree of erosion measured along south-facing hill slopes is usually much higher (even twice higher) than in the north-facing slopes under various types of vegetation cover. 3.3.2. Vegetation characteristics affecting soil erosion Indicators of soil erosion related to the existing vegetation can be considered in relation to: (a) fire risk and ability to recover, (b) erosion protection offered to the soil, and (c) percentage plant cover. Forest fires are of the most important causes of land degradation in hilly areas of the Mediterranean region. Fires have become very frequent especially in the pine-dominated forests during the last decades with dramatic consequences in soil erosion rates and biodiversity losses. The frequency of fire occurrence is lower in grasslands, and 37 mixed Mediterranean macchia with evergreen forests. Also, Mediterranean pastures are frequently subjected to man-induced fires in order to renew the biomass production. The Mediterranean vegetation type is highly inflammable and combustible due to the existence of species with high content of resins or essential oils. Conversely, it is known that vegetation has a high ability to recover after fire and the environmental problems related to fire normally last for only a limited number of years after the fire occurred. There are several factors, which affect the process of the recovery, apart from the fire and site characteristics, which can be both natural and anthropogenic. Years of unusual drought or sites that cannot be affected from the moist sea winds during summer show a slower rate of recovery. Human interference, such as livestock grazing or change in the land use pattern may damage irreversibly the recovering vegetation. Particularly important are the time intervals between subsequent fires. The ability of the ecosystems to recover is not unlimited and a fire frequency beyond a certain threshold can also lead to a permanently degraded state. This can be due both to the nutrient and seed bank depletion and to increased erosion. These processes have already led to severe degradation of extensive hilly areas in the Mediterranean region. Vegetation and land use are clearly important factors controlling the intensity and the frequency of overland flow and surface wash erosion. Extensive areas cultivated with rainfed crops such as cereals, vines, almonds and olives are mainly confined to hilly lands with shallow soils very sensitive to erosion. These areas become vulnerable to soil erosion because of the decreased protection by vegetation cover in reducing effective rainfall intensity at the ground surface. Almonds and vines require frequent removal of perennial vegetation using herbicides or by tillage. In fact, soils under these crops remain almost bare during the whole year, creating favourable conditions for overland flow and soil erosion. Erosion data measured along the northern Mediterranean region and the Atlantic coastline located in Portugal, Spain, France, Italy and Greece in a variety of landscapes and under a number of land-uses representative of the Mediterranean region (rainfed cereals, vines, olives, Eucalyptus plantation, shrubland) showed that the greatest rates of run-off and sediment loss were measured in hilly areas under vines. Areas cultivated with wheat are sensitive to erosion, especially during winter, generating intermediate amounts of run-off and sediment loss especially under rainfalls higher than 380 mm per year. Olives grown under semi-natural conditions, particularly where there is an understorey of annual plants greatly restrict soil loss to negligible values. Erosion in shrublands increased with decreasing annual rainfall to values in the range 280-300 mm, and then decreased as rainfall decreased further. Several hilly areas under natural forests around the Mediterranean region have been reforested with exotic species such as Eucalyptus. Such soils are undergoing intense erosion as compared with soils left under natural vegetation. However, the measured rates of erosion under Eucalyptus are relatively lower than those measured under vines, almonds and cereals. Soil erosion data measured from various types of vegetation and certain physiographic conditions showed that the best protection from erosion was measured in areas with a dominant vegetation of evergreen oaks, pines and olive trees under semi-natural condition. Pines have a lower ability to protect the soils in southern aspects due to the higher rate of litter decomposition and the restricted growth of understorey vegetation. Deciduous oak trees offered relatively low protection from erosion in cases where the falling leaves did not cover the whole soil surface. The main factors affecting the evolution of the Mediterranean vegetation, in the long term, are related to the irregular and often inadequate supply of water, the long length of the dry 40 Fires have become frequent in the pine-dominated forests during the last fifty years. Most of the fires can be attributed to the carelessness of people. The majority of fires occur in areas with high xerothermic indices and moisture deficits. Soil dryness and wind speed are the principal factors of fire evolution. The areas affected by forest fires are increasing dramatically throughout the Mediterranean basin. In the period from 1960 to 1975, the average rate of burning was 200 000 ha/yr, from 1975 to 1980 470 000 ha/yr, and 660 000 ha/yr from 1981 to 1985. Erosion rates seem to be enhanced after fires. The increased erosion rates are only partly due to the removal of vegetation. More important seems to be the forming of an impermeable subsurface layer, which decreases infiltration rates, while causing a quick saturation of the upper layers leading to overland flow and erosion. In contrast aggregate stability increases after fire and that increase is more pronounced after severe burns. The management quality can be related to the intensity of land use and to the applied measurements for environmental protection related to certain policies. Land use can be classified according to several criteria leading to hierarchies of land use types. The number of criteria employed is dictated by the level of detail desired as well as by the availability of the proper data. The principal classification criterion is the main purpose for which land is used. Based on this criterion, the land use types can be distinguished as following: • Agricultural land (cropland, pasture or rangeland) • Natural areas (forests, shrubland, bare land) • Mining land (quarries, mines, etc.) • Recreation areas (parks, compact tourism development, tourist areas, etc.) • Infrastructure facilities (roads, dams, etc.). Using the above classification of land use on land parcels, allows the intensity of land use and the enforcement of policy on environmental protection to be assessed. The intensity of land use of a cropland can be evaluated on the basis of the frequency of irrigation, degree of mechanisation of cultivation, application of fertilisers and agrochemicals, types of plant varieties used, etc. In the degree of mechanisation, the following characteristics should be included, type of tillage instrument, plough depth, wheel speed of the tractor, direction of tillage operation, etc. The intensity of land use of a pastureland can be defined by estimating the sustainable stocking rate (SSR) and the actual stocking rate (ASR) for the various land parcels under grazing. The ratio of ASR/SSR can be used to assess the intensity of land use. In natural areas such as forests, shrubland etc., the intensity of land use can be defined by assessing the actual (A) and sustainable yield (A/S). Then, the intensity of land use can be classified based on the ratio A/S. The intensity of land use for areas with mining activities can be defined by evaluating the measurements undertaken for soil erosion control such as terracing, vegetation cover, etc. Then, the intensity of land use can be classified based on the evaluated degree of land protection from erosion. In areas undergoing active recreational use such as skiing, motor rallies etc., the intensity of land use can be evaluated by defining the actual and the permitted number of visitors per year (A/P). Then the land use intensity can be classified based on the ratio A/P. Particular attention must be given to the policies related to soil protection such as policies supporting terracing, policies favouring extensive agriculture, etc. Of course their 41 effectiveness depends on the degree to which they are enforced. Therefore, rating of policies can be based on the degree to which they are enforced. Hence, the information must be collected on the existing policies and their implementation /enforcement. 3.4. Conclusions of review of indicators in relation to land use Many of the soil erosion indicators proposed by the EEA relate to land use and land use intensity. Land use and vegetation cover, in general, are the major input in defining actual soil erosion risk. It is therefore advocated to use regularly updated land cover data, such as established in the CORINE land use map, in combination with remotely sensed products such as the Normalised Difference Vegetation Index (NDVI) in order to capture seasonal variations in land cover. The proposed indicators of intensification of agriculture can be considered as a good basis for assessing soil erosion risk but they require further expansion with other factors that affect land cover, existing policies for the protection of soils and the degree of enforcement of such policies. Other human activities that affect land use and determine land use intensity include infrastructure, recreation, mining activities or forest management. Land cover is affected by different environmental and socio-economic factors, such as precipitation, vegetation type and management quality, which require monitoring in order to understand the complex relationship with soil erosion. Concerning management, tillage erosion is a prime example of human induced erosion. 42 4 Regional assessment of the extent of soil erosion by water A regional soil erosion assessment, providing an estimate of the area affected by soil erosion and the expected magnitude, is needed in order to make objective comparisons that may provide a basis for further environmental analysis, economic statements or policy development. Suitable assessment methods need to be developed to this purpose. This section deals primarily with assessing the extent of soil erosion by water as this is the most important form of soil erosion in Europe. Four alternative methods for carrying out regional assessment are compared The GLASOD maps and HOT SPOTS map can be classified as methods based on distributed point data, while the RIVM and CORINE maps can be classified as factor or indicators based maps. A description of the processes of soil erosion, crucial to an understanding of the following sections, can be found in Annex V. 4.1. Alternative assessment methods Assessments of soil erosion at a European scale are required for a number of reasons: 1. To make objective comparisons of the soil resource, taking account of past erosional degradation 2. To estimate the average rate of erosion to estimate the rate of loss of soil resource and its economic cost 3. To estimate the probability and distribution of severe erosion events, to evaluate the implications for loss of production and offsite deposition 4. To provide an objective basis for allocation of resources for remediation, mitigation or more detailed research and assessment 5. To assess the impact on the soil resource of future climate and/or land use change, due to global warming, possible policy changes and economic conditions. Assessment of soil erosion may be based on a range of methodologies. Some of these are based on the collection of distributed field observations, others on an assessment of factors, and combinations of factors, which influence erosion rates, and others primarily on a modelling approach. All of these methods require calibration and validation, although the type of validation needed is different for each category. There are also differences in the extent to which the assessment methods identify past erosion and an already degraded soil resource, as opposed to risks of future erosion, under either present climate and land use, or under scenarios of global change. 4.1.1. Distributed point data On important form of erosion assessment is from direct field observations of erosion features and soil profile truncation. Erosion features consist of rills and gullies, some of these ephemeral, and associated deposition in swales and small valleys. Soil profiles may show local loss of upper horizons, or burial by deposition from up-slope. Deposited material may provide dateable material, which can indicate when erosion occurred, but much of this evidence is cumulative over the period since cultivation began, or in some cases over the 45 Figure 4-1: Potential versus actual erosion risk as estimated by the CORINE methodology (CORINE 1992). 4.2.1. Methodology For one of these priority topics, soil erosion, a new methodology was developed, which provides a factor-based assessment of risk Figure 4.2). It was recognised that there was no suitable Europe-wide map of erosion, and that existing maps differed widely in methodology and scales of assessment. The methodology used was based on a simplification of the Universal Soil Loss Equation (USLE), a regression based model, for which there is a massive data base for US conditions, but little systematic data for Europe. E=K R S P V [4.1] Where E is the annual soil loss K is soil erodibility, R is rainfall erosivity, S is the slope length factor P is the crop management practice factor and V is the vegetation cover factor. 46 Figure 4-2: Methodology for CORINE Soil Erosion assessment (CORINE 1992). Land Cover, V 1 for fully protected 2 for not fully protected Actual Soil Erosion Risk, EA 0 for EP.V = 0 1 for EP.V = 1-2 2 for EP.V = 3-4 3 for EP.V >=5 Slope angle, S 1 for <5% 2 for 5-15% 3 for 15-30% 4 for >30% Potential Soil Erosion Risk, EP 0 for K.R.S = 0 1 for 0 < K.R.S < 5 2 for 5 < K.R.S < 11 3 for K.R.S > 11 Erosivity, R 1 for F.B<4 2 for 4<F.B<8 3 for F.B>8 Bagnouls-Gaussen Aridity Index, B 1 for Σ(2Ti-pi) =0 2 for 0<Σ(2Ti-pi)<50 3 for 50<Σ(2Ti-pi)<130 4 for Σ(2Ti-pi)>130 Fournier Index, F 1 for Σpi 2/Σp <60 2 for 60<Σpi 2/Σp<90 3 for 91<Σpi 2/Σp<120 4 for Σpi 2/Σp>120 Erodibility, K 0 for ST.SD.SS =0 1 for 0< ST.SD.SS <3 2 for 3< ST.SD.SS <6 3 for ST.SD.SS >6 Soil Stoniness, SS 1 for >10% 2 for <10% Soil Depth, SD 1 for >75 cm 2 for 25 – 75 cm 3 for <25 cm Soil Texture, ST 0 for Bare rock 1 for C, SaC, SiC 2 for SsCL,CL,SiCL,Lsa 3 for SaL, L, SiL, Si The USLE (Equation [4.1]) is intended to provide an estimate of average annual erosion loss in tons per unit area.The CORINE soil erosion methodology is a considerable simplification 47 of the USLE (CORINE 1992; Briggs and Giordano 1995). Erodibility is estimated from soil texture, depth and stoniness. Erosivity is estimated from the Fournier and Bagnouls-Gaussen climatic indices. Slope gradient is included, but without a slope length correction, and vegetation and crop management are collapsed into two categories of protected, and not fully protected, using data from the associated CORINE land cover database. These factors are combined to estimate three categories of potential and actual soil erosion risk. Potential risk excludes vegetation factors, and so identifies land at risk, while Actual risk includes the vegetation factor to indicate whether the potential is being realised. A map showing the assessment for southern Europe is provided in Figure 4-1 and the overall scheme is summarised in Figure 4-2. 4.2.2. Advantages and limitations The CORINE soil erosion assessment has the great advantage of simplicity, in that it provides a clear forecast, on an objective basis, for the whole of the area studied. The method is based, at least in principle, on a well-established technology, the Universal Soil Loss Equation, which has been very widely used, both in America and worldwide. Being based on a factor method within a 1 km GIS base, the method can be applied at a resolution, which allows discrimination within regional areas. The method correctly identifies the areas of the Mediterranean, which have the highest risk of erosion. As a product of its time, it has considerable merit, and could be improved with the more detailed land cover classification now available, providing refinement in the USLE land cover and crop management factors. However, the USLE, although still widely used on account of its simple structure, is now widely regarded as a post-mature technology, and cannot therefore recommended as the best basis for estimation of erosion risk. Furthermore mapping of USLE forecasts on national scales, for example for Italy (van der Knijff et al 2000) shows wide discrepancies between CORINE and USLE forecasts, so that CORINE may not even correctly represent the USLE factors. The CORINE report concedes (p.92) that ‘future development of this work would allow more sophisticated models of soil erosion to be used. Particularly on improving the factors used in the procedure, notably in the calculation of erosivity and soil erodibility, and in the classification of land cover’ (CORINE 1992). On a qualitative basis, comparison of the Erosion maps of southern Europe appear to show too great a dependence on the climatic factors in determining erosion risk, with relatively less weight given to important factors of erodibility and land cover. For use in the future, the CORINE assessment also has the limitation that it is restricted to southern Europe, whereas present needs for erosion data apply to the whole of the European area. 4.3. The ‘Hot Spots’ approach An analysis and mapping of soil problem areas (Hot spots) in Europe was published in the EEA-UNEP joint message on soil (EEA 2000). This addresses a number of soil problems, and only soil erosion aspects are reviewed in this section. The purpose of the study was to support the joint message on the need for a pan-European policy on soil, identifying ‘hot spots’ of degradation in Europe and examining environmental impacts leading to change and particularly degradation of soil function. The work involved compilation and analysis of data available at the EEA, together with additional data from the scientific literature. These data were incorporated into a GIS (ArcView) for manipulation and display. The hot-spot maps aims to present a kind of ‘spatial indicator’ that would enable the 50 module generates a water erosion risk index based on three main parameters: terrain erodibility, rainfall erosivity, and land use pressure. Figure 4-4: Water erosion vulnerability for 2050, according to the baseline scenario5 by RIVM (RIVM, 1992). ARG note: possibly change map and reference to the Turn of the century report page 190, maps 3.6.1 and 3.6.2 The methodology is described below and summarised in Fig. 4-5: 1. Terrain erodibility is based on soil type and landform, which are regarded as constant parameters. Land form is classified into general types (flat, undulated, mountainous, etc.) by using the difference between minimum and maximum altitudes for each grid cell, using the 10 minute grid elevation dataset of the Fleet Numerical Oceanography Center (FNOC) which provides 9 points per 50 km grid cell. Soil type is derived from the FAO Soil Map of the World and is composed of soil depth, soil texture, and bulk density. General averages for these characteristics are supplied by the WISE soil profile dataset. 2. Rainfall erosivity is represented by the month with the maximum rainfall per rain-day. This is considered to be indicative of rainfall erosion potential. Data on precipitation 5 In the last EEA state of environment report (EEA 1999a) an increase in the risk of water erosion was expected by the year 2050 in about 80 % of EU agricultural areas, as an impact of climate change. The increase would mainly affect the areas where soil erosion is currently severe. These results were produced jointly with the Commission, based on ‘business-as- usual’ socio-economic and energy developments which did not assumed that the Kyoto targets would be met (pre-Kyoto EC energy scenarios). 51 and number of wet days are derived from the IIASA climate database for mean monthly measured climate variables from an array of weather stations for the period 1931 to 1960. Precipitation is considered a dynamic variable, while the number of wet days is assumed to remain constant. 3. The potential erosion risk derived from these two factors is then converted to actual erosion risk by a land cover factor, representing the degree of protection afforded by various land covers (agricultural crops) from land cover maps. Natural vegetation with a closed canopy (e.g. forests) is assumed to provide optimal protection (no risk) and natural vegetation with a more open structure (e.g. shrubs) is assumed to provide sub- optimal protection (low risk). Land cover maps for the IMAGE model are derived from several sources including Olson’s land cover database and statistical information from FAO. 4.4.2. Advantages and Limitations The main advantage of the RIVM approach lies in its potential for integration with other environmental factors within an integrated model of the physical and economic environments, and the IMAGE model used is not evaluated here. Nevertheless these advantages cannot be fully realised unless the underlying model modules are themselves of an acceptable standard. The RIVM soil erosion model is a factor model, like CORINE, but, although initiated 6-8 years later, is in many ways a still more simplified approximation to the imperfect USLE model. If Figure 4.5 is compared with Figure 4.2, the similarities and differences are immediately evident. It may be seen that the soil erodibility takes a similar form to CORINE or USLE, with components for soil type, and a simplified gradient and index. The rainfall erosivity component is seen as an inadequate representation, which contains neither the theoretical basis underlying USLE nor the fair empirical alternatives provided in CORINE. Only land use provides an improvement on CORINE, due to the availability of better land cover data than was available early in the CORINE project. The RIVM method exploits the potential, inherent in any physically based or factor based assessment, of providing scenario analysis, through the inclusion of two dynamic components, the monthly rainfall totals (affecting erosivity) and land cover (affecting the assessed actual erosion risk). 52 Landform (from 10’ DEM) Relief range (for 9 point s) Soil Type (from FAO map) Dept h Text ure Bulk Densit y Rainfall Erosivity (IIASA) Max ( r rain-day) Potential Erosion Risk (IMAGE) Crop prot ect ion Low for open shrubland 0 for closed canopy forest Actual Erosion Risk Rain pe Land Cover Figure 4-5: Summary of RIVM methodology for water erosion assessment The RIVM approach is therefore seen to share some of the advantages of all methods which use distributed data sources, by providing an objective assessment across the European area. However, neither the 50 km resolution nor the implementation of the factors contributing to erosion is seen as providing a state-of-the-art assessment. 4.5. The GLASOD approach The main objective of the GLASOD project was to strengthen the awareness of decision makers on the risks resulting from inappropriate land and soil management to the global well- being. To achieve this, the United Nations Environment Programme (UNEP) commissioned the International Soil Reference and Information Centre (ISRIC), in 1988 to coordinate a worldwide programme in cooperation with a large number of soil scientists throughout the world to produce, on the basis of incomplete existing knowledge, a scientifically credible global assessment of the status of human-induced soil degradation within the shortest possible time frame. The task was subcontracted to correlators in 21 regions to prepare, in close cooperation with national soil scientists, regional soil degradation status maps, and these regional maps were correlated to provide the GLASOD world map of soil degradation. It is important to recognise the limited aims of the project, and to observe that GLASOD is the only approach which has, to date been applied at a world-wide scale. It is based on responses to a questionnaire, which has been sent to recognised experts in all countries (Oldeman et al 1991). It thus shares with the Hot Spots approach dependence on a set of expert judgements, but can provide very little control or objectivity in comparing the standards applied by different experts for different areas. The information and data on erosion and physical degradation in the Dobris assessment (EEA 1995) are based on an updated version of the European part of the Global Assessment of Soil 55 models. However, it was the first comprehensive global overview on soil degradation, which created awareness and highlighted the need for a more objective approach and for validation. Updates for specific regions have been made under the SOVEUR and ASSOD programmes (see Annex III, workshop paper by Van Lynden). 4.6. Comparative assessment of the four methodologies None of the four approaches reviewed here achieves state-of-the-art forecasting for soil erosion risk assessment across Europe (nota bene: an interpolation with a colour for a region where no observation was made is also a forecast). Because soil erosion events are associated with the incidence of storms, which are patchy in both time and space, site data must be widespread and long-continued to allow effective interpolation between available sites. Thus methods based on questionnaire surveys (GLASOD) or erosion measurement sites (Hot Spots) are likely to be inadequate on their own. In addition, differences between expert assessments and measurement methods reduces the comparability between the limited data available. Methods based on factors or indicators have the immediate benefit of accessing distributed data sources that are available at a European scale in electronic form (GIS). These include climate data, DEMs and soil maps. All of the mapping methods appear to use implicit or explicit reference to at least some indicators, particularly to soil classifications, but only CORINE makes explicit use of an adequate range of relevant indicators. However, CORINE is an implementation, which is imperfect for historical reasons of data availability, of a model (USLE) which is now no longer considered as state of the art. For these reasons, although it perhaps gives the best indication of the Europe-wide distribution of soil erosion of the four methods surveyed, it is now in need of replacement, and appears not to represent expert opinion of variations in erosion rate within each national region. 4.7. Options for the future It is clear that the widespread availability of GIS data for key controlling variables strongly favours a factorial or modelling base for future assessments of soil erosion. The difficulties associated with a modelling approach should not, however, be under-estimated. It is essential that a suitable model should: 1. Represent the state-of-the art in current understanding of soil erosion 2. Combine sufficient simplicity for application at a European scale with a proper incorporation of the most important processes 3. Have the potential for downscaling to field or plot scales where explicit validation can be made with field monitoring data, to make full use of experimental sites available Current thinking on modelling (COST623 2001) recognises the importance of runoff forecasting as a critical control on erosion loss. Simple runoff models are based on a runoff threshold or infiltration equation approach, and vary in complexity from the RDI model (Kirkby et al 2000) to the USDA WEPP (Nearing et al 1989) model. There is a trade-off between a simple model, which can be applied across a continuous range of parameters, for each cell within a European grid (as in the EC PESERA project); and a more complex model, applied for a finite number of parameter steps, the permutations of which are then repeated at many sites across a region (the MIR approach proposed by Brazier et al 2001). In either case, there is then the additional need to ensure that there is adequate investment in validation 56 against existing field data, although recognizing its variations in quality and methodology. Present-day soil erosion models have substantially aided insight into erosion processes, but are designed to assess soil erosion risk at small spatial units. In addition, these localised studies may not be representative of the continental and regional scales required by policy makers to set up an adequate soil conservation strategy. Moreover, it is often technically and financially infeasible to acquire the necessary input data to run detailed soil erosion models for decision-making at regional, national or pan-national level. For application at a broad regional scale, current models are severely limited by their high data demand and, in many cases, by a focus on individual events rather than on long-term averages or cumulative impact. This prevents the application of the best American models, such as WEPP (Nearing et al 1989) and KINEROS (Woolhiser et al. 1990) or other EU-funded models, such as the EUROSEM, EUROWISE and MEDALUS (MEDRUSH) models. The CORINE and USLE- derived models (RUSLE etc) are more appropriate in their data needs, but all are now recognised as lacking a physical basis which can be linked, more or less explicitly, with current concepts and research in soil erosion, and which offers the possibility of direct provision of physical and socio-economic scenarios. The 5th framework project ‘Pan-European Soil Erosion Risk Assessment’ (PESERA) will produce a regional model with a physical basis that can be applied to larger areas and can be used for scenario analysis and impact assessment. Earth observation techniques and the increased use of Geographic Information Systems have greatly improved the availability and methods to process and analyse spatial data. In concert with the improved understanding of soil erosion processes, the development of a spatially distributed process-based model to assess soil erosion risk over large areas is therefore the next challenge. In the face of an inevitable uncertainty, the concern will be to safeguard the model’s robustness based on a well-developed strategy of sensitivity analysis. Measured soil erosion data will play a crucial role in evaluating the model through quality assurance in the absence of any measurements. The model to be developed will produce quantitative results with a known reliability, and can be continuously upgraded with more accurate or detailed data upon their availability. The latter will evoke the somewhat under-estimated challenge of reconciling the model with high- volume data sets. More details on the PESERA project can be found in Annex III (Ann Gobin’s and Mike Kirkby’s presentation). 4.8. Conclusions and recommendations on implementation of regional assessments A number of recommendations can be made from the assessment of existing methods in this section. The most immediate is that there is scope and need for an improved assessment method, since all show serious shortcomings, and only a moderate level of agreement about the areas most seriously affected by soil erosion in Europe. The scope for a new assessment is based on the emergence of better models at appropriate scales, which can build on the data and expertise developed through the CORINE project, to develop a physically based forecast for the distribution of water erosion across Europe. The need for a new assessment is based on the large variation between current maps, which shows no clear consensus on the areas most at risk. Additional recommendations relate to the specification of erosion risk, which is defined in significantly different ways for the various assessment methods. It is suggested that evidence of historical erosion should be used to modify soil databases, and as a gross qualitative 57 indicator that an area is susceptible to erosion under certain circumstances (which may no longer apply). All of this information should be included in an assessment of the existing soil resource, and this is considered to be separate from an assessment of soil erosion risk. Soil erosion risk refers to the expectation of future loss, under both present conditions and under different climate (due to global change) and land use (due to economic circumstances, global change or policy implementation). This can most usefully be expressed in two ways. First as an estimate of long-term average rates of soil loss, and second as the loss expected in an extreme event (for example with 100 years average recurrence interval). These assessments can then be directly related to the long-term loss of soil resources in relation to present soil depths, and to the likely costs of locally severe off-site deposition and pollution. 60 cropping from year to year. 5.4. Discussion on questions What is soil erosion? It was recommended to concentrate on present-day soil erosion for policy purposes. Because of erosion's patchiness, rates are only meaningful for very small areas. Policy-makers are interested in a European-wide assessment of the problem at present and in the future. This requires a regional assessment in terms of soil erosion risk. Mapping actual erosion will be a very time-consuming and costly operation. Moreover, the recognised patchiness in time and space will always call for continuous updates. A risk assessment will enable a transparent and objective comparison between regions. The underlying model in a risk assessment translates what experts use into mathematical algorithms. However, a mapping instruction to map out actual erosion features at a detailed scale could be an option where more details are required on the actual state of the problem and where funds are available to undertake this expensive operation. One very important remark is that a programme to monitor soil erosion across different agro- ecological regions and under different land uses should underpin both mapping exercises and regional soil erosion risk assessment methods. Only then a sound approach is ensured of estimations and mapping features that are directly validated and compared with measurements. Moreover, measuring campaigns may lead to new insights and therefore to better mapping and risk assessments. Indicators of state and impact are the most important. However, factors underlying the causes of soil erosion such as pressure indicators should be clarified and communicated to the policy-maker. It is important to link each indicator to the general policy-framework. Headline indicators and sub-indicators should be identified and prioritised. A major concern is the link between different indicators of one category (e.g. driving forces) that is not expressed nor explored within the DPSIR framework. Agriculture in general is a very important driving force for soil erosion. An example for some of the less-favoured areas in the Mediterranean showed that with increasing subsidies stocking rates increased and resulted in overgrazing and subsequently more erosion. A similar scenario is foreseeable if farmers are compensated for soil erosion. In a situation of financial compensation for soil eroded land, incentives for farmers to practise soil conservation will be lacking. Care should be taken in formulating the necessary remedial measures and encourage farmers in practising them. 61 6 Regional and spatial assessment methods of soil erosion and data availability 6.1. The GLASOD map The GLASOD-map produced by ISRIC was the first effort to produce a global assessment of human-induced soil degradation on the basis of incomplete expert knowledge within the shortest possible time frame. This approach provides an overview at a global scale of human induced soil degradation, and can be used to identify hotspots and awareness raising for international policy makers. Major critiques on the map relate to the methodology and are reflected in a strong correlation between country boundaries and erosion risk. The GLASOD approach has been further developed by other programmes such as ASSOD (assessment of soil degradation in South and South-Eastern Asia, van Lynden and Oldeman 1997) and SOVEUR (assessment of soil degradation in Central and Eastern Europe, Batjes 2000) that are linked to GIS and database technologies. But GLASOD reliance on qualitative data means that the approach should not be adopted in isolation. Another suggested approach is to use the 1:1 million scale Soil Map of Europe as a base for a rapid GLASOD-type assessment of erosion, i.e. what is the type of erosion and its extent within a soil map unit and what are the causes of that erosion. A major disadvantage of expert mapping is that the policy-maker does not know what the underlying criteria were to produce the map (did the expert use soils, land use or a combination, etc.). 6.2. The Hot spots map The Hot Spots map (EEA 2001) is an empirical approach using measured data and expert opinion. This approach was adopted in view of the difficulty modelling approaches have in dealing with erosion’s spatial and temporal variability, and the generally poor job these models make of modelling gully erosion. Three categories are presented: zones (expert opinion); hotspot areas (based on De Ploey map); and locations (published erosion rates). An obvious disadvantage is that there is a lack of reliable data to give an adequate picture of erosion hotspot locations across the whole EU. Additionally, because of erosion's spatial patchiness, it is problematic to link erosion rates measured at specific locations with the severity of erosion in the hotspot areas. Moreover, spatial links between the different hotspot areas are difficult to establish. Therefore the usefulness of the Hot Spots map to policy- makers was questioned. The problem with the hot spots map is not its aim, i.e. to bring out where erosion is, or is most likely to occur, but its scale. Small areas, e.g. soil-landscape units cannot be brought out at this scale. However, a framework such as a 1:1M, or preferably a 1:250 000, soil/land use map could form the basis for assessing erosion on which hotspots could be portrayed, i.e. very often erosion will equate with particular soil/land use associations. Such a map could be similar to the 'actual' and 'potential' erosion maps produced for England and Wales which classify soil/land use associations from very low, low, medium, high to very high risk. It would bring together both expert views as well as quantitative work. Such an approach would bring together both the GLASOD and Hot Spots methodologies. 62 An evaluation of the Hot-Spot map was carried out by EIONET in Spring 2001. The results of the evaluation are published in EEA 2002b. 6.3. Regional assessment of the extent of soil erosion by water A comparison of existing maps for soil erosion assessment at a European scale was made. Four specific approaches – GLASOD, Hot Spots, RIVM and CORINE, used by EEA to obtain a European-wide assessment of soil erosion, were related to two assessment methods (Distributed point data and Factor or indicator mapping). Regional process modelling (RDI model and PESERA model) was presented as a suitable alternative for future regional erosion risk assessment. A description of the RDI and PESERA model is given in Annex III (regional soil erosion risk assessment by Anne Gobin and Mike Kirkby). It was pointed out that the four approaches used by EEA each served specific but different purposes. For instance, the Hot Spots map was aimed at locating soil erosion problem areas, whereas the RIVM map illustrated the impacts of global change on soil erosion. However, the common objective of all these specific approaches (including regional process modelling) is a regional assessment of soil erosion. Some of them consider soil erosion risk (RIVM; CORINE), while others consider actual erosion (GLASOD, Hot-Spot map). The process modelling method has the advantage of producing an indicator of state with the possibility for analysing different scenarios and assessing impacts (i.e. estimate what may happen in the future). The major objective of scenario analysis is to reduce soil erosion through policy-making. Clearly, the policy-maker has a major impact on land cover and land management through various land use policies. Any changes in these two factors affect soil erosion. The focus remains on indicators that are relevant to human activities. Modelling efforts should be thoroughly validated against erosion measurements, and a clear distinction should be made between modelled erosion risk and present-day erosion rates. Rainfall intensity is a crucial input to any soil erosion model and is incorporated in the PESERA model through rain distribution as a surrogate (Kirkby et al. 2000). All current erosion modelling approaches have severe limitations in capturing erosion's spatial variability. Moreover, any model based on the USLE will not include gully erosion, which as Poesen et al. (1996, 1998) have shown, can be a major contributor to total erosion in (at least some parts of) Europe. The PESERA model was presented as a potential solution to future regional erosion risk assessment. Three different views were adopted among the workshop participants concerning regional erosion assessment: (1) representation of real measurements, (2) expert judgement and (3) modelling approaches (whether factorial or physically based). The first two groups represent anti-modelling views. There was a general consensus, however, that both measurements and expert judgement remain a vital part in factorial or process modelling of regional erosion. It was the role or weight that is given to the different components in the process of regional assessments that remained a point of discussion. 6.4. General discussion on regional/spatial soil erosion indicators For policy purposes, there is a need to define a method which could be used to assess the present state of soil erosion but also to predict future responses. This calls for the definition of an indicator and a calculation procedure. The use of expert-based maps versus indicator or model-based maps were discussed in detail. 65 be taken with the interpretation of off-site indicators of impact (see section 5.2). Meaningful on-site indicators are more difficult to develop. Crop productivity springs to mind as a clear on-site effect. However, relating actual yield to soil loss is extremely difficult in European high-input agricultural systems. Moreover, it will have to be monitored over a long time period. In Mediterranean regions, productivity should be confronted with the available water capacity calculated over soil depth. Loss of soil fertility or soil quality could be considered, but are difficult to measure. Particularly for wind erosion, costs of re-seeding could be estimated. Data on land management changes (e.g. tillage practices) and in particular conservation practices should be collected in a systematic manner. This could be realised in conjunction with the IACS system. This type of indicator accentuates the response rather than the impact of soil erosion. 66 PART III RECOMMENDATIONS FOR FURTHER WORK 8 Recommendations to the EEA All recommendations are related to accelerated soil erosion (i.e. where the natural soil erosion rate has been significantly increased by human activities that cause changes in land cover and management). In a first instance, the EEA should focus on soil erosion by water and adding soil erosion by wind or tillage erosion in a later phase. The set of recommendations follow the main chapters of the report. They relate to the following categories: general, DPSIR framework, proposed indicators, land use and soil erosion indicators, and regional erosion assessment. 8.1. General recommendations The following general recommendations are related to the general reporting and networking activities. 1. Since soil erosion has impacts on several media, in particular on water quality, working links should be developed with other ETCs and specifically with the ETC on Water. The water framework directive recognises the relevance of agriculture as a major source of water pollution. 2. Working links with groups of experts contributing to the development of international initiatives - such as the COST Action 623 on 'Soil Erosion and Global Change', the European Society for Soil Conservation, IGBP-GCTE Focus 3 'Soil Erosion Network' (COST623 2001), the EUROSTAT projects IRENA and LUCAS - should be maintained. 3. Institutional links with data providers should be strengthened if the EEA is to provide policy-makers and the general public with information on the state of the environment. A general complaint was that data, and particularly statistical data, exist but are often not accesible. 8.2. Recommendations related to the DPSIR Framework: The DPSIR framework is an excellent approach onto which further extensions and strategies of reporting on soil erosion can be built. The following recommendations are made to the EEA in an attempt to extend the framework: 1. Although the DPSIR framework does not aim at understanding or analysing soil erosion as a process, policy-makers could immensely benefit from exploring underlying factors that influence soil erosion. These factors include topography, soil, climate, land cover (including vegetation), land use and land management. 2. Particularly physical indicators should be fully explored and explicitly mentioned in the DPSIR assessment framework. Climate change is considered as a driving force but only in the sense that it relates to human activities. Important physical factors that 67 influence soil erosion are topography, soil type, soil vulnerability and climatic factors (particularly rainfall). These factors should not be separated from the identified pressure indicators. At the same time, headline indicators and sub-indicators should be identified and prioritised. 3. All factors that change land cover, land use and land management should be included as driving forces. At present, only agricultural intensification is seen as the most important driving force (EEA-ETC/S 1999; EEA 2000). A revised DPSIR scheme, presented in Figure 2-5, has therefore been proposed, but could certainly be elaborated further upon. Examples of driving forces to be included are human population, land development, tourism, transport, natural events and climate change. 4. The revised DPSIR framework (Figure 2-5) presents land cover change and precipitation as the most important pressure indicators of soil erosion, as they are seen to be directly influencing the degree of soil erosion. 5. The general DPSIR framework lends itself to systems analysis and as such is very useful in describing the relationships between the origins and consequences of environmental problems. Obviously, the real world is more complex than can be expressed in simple causal relationships. Linkages between the different types of indicators are explored through the DPSIR chain. However, the linkages deserve further attention, not least to capture the dynamics of the system. Moreover, linkages within one type of indicators (e.g. pressures) are not explored, despite their repeatedly reported importance. 6. There is a huge difference between measured erosion, actual erosion risk and potential erosion risk. Indicators describing the driving forces and pressures may affect the risk of soil erosion, but they may not affect soil erosion in itself at present. A mechanism is therefore needed to jointly estimate the potential and actual risk, based on links between the identified driving force and pressure indicators, and on an estimation or measurement of what is actually happening. 7. In the different reports made by EEA, it is recognised that a distinction ought to be made between on-site and off-site impacts of soil erosion. This distinction, however, already applies at an earlier stage in the DPSIR chain, namely at the stage of state indicators. Soil erosion can be measured in terms of actual sediment loss per unit area (on-site) or in terms of sediment delivery into streams or rivers (off-site). 8. At present, there is no reporting mechanism in place to assess whether existing measures are leading to improvement of soil conditions or to gauge the level of implementation of existing legislation. This could be a focal point of action. 9. The DPSIR framework does not enable the identification of actors related to the perceived environmental problem. The identification of several actors related to the environmental problem of soil erosion requires a stakeholder analysis, which ultimately helps formulate sound policies for remediation and mitigation strategies. 8.3. General recommendations related to the proposed indicators Recommendations related to the indicators proposed by EEA-ETC/S (1999; 2000) (see also table 2-1) are presented below. A separate section is devoted to the indicators of state (8.5). A number of recommendations are also provided, which are related to land use issues in the indicators for soil erosion (see section 8.4). 1. Driving forces, other than agricultural intensification, should be included (see above). 2. Driving forces or pressures should never be evaluated alone in relation to erosion. In order to understand the complexity of accelerated erosion, it is necessary that at least some of the indicators identify the causes of soil erosion. Physical factors that 70 4. Field observations are invaluable as soil erosion indicators of state. However, the impossibility of making truly objective comparisons between and often within areas calls for standardised approach to record and particularly map the observations. 5. In conjunction with soil erosion measurements and observations, data on climate, topography, soil and land use should be carefully documented for each observation or measurement. The erosion type, scale of measurement/observation, study period should be well document. This requires the set-up of a comprehensive database, including meta-data. 6. A Europe wide monitoring network for soil such as proposed by EEA (2001b) should include soil erosion, covering the most affected areas (hot-spots). A standardised approach to record soil erosion should be defined. 7. Questionnaire based mapping approaches provide quick results for creating awareness, but should be avoided in the future whilst not rejecting field observations and measurements. 8. The temporal and spatial patchiness of soil erosion favours a risk analysis approach in order to make comparisons between regions and to complement field measurements and observations. The underlying model in a risk assessment should ideally translate experts’ knowledge into mathematical algorithms. The widespread availability of GIS data for key controlling variables strongly favours a factorial or modelling base for assessments of soil erosion. 9. Factorial models are useful for identifying the extremes of low and high erosion, but less satisfactory in identifying the gradation between the extremes. There are difficulties about combining different factor ratings into a single scale, about the individual weightings and about the assumed linearity and statistical independence of the separate factors. A process modelling approach is therefore recommended in case the full spectrum of soil erosion has to be assessed. 10. The difficulties associated with a process modelling approach should not be under- estimated and a suitable model should (a) represent the state-of-the art in current understanding of soil erosion, (b) respond explicitly and rationally to changes in climate and land use (c) combine sufficient simplicity for application at a regional scale and (d) relate coarse scale forecasts to measured erosion rate data so that explicit validation can be made with field monitoring data, to make full use of experimental sites. The process modelling method has the advantage of producing an indicator of state with the possibility for analysing different scenarios, which in turn enables the formulation of soil conservation policies. The PESERA project has adopted this modelling approach. 11. Modelling efforts should be thoroughly validated against erosion measurements, and a clear distinction should be made between modelled erosion risk and present-day erosion rates. 12. A programme to monitor soil erosion across different agro-ecological regions and under different land uses should underpin both mapping exercises and regional soil erosion risk assessment methods. Only then a sound approach is ensured of estimations and mapping features that are directly validated and compared with measurements. Moreover, measuring campaigns may lead to new insights and therefore to both better mapping and risk assessments. 13. Erosion literature commonly identifies ‘tolerable’ rates of soil erosion, but these rates usually exceed the rates, which can be balanced by weathering of new soil from parent materials, and can only be considered acceptable from an economic viewpoint. Tolerable soil loss rates should be developed but at the same time carefully evaluated by experts. 71 9 Conclusions The DPSIR framework is an excellent approach onto which further extensions and strategies of reporting can be built. A revised framework presents changes in land cover and precipitation as the most important pressure indicators of soil erosion. The DPSIR framework sets a good basis for identifying the different factors influencing soil erosion, but doesn’t explicitly allow for the identification of actors in the DPSIR chain nor does it explicitly distinguish the causes of soil erosion. Driving forces and related pressure indicators other than ‘agricultural intensification’ should be included. However, their relationship with soil erosion is complex. Physical factors that cause erosion should be included, i.e. topography, soils, climate and land cover, and their interaction with pressures should be analysed. The identified indicators of state and impact are difficult or expensive to measure and the data are usually not readily available. Indicators of response are prevention and control measures, which are rarely in place at present. Generally, it was concluded that the indicators should be developed according to the following properties and procedures: quantitative, objectively calculated, validated against measurements and evaluated by experts. Land cover type and change, land management and land use are the best pressure indicators for soil erosion. Land cover type and change can be monitored by combining CORINE LC with earth observation derived indices. In addition, land use and management information can be derived from NUTS together with the Farm Structure Survey data from Eurostat. The statistical data should be spatialised and disaggregated to the maximum possible. A regional assessment using a combination of modelling, expert estimates and other methods should be developed in order to provide a general view and identify the hot-spots areas where to undertake a detailed soil erosion monitoring programme. Regional soil erosion assessments enable estimates of the area that is affected by soil erosion and the expected magnitude in a particular area, both of which are required to formulate sound soil protection policies. Indicators of state should reflect all four strategies of regional soil erosion assessment, i.e. distributed point data, expert mapping, factor mapping and process modelling. The four different methods described in this report are not mutually exclusive and each provides a different emphasis. Erosion rate measurements and field observations provide an unambiguous measure of actual erosion, where they exist. However, apart from the time and expenses involved, spatial interpolation is not justified due to the sporadic distribution and episodic occurrence of soil erosion. Factorial approaches provide a measure of erosion risk and can only be recommended for identifying the extremes of low and high erosion, but not for the gradation between the extremes. A process modelling method is recommended for modelling soil erosion risk in relation to climate and land use changes. Field campaigns are necessary and databases should be made with erosion measurements, field observations, data on underlying factors influencing erosion (climate, topography, soils and land use) and related meta-data (period of record, erosion type, etc.). 72 10 References Alcamo, J. (Ed) (1994). IMAGE 2.0: Integrated Modelling of Global Climate Change. Kluwer, London. Arnoldus, H.M.J., 1978. An approximation of the rainfall factor in the Universal Soil Loss Equation. In: De Boodt, M. & Gabriels, D. (eds): Assessment of erosion, p. 127-132. Wiley, Chichester. Arnoldus, H.M.J., 1978. An approximation of the rainfall factor in the Universal Soil Loss Equation. In: De Boodt, M. & Gabriels, D. (eds): Assessment of erosion, p. 127-132. Wiley, Chichester. Batjes, N.H., 2000. Degradation and vulnerability database for Central and Eastern Europe — Preliminary Results of the SOVEUR Project. Proceedings of concluding workshop (Busteni, 26-31 October 1999), FAO, Rome, RISSA, Bucharest, and ISRIC, Wageningen, pp. viii + 99. Boardman, J. (1998). An average soil erosion rate for Europe: myth or reality? Journal of Soil and Water Conservation 53(1), 46-50. Boardman, J. and Favis-Mortlock, D.T. (eds) (1998). Modelling Soil Erosion by Water, Springer-Verlag NATO-ASI Series I-55 Berlin. 531 pp. Brazier, R.E., Rowan, J.S., Anthony, S.G., and P.F. Quinn, 2001. MIRSED: towards an MIR approach to modelling hillslope soil erosion at the national scale. Catena 42: (1) 59- 79 Briggs, D.J., and A. Giordano, 1995. CORINE Soil Erosion report, European Commission, 124 pp. Burt, T.P. (1994). Long-term study of the natural environment — perceptive science or mindless monitoring? Progress in Physical Geography 18(4), 475-496. CORINE (1992). CORINE Soil Erosion Risk and Important Land Resources in the Southern Regions of the European Community, Publication EUR13233 EN, Luxembourg. 97 pp. + maps. COST623 (Soil Erosion and Global change), 2001. Mid-term report on research cooperation in Europe. Report to European Commission. De Jong, S.M., 1994. Applications of reflective remote sensing for land degradation studies in a mediterranean environment. Nederlandse geografische studies 177. De Jong, S.M., Brouwer, L.C. & Riezebos, H. Th., 1998. Erosion hazard assessment in the Peyne catchment, France. Working paper DeMon-2 Project. Dept. Physical Geography, Utrecht University. De Ploey, J., 1989. A Soil Erosion Map for Western Europe. Catena Verlag. De Roo, A.P.J., 1993. Modelling Surface Runoff and Soil Erosion in Catchments Using Geographical Information Systems; Validity and Applicability of the ‘ANSWERS’ Model in Two Catchments in the Loess Area of South Limburg (The Netherlands) and one in Devon (UK). Nederlandse Geografische Studies 157. EEA, 1995. Europe's Environment: The Dobris Assessment, European Environment Agency, Copenhagen, Denmark. 676 pp. EEA, 1998. Europe’s Environment: The Second Assessment. Environmental assessment 75 “Erosion des Sols” en France. INRA, Orléans. Morgan, R.P.C, Morgan, D.D.V. & Finney, H.J. (1984). A predictive model for the assessment of soil erosion risk. Journal of agricultural engineering research 30, p. 245- 253. Morgan, R.P.C. (1992). Soil Erosion in the Northern Countries of the European Community. EIW Workshop: Elaboration of a Framework of a Code of Good Agricultural Practices, Brussels, 21-22 May 1992. Morgan, R.P.C. (1995): Soil Erosion and Conservation. Second Edition. Longman, Essex. Nearing, M.A., G.R. Foster, L.J. Lane and S.C. Finkner, 1989. A process-based soil- erosion model for USDA-water erosion prediction project technology. ASAE Trans, 32, 1587-1593. OECD (1993): OECD Core Set of Indicators for Environmental Performance Reviews. Environment Monographs No. 83, Paris OECD (1999). Environmental Indicators for Agriculture: Concepts and Frameworks, Vol. 1. Organisation for Economic-Co operation and Development, Paris Oldeman, L.R., Hakkeling, R.T.A. and Sombroek, W.G. (1991). GLASOD World Map of the Status of Human-induced Soil Degradation (second revised edition), ISRIC, Wageningen; UNEP, Nairobi Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., Crist, S., Shpritz, L., Fitton, L., Saffouri, R. and Blair, R. (1995). Environmental and economic costs of soil erosion and conservation benefits. Science 267, 1117-1123. Poesen, J., Boardman, J., Wilcox, B.P. and Valentin, C. (1996b). Water erosion monitoring and experimentation for global change studies. Journal of Soil and Water Conservation 51(5), 386-390. Poesen, J., Vandaele, K. and Van Wesemael, B. (1996a). Contribution of gully erosion to sediment production on cultivated lands and rangelands. In, Walling, D.E. and Webb, B.W. (eds), Erosion and Sediment Yield: Global and Regional Perspectives, IAHS Publication no. 236, Wallingford, UK. pp. 251-266. Poesen, J., Vandaele, K., Van Wesemael B., 1998. Gully erosion: importance and model implications. In: Boardman, J., Favis-Mortlock, D., Modelling soil erosion by water, Springer-Verlag Berlin Heidelberg. Renard, K.G., Foster, G.R., Weessies, G.A., McCool, D.K., Yoder, D.C. (eds) (1997). Predicting Soil Erosion by Water: A guide to to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). U.S. Department of Agriculture, Agriculture Handbook 703. RIVM, 1992. The environment in Europe: a global perspective. Report 481505001, RIVM, Bilthoven, The Netherlands. Van Der Knijff, J.M., Jones, R.J.A. and Montanarella, L. (1999). Soil Erosion Risk Assessment in Italy. EUR 19044 EN, 52pp. Van Der Knijff, J.M., Jones, R.J.A. and Montanarella, L. (2000). Soil Erosion Risk Assessment in Europe, EUR 19044 EN, 34pp Van Lynden, G.W.J. (1995). European soil resources. Nature and Environment No. 71. Council of Europe, Strasbourg. 76 . Van Lynden, G.W.J., Oldeman, L.R., 1997. The assessment of the status of human- induced soil degradation in south and southeast asia (ASSOD). ISRIC, Wageningen, The Netherlands. Van Muysen, W., Govers G., Bergkamp, G., Poesen J., Roxo, M., 1999. Measurement and modelling of the effects of initial soil conditions and slope gradient on soil translocation by tillage. Soil and Tillage Research 51: 303 Wischmeier (1975): in B.A. Stewart et al.: Control of Water Pollution from Cropland. Vol. I, A Manual for Guideline Development. Agricultural Research Service, Hyattsville, MD. Wischmeier (1976): in B.A. Stewart et al.: Control of Water Pollution from Cropland. Vol. II, An Overview. Agricultural Research Service, Washington, D.C. Wischmeier, W.H. & Smith, D.D. (1978). Predicting rainfall erosion losses –a guide for conservation planning. U.S. Department of Agriculture, Agriculture Handbook 537 Woolhiser, D.A., R.E. Smith and D.C. Goodrich.1990. KINEROS, A kinematic runoff and erosion model: documentation and user manual. USDA-Agricultural Research Service, ARS-77,pp 130. Yassoglou, N., Montanarella, L., Govers, G., Van Lynden, G., Jones, R.J.A., Zdruli, P., Kirkby, M., Giordano, A., Le Bissonnais, Y., Daroussin, J. & King, D. (1998): Soil Erosion in Europe. European Soil Bureau. 77 ANNEXES Annex I List of Participants Participants Address Anna Rita Gentile Anna.Rita.Gentile@EEA.eu.int European Environment Agency Kongens Nytorv 6, 1050 Copenhagen K, Denmark Tel: +45 33 36 71 00; Fax: +45 33 36 71 99 Anne Gobin Anne.gobin@geo.kuleuven.ac.be Laboratory for Experimental Geomorphology Redingenstraat 16, 3000 Leuven, BELGIUM Tel: (32-16)326433; Fax: (32-16)326400 Gerard Govers Gerard.govers@geo.kuleuven.ac.be Laboratory for Experimental Geomorphology Redingenstraat 16, 3000 Leuven, BELGIUM Tel: (32-16)326433; Fax: (32-16)326400 Paul Campling Paul.campling@agr.kuleuven.ac.be Ground for GIS Vital Decosterstraat 102, 3000 Leuven, BELGIUM Tel: (32-16)329732; Fax: (32-16)329700 Simon Turner Simon.Turner@adas.co.uk ADAS, Wergs Road, Woodthorne, Wolverhampton, England, UK, WV6 8TQ Hester Lyons Hester.Lyons@adas.co.uk ADAS, Wergs Road, Woodthorne, Wolverhampton, England, UK, WV6 8TQ Luis Carazo Luis.Carazo_jimenez@cec.eu.int Commission of the European Communities Unit B1, Water, the Marine & Soil BU-9 Office N0 3/137, Rue de La Loi 200, 1049- Brussels, BELGIUM Tel: 00 32 2 2960066 Pierpaolo Napolitano napolita@istat.it ISTAT DISS VIA A.RAVA',150 00142 ROMA, ITALY phone: 00390659524347 fax: 0039065943257 philipp.schmidt-thome@gsf.fi Philipp Schmidt-Thomé P.O. Box 96 FIN-02151 ESPOO Finland Mobile phone +358-40-542 4192 Office +358-20-5502163 Fax +358-20-55012 Robert Jones Robert.jones@jrc.it Commission of the European Communities - DG Joint Research Centre Environment Institute, European Soil Bureau TP 262, Via E. Fermi, Ispra (VA) 21020, ITALIA Tel: (39-332) 786 330; Fax: (39-332) 789 936 Godert van Lynden Vanlynden@isric.nl International Soil Reference and Information Centre PB 353, Duivendaal 7-9, 6700AJ Wageningen, THE NETHERLANDS Tel: (31-317)471715; Fax: (31-317)471700 80 Questions to guide the discussions In the evaluation and discussions during the workshop, the following questions were used to guide the review. 1. What is soil erosion? 2. What information does a policy maker need to assess soil erosion and its current impacts, and to formulate remedial measures in Europe? 3. Is the conceptual framework (DPSIR; MF-MI approach) adequate to describe soil erosion in Europe (its state, impacts on the soil resource and on other media, the causes and measures)? 4. Is the list of proposed indicators for soil erosion adequate? How many and which type of indicators should be advocated? (ideas for change) For each indicator in the list: • Is the indicator adequate? • Are the data used adequate? • Are the conclusions and is the assessment correct? • What else should be taken into account? 5. What are the driving forces of soil erosion? (with specific attention to agriculture) 6. Are the drivers of soil erosion sufficiently known and how do they link to the phenomenon? 7. Are there other quantifiable indicators of impact apart from the proposed indicator ‘removal of sediment deposits’? 8. Is the assessment of soil erosion in Europe correct? Are the methods used scientifically sound? 9. What are the recommendations for further work? A specific point of discussion was the indicator of state for soil erosion. Soil erosion is recognised to be highly variable in both space and time. The following questions were used as guidelines to discuss the assessment of soil erosion. 1. Which erosion types should be considered? (wind, water, gully, mass movements, active versus non-active erosion (old gullies), 2. How can or should tillage erosion be incorporated in the framework? 3. Can thresholds be derived for policy purposes? How should these be set? 4. What should be the preferred scale for assessing soil erosion taking into account its use for policy-makers? (nested strategies at multiple scales, …) 5. How should the extent of the erosion problem be mapped in relation to the severity or frequency? 6. What are good indicators of state and impact? 80 81 Annex III Background papers presented at the workshop State-play of EEA work on soil erosion indicators Anna Rita Gentile Project manager for soil and contaminated sites European Environment Agency This first presentation provided some background information and focused on the objectives of the EEA work programme on soil and a description of the European soil monitoring and assessment framework. The state-of-play of EEA work on indicators for soil erosion and EEA expectations from the workshop were also discussed. In particular, EEA organised the workshop with the aim to take stock of the work done, get expert advice on how to proceed with the work on soil erosion, connect with other relevant initiatives on soil erosion at the European and national level and help to define the work plan of the new European Topic Centre on Terrestrial Environment. A selection of overheads is included below. 3 MDIAR chain: getting from BAI to BNI Best Available Env Information Current env/econ information system in Europe Best Needed Env and Econ Information 4 Monitoring Data Information Assessment Reporting R: Reporting A: Assessment I: Information D: Data M: Monitoring MDIAR stands for: M: D: I: A: R: So far, EEA has collected information based on the ‘best available’ data. However, this approach, although allowing for the provision of timely information, has showed some limitations. For example, it may not help rationalise ongoing data collection and monitoring activities at the national and European levels, possibly covering subjects that are not needed, while resources should be better employed to fill data gaps in other priority areas (BTG, 1998). In order to help streamline monitoring, assessment and reporting activities, a broader approach is required. In the long-term, the objective is to focus on the ‘best needed’ data. This shift should be obtained by building stronger links to EU policy needs, by focussing on the assessment of the environmental impacts of soil degradation and by undertaking a more detailed analysis in hot spot areas. 81 82 6 Objective of EEA work on soil (1) Contribute to EEA reporting through the provision of policy- relevant information on soil- related issues… 7 Objective of EEA work on soil (2) …by analysing not only state and trends in the soil environment but also assessing: • the causes of changes in state and related pressures acting upon the soil; • the impacts on soil functions and other environmental media; • the effectiveness of society’s responses (policy measures). 9 How is the work organised? Conceptual and operational framework for the assessment and the monitoring of soil in Europe based on the general EEA "monitoring to reporting framework" 10 European framework for the assessment and the monitoring of soil (1) The framework is both: • Conceptual (DPSIR, MF/MI) • Operative 11 European framework for the assessment and the monitoring of soil (2) based on: • EEA ’from monitoring to reporting framework’ (MDIAR) • political framework (definition of priority issues; based on needs) • Identification of relevant indicators through use of analytical tools (e.g. DPSIR,MF/MI) 12 Implementation • European Topic Centre on Soil (1996- 1999) • Working groups on soil indicators (2000) • European Topic Centre on Terrestrial Environment (2001-2003) 82 85 3 Soil erosion - State • Water erosion risk in agricultural areas, 2050 (map) • Change in water erosion risk in agricultural areas due to climate change, 1990-2050 (map and tables) • Area affected by water erosion in selected countries in the period 1990-1995 • Soil loss due to water erosion in selected countries in the period 1990-1995 • Probable problem areas of soil erosion in Europe (map) 4 Changes in water erosion risk in agricultural areas due to climate change in the period 1990-2050 P rojecte d change s in risk to so il e rosion in ag ricu ltura l are as in the p eriod 1990-2050 -60% -40% -20% 0% 20% 40% 60% Austri a Belg iu m Denm ark F ra nce Germ any G ree ce Ire land Ita ly Luxembourg Neth er la nds Portu ga l Spa in UK All c ountr ie s n.a. very high high moderate low Projected changes in water erosion risk in agricultural areas in the period 1990-2050 as % of total land area Source: European Commission, 1999;EEA data elaboration 85 86 5 Soil erosion - Impacts • Landslide and flooding events in Italy in the last 100 years • Hydro-geological risk in Italian municipalities • Estimated organic carbon in topsoils of Southern Europe 6 Slope stability 223 136 450 569 222 1232 1519 1282 2068 2653 10354 112 163 653 547 313 1203 1381 1107 618 485 6582 0 2000 4000 6000 8000 10000 12000 1900-1909 1910-1919 1920-1929 1930-1939 1940-1949 1950-1959 1960-1969 1970-1979 1980-1989 1990-1999 Grand Total Pe rio d Number of events Total floods Total landslides Landslide and flooding events in Italy in the last 100 years Sources: EEA data elaboration from AVI database, 1999 Sources: Italian Ministry of Environment, 1999 Hydro-geological risk in Italian municipalities 86 87 7 Soil erosion - Data sources • Landuse statistics (Eurostat- OECD) • Published national data • European datasets (CORINE soil erosion risk) • Global datasets (GLASOD) • Outlooks/Models (IMAGE 2) • Ad-hoc data collections 8 Soil erosion - Data gaps • Data are patchy • Data are not comparable • Few time series available • Access to existing data is difficult • Few data on each of the elements of the DPSIR chain • Data mostly available at field level 9 Soil erosion - Data needs • European coverage • Comparable data • Time series (trends) • Information on pressures (e.g. landuse changes, agricultural practices), state (e.g. Soil loss due to erosion,including wind erosion), impacts (e.g. floods, sedimentation, loss of fertility), responses (e.g. Public expenditures, soil conservation measures) • Geo-referenced information at the regional level (e.g. ’hot-spots’) 90 Data quality issues The spatial and temporal occurrence of erosion Erosion is patchy in both space and time (cf. Figure 1). Loss of soil can be highly variable even in areas of severe erosion. For water erosion for example, the vagaries of topography concentrate erosive flows so that severe erosion in one field can be found side-by-side with almost untouched areas. Similarly, several years can pass between major erosion events (water or wind) even in erosion-prone regions. Long time series of measurements of erosion are therefore required to adequately estimate erosion rates. Temporal distributions of erosion are highly skewed, so that calculation of long-term average values for erosion is statistically dubious (use of the median is preferable, but uncommon). Precise delineation of erosion hotspots is therefore futile. Additionally, even within an area which is designated as a hotspot it may well be that only a minority of fields will show obvious erosion at any time. Also since almost all erosion monitoring studies operate for only a relatively short period, any assessment of erosion rates for these hotspot areas is fraught with uncertainty. Spatial considerations regarding data collection To a large extent6, erosion is independent of national boundaries. However field measurements of soil erosion may be obtained during a study which is funded or sponsored by a particular country, or by a scientist who works within well-defined regional boundaries. The emphasis placed upon erosion studies also varies markedly from country to country. As a result, the availability of data on European erosion varies strongly from region to region. There is thus some risk both of spurious ‘hotspots’ being generated simply by the presence of abundant data for an area, and also of the inverse problem: lack of data resulting in under-emphasis of an area’s erosion problems. Any Europe-wide study of erosion must therefore exercise discrimination in the face of possibly artefactual positive or negative hotspots. Techniques of data collection Techniques of data collection are an issue with respect to the erosion rates quoted here. Even for the same location, erosion rates obtained by different methods7 are likely to vary. This study has — unavoidably — had to draw upon data for erosion rates which was obtained by a range of methods. While in some cases it is possible to reconcile such methodological variations, in general the result is to increase the uncertainty associated with rates assigned to mapped hotspots. Soil loss rates calculated from plot-sized areas (the most common among the studies reported here) can be up to one or two orders of magnitude higher than sediment yields calculated from catchments. However, results from small areas such as plots do not include the contribution which talweg (valley- bottom) gullying can make to total erosion: this may be over 40 % in N Europe and over 6 Except where trans-border land use is strongly influenced by differing national policies. 7 For example, by field survey of rill depths, collection of sediment lost from a plot, or aerial photography. 91 80 % in S Europe (Poesen et al. 1996a). Due to erosion’s temporal variability, soil loss rates from single events are generally not reported here, except where measured data is scarce e.g. Eastern Europe. Other issues regarding data Most of the source publications for this map are in English. This is a definite limitation, although ameliorated to some extent by the use of English publications which summarise earlier non-English work. The design and use of this map As described previously, this map shows erosion on a three-level spatial hierarchy. For proper use of this map it is vital to remember the following caveats: hotspot boundaries are rather arbitrary even within a hotspot area, erosion occurs patchily there is a considerable variability and uncertainty associated with all cited rates of erosion. expert judgement has played a major role in the methodology used here. This is unavoidable, given the complex nature of erosion’s occurrence and the limitations of currently available data. Thus it is important to note that, just because erosion is not indicated at a particular location on this map, this does not imply that no erosion occurs there. For example, erosion occurs regularly in Denmark (Hasholt 1988; 1998) but does not appear to be a major problem there. Boundaries for water and wind erosion hotspot areas in western and southern Europe are in most cases modified from De Ploey (1989), while others have been deduced from the publications cited. Those for eastern Europe are also derived from individual publications, interpolated as necessary. Interpretation of the map General There are three broad zones of erosion in pan-Europe: a southern zone, a northern loess zone, and an eastern zone8. In the southern zone, severe water erosion results from intense seasonal rainfall. This is often associated with overgrazing or a move away from traditional crops. Erosion here may be of considerable age. The principal impact is on- site: soil productivity decreases as a result of thinning. The northern zone has moderate rates of water erosion. This mostly results from less intense rainfalls falling on saturated, easily erodible soils. There is also local wind erosion of light soils. Impacts here are mainly off-site, as agricultural chemicals from the north’s more intensive farming systems are moved into water bodies along with eroded sediment. Partially overlapping these two zones is the eastern zone, where former large state-controlled farms produced considerable erosion problems. Eroded sediments here may also be contaminated from 8 Iceland was not included in the original study. 92 former industrial operations. Other, relatively minor, areas of erosion occur outside these zones. Within all three zones, there are ‘hotspot areas’ where erosion is more serious. The coverage of reliable measurements of erosion is very patchy, and to an extent reflects the activities of particular workers rather than the severity of the problem. Rates of erosion As noted on the Methodology, regional rates inferred from this map must be very tentative. Nonetheless there is some indication that average rates of soil loss are higher in southern and eastern Europe than in the north-west. This is conventional wisdom; however rates for the south appear to be generally much lower than (for example) the 27 t ha-1yr-1 for the whole of Spain which was quote in Europe’s Environment: The Dobríš Assessment (EEA 1995, p.155). Impacts of erosion The impacts of erosion are not a simple function of erosion rate. These impacts can be categorised as ‘on-site’ and ‘off-site’. Off-site problems of water pollution from agricultural chemicals can result even from very low rates of soil loss (Harrod, 1994). Erosion’s impacts across Europe can be very generally summarised as described in Table 2. Table 2: Erosion’s impacts across Europe Zone Short-term (i.e. decades) Long-term (i.e. centuries) North-west Europe Off-site: water pollution from agricultural chemicals On-site: loss of soil productivity Southern Europe On-site: mainly loss of soil productivity On-site: loss of soil productivity Eastern Europe On-site: loss of soil productivity Off-site: water pollution from former industrial waste, as well as agricultural chemicals On-site: loss of soil productivity Policy implications EU recognition of the impacts of soil erosion has to date largely been confined to the south of Europe e.g. the CORINE9 and MEDALUS10 studies (Stanners and Bourdeau, 1995). This is principally due to a focus only on on-site effects. However, if off-site impacts are also considered, then there is a need for greater EU acknowledgement of erosion problems elsewhere in Europe. At a national level, there has been some progress in this direction. For example, a recent publication11 from the UK Ministry of Agriculture, Fisheries and Food states that ‘… erosion of agricultural land is more widespread in this 9 See CORINE (1992). 10 For example, Mairota et al. (1997). 11 Ministry of Agriculture, Fisheries and Food (1997). 95 global scale in a relatively rapid (three years) and cheap (around US$ 300.000) manner. It raised awareness on soil degradation problems and created wide interest among scientists and general public. It provided an overview for national and regional planning and enabled identification of “hot spots” for further study. From the received feedback it was clear that GLASOD responded to a strong apparent need for a global overview. Multiple requests were received for national breakdowns or new assessments at country level. GLASOD also showed the need for an assessment of measures to control degradation, i.e. showing some “good news”. At the same time, the need for an additional more objective/qualitative approach (especially for more detailed scales) as well as the need for data validation and updating also became obvious. Who are the users? GLASOD has a wide range of potential and proven users, such as: • International policy makers and planners (e.g. UNEP, FAO, WRI) • National policy makers and planners • International conventions and programmes (CCD, Kyoto protocol, UN-CPB, IGBP) • Researchers at national and international level (NARI’s, CGIAR, universities) • Education professionals (teachers, professors, etc.) and students • Environmental organisations (general public awareness) Limitations and problems As a “quick and dirty” methodology, GLASOD (and its derived successors) also has several limitations that need to be taken into consideration. Some of these limitations were overcome in subsequent assessments. • The small scale makes GLASOD less appropriate for national breakdowns • The expert judgement approach can lead to subjectivity • Cartographic restrictions at the time of publication limited the number of attributes on the map • The representation of the map items causes a visual exaggeration: each polygon which is not 100 % stable shows a degradation colour, even if only 1 to 5 % of the polygon is actually affected • Extent was expressed in classes rather than percentages • The map has a complex legend: extent and degree (severity) are aggregated for four major degradation types (water and wind erosion, physical and chemical deterioration) • Only the “dominant” main type of degradation is shown in colour • Degradation sub-types are only shown by codes printed in each polygon • GLASOD presented only “bad news” (doom scenario) 96 Follow-up of GLASOD / derived initiatives ASSOD (1997): Assessment of the Status of Human-induced Soil Degradation in South and Southeast Asia In 1993 an Expert consultation of the FAO-supported Asian Network on Problem Soils recommended the preparation of a South and Southeast Asian soil degradation status assessment (ASSOD) at a scale of 1:5 million. This study was commissioned by UNEP to ISRIC and carried out in close co-operation with FAO and national institutions in 16 countries. The project used a modified GLASOD methodology, with more emphasis on the impact of degradation on productivity and on the rate of degradation and used a 1:5M physiographic base map following criteria outlined in the Global and National Soil and Terrain Digital databases methodology (SOTER, van Engelen and Wen, 1995). The ASSOD project finished in 1997. All information was stored in a digital database, which is linked to physiographic units through a GIS, This enables a more flexible production of outputs: thematic or regional maps, no restrictions to number of attributes per polygon, less complex legend. Figure 2: The ASSOD map for south and southeast Asia. SOVEUR (2000): “Mapping of Soil and Terrain Vulnerability in Central and Eastern Europe" In 1997 the FAO and the ISRIC initiated the project on "Mapping of Soil and Terrain Vulnerability in Central and Eastern Europe" (SOVEUR). There were three main activities in the project: 97 • Development of a soils and terrain digital database, at scale 1:2.5 million, for the countries under consideration, using the uniform methodology of SOTER. • Assessment of the status of soil degradation, with special focus on diffuse pollution, according to a modified GLASOD methodology • Providing the soil geographic and attribute data for an assessment of the vulnerability of soils to selected categories of pollutants. Implementation The SOVEUR project has been implemented in close collaboration with specialist institutes from 13 countries: Estonia, Latvia, Lithuania, Poland, Czech Republic, Slovakia, Hungary, Rumania, Bulgaria, Belarus, Ukraine, Moldavia, and (the European part of) The Russian Federation. Initial results were presented and discussed during an international workshop in October 1999. Thereafter, the assessment has been finalised. In December 2000 the databases and technical documentation have been released on a CD- ROM in FAO's Land and Water Media Series (no.10). This CD-ROM contains information in the form of databases, maps and reports on soil, on the soil degradation status and gives a soil vulnerability assessment for eleven metals in thirteen countries in Central and Eastern Europe. Beneficiaries Target beneficiaries are Ministries and planning bodies in the collaborating countries who can use the definitive databases and derived maps for policy formulation at the national level, for instance by identifying areas considered most at risk. The project further contributes to strengthening of the capabilities of national "environmental" organisations in Central and Eastern Europe, and it can play a significant role in enhancing scientific co-operation within Europe on issues of soil degradation and pollution. Further, it is an integral part of a global programme on the development of a world soils and terrain information system, a world assessment of the status and risk of soil degradation, and studies of the potential productivity assessment of the land (cf. UNCED 1992). General degradation guidelines Based on the experiences with GLASOD (Fig. 1), ASSOD (Fig. 2) and SOVEUR (Fig. 3), guidelines for the qualitative assessment of soil degradation have been developed that are generally applicable, scale-independent and offer links to other standardised methodologies (SOTER, WOCAT). WOCAT: since 1992, ongoing In response to the “bad news” of GLASOD a new project was initiated to investigate what measures are being taken to combat degradation. A consortium of various national and international organisations, institutions and individuals, guided by a management board is undertaking an inventory of SWC world-wide. Through the collection, analysis and dissemination of existing experiences, it is expected that mistakes and duplication of efforts can be minimised. WOCAT is using a set of comprehensive questionnaires on technologies, approaches, mapping respectively that serves as a framework for the evaluation of soil and water conservation and a methodology for data collection at the same time. This information is stored in a MS-ACCESS database with a user-friendly menu for storage, analysis and output of data. Regional and national training workshops