This paper will refer specifically to the pilot project for which the mechanisms' developed to manage and reconcile agriculture on land contaminated due to heavy metals, have been implemented. Extensive data sets pertaining to soil analysis, geology, internal transport networks, industrial locations and agricultural practices in a variety of forms, were identified or collected and integrated within a Geographical Information System (GIS). The spatial distribution of heavy metal contamination throughout the Gmina were determined utilising GIS functionality; based on the land classification scheme devised by ADAS.
1. INTRODUCTION The Katowice Voivodeship has been regarded as one of the most ecologically damaged areas in Europe. The average human life expectancy is one year shorter than in other parts of Poland with a particularly high death rate in the 30 to 59 age group. Infant mortality is also higher than the national average and birth defects occur in 60% more cases than in Poland as a whole. Industrial activity in the region is considerable with approximately 17% of the nation's industrial production undertaken in the area. Contamination of land with heavy metals has been identified in some areas of the Voivodeship due primarily to past and present industrial activity (ADAS HASKONING, 1994). Road traffic has also made a contribution. As a result significant areas of agricultural land in the central part of the province are directly affected by heavy metal contamination. In terms of land use, arable crops represent the main farming enterprise. There is also a significant area of grassland. The agricultural structure of the Voivodeship is dominated by small farms in the private sector with cereals, potatoes and fodder crops being the staple agricultural production. These products are likely to represent a high proportion of the potential dietary intake of contaminants by high risk consumer groups (i.e. those growing a high proportion of their own food). In August 1994, ADAS together with HASKONING were commissioned by the European Union (EU) under the PHARE programme to advise on how current agricultural practices in the Katowice Voivodeship could be altered to reduce the risk of human exposure to high levels of heavy metals, through the consumption of locally grown foods. The ultimate aim was to develop management practices to control and ultimately to eliminate food production in those areas contaminated by heavy metals. A general study of the entire Katowice Voivodeship was undertaken to provide mechanisms for controlling food production on land classified as contaminated due to heavy metals. Additionally, a pilot project was carried out to implement these mechanisms in the Voivodeship's Tarnowskie Gory Gmina. The first stage of the programme involved the development of a land classification scheme for heavy metal contamination for the Voivodeship. The scheme was based upon International standards and those that currently exist in EU Member States. Soil concentrations of heavy metals, deposition of metals from industry and road traffic are considered with specific advice given on suitable agricultural uses for each land contamination class. A framework to enable the region's agricultural authorities to develop controls and advice for local farmers and allotment holders was subsequently defined. The implementation of the land classification scheme and the mechanisms for advising food producers in the Gmina provided a way to practically test and revise the food production control strategies proposed. In addition, it demonstrated the rapid improvements that could be made to agricultural practices for a typical Gmina. This paper will focus on the implementation of the devised land classification scheme for the Gmina and the identification of food producing activities at risk due to this contamination under the pilot project. Details on the land classification scheme and the management mechanisms to control food production on land identified as contaminated will not be given. 2. METHODOLOGY 2.1 USE OF GEOGRAPHICAL INFORMATION SYSTEMS (GIS) The implementation of the land classification scheme required the analysis and comparison of several datasets. Geographical Information Systems (GIS) is specifically designed to analyse spatial data and was therefore central to the pilot project. The GIS also provides the flexibility to quickly re-calculate the classification scheme as new data is collected. By utilising the functionality of the GIS the areas of land falling in each class defined by the land classification scheme were identified. The areas experiencing contamination due to heavy metals were thus determined and overlayed with typical agricultural production patterns. Food production activities in the identified contaminated areas were therefore highlighted allowing appropriate advice on the agricultural management and cropping patterns to be efficiently targeted. 2.2 LAND CLASSIFICATION SCHEME FOR CONTAMINATED LAND A four category land classification scheme was proposed for agricultural land throughout the Voivodeship. Although primarily based on the concentrations of lead and cadmium in the soil this classification scheme also considers atmospheric inputs and industrial sources (ADAS HASKONING, 1995). Essentially the classification scheme denotes that land in Class 1 is clean and that no restriction of food production is needed. In contrast, land in Class 4 is considered as contaminated due to heavy metals and recommends that restrictions on the production of foods and livestock grazing should be implemented. Details are given in Table 1. The implementation of this land contamination classification for the pilot project required information on soil metal concentrations, road traffic volumes and the atmospheric deposition of metals. It was considered that the only industrial source of concern in the Gmina was the HUTA zinc smelter in Miasteczko Slaskie. The effects of the smelter were incorporated in the classification implementation via the atmospheric deposition of its heavy metals. Table 1: Land Classification Class 1: The cleanest land where all food can be grown to a 'premium standard' Class 2: All crops can be grown. The current average dietary intake in the Voivodeship is maintained and may be expected to show a small positive improvement as Class III is introduced and the most contaminated produce is removed from supply. (The diet of the children of farmers consuming a very high proportion of home grown food may require adjustment at the top end of the range). Class 3: In consultation with nominated experts only a restricted range of food crops should be grown. Random official monitoring of produce should be introduced. Soil contamination should be minimised. The land should be considered for alternative use, such as forestry, industrial or other non- food crops. Class 4: In consultation with nominated experts food production should be discouraged on these most contaminated soils. It might be possible to grow a very limited range of food crops but only under strict monitoring. Such crops would be better fed to livestock. Livestock should not be grazed continuously on these soils as the risk of metal ingestion from the soil causing health problems is high. 2.2.1 SOIL ANALYSIS The foundation of this land classification relies upon the spatial distribution of soil concentrations of cadmium and lead. Soil analysis data from various Polish institutes were made available to the pilot project and their spatial location known (provided in digital format or on original maps). Additional data from Imperial College, London were also made available. A total of 631 soil sample results were used. Once implemented within the GIS appropriate interpolation functions were employed to determine the spatial patterns of lead and cadmium values across the Gmina. These interpolated metal concentrations were used to determine a preliminary classification according to the thresholds outlined in Table 2. 2.2.2 ROAD TRAFFIC VOLUMES Road vehicles are an internationally recognised source of both lead and poly aromatic hydrocarbons (PAHs). It was estimated that road traffic contributes about 48% of all lead emitted to the environment in the Voivodeship. Work by IETU investigating the lead and cadmium content in soils at increasing distances from roads in 30 locations (Gzyl, 1992) determined that lead concentrations within two and five metres to the road were notably higher than background concentrations of lead. The effects of road traffic were therefore considered to be an important factor influencing land contamination and was consequently incorporated as a factor in the land classification scheme. All land within 30 metres of a major road (with over 20,000 vehicles per day) were considered to be more polluted and to have an influence on the land class (see Table 2). Although the location of all the major roads in the Gmina were known and implemented in the GIS (digitized from scanned maps) it was concluded that less than 20,000 vehicles travelled them per day. Hence, the implementation of the classification scheme for Tarnowskie Gory Gmina did not incorporate the effects of road traffic pollution. Table 2: Proposed Land Classification based on Lead and Cadmium Values 2.2.3 ATMOSPHERIC DEPOSITION The atmospheric deposition of lead and cadmium are important factors in the land classification scheme. Therefore, it was necessary to model the atmospheric deposition of these pollutants in the vicinity of known industrial emitters. Although in recent years the concentrations of dust, lead, cadmium and other elements being emitted throughout the Voivodeship have been greatly reduced, the incorporation of atmospheric dispersion model results were a vital component of the classification scheme. The UK's Meteorological Office was commissioned to undertake the necessary atmospheric modelling of lead and cadmium deposition from the HUTA zinc smelter in the Gmina. These results were provided as a series of graphs which were incorporated in the GIS. The land classification scheme was employed on these model results with key levels of dry atmospheric deposition of 100 and 1,000 mg/m2/year increasing the contamination classification of land. Each component of the land classification scheme was implemented separately within the GIS utilising the appropriate spatial datasets. Each of these results (for lead and cadmium values from soil analysis and atmospheric pollutants) were subsequently combined to produce an overall land contamination distribution for the entire spatial extent of the Tarnowskie Gory Gmina. Thus the changing patterns of clean to heavily contaminated land were identified. 2.3 AGRICULTURAL PRODUCTION The ingestion of toxic substances in the form of heavy metal concentrations in food in the areas of Poland studied under this programme poses a potential risk to the long term health of some people, notably those consuming a high proportion of home produced food. The need to identify the main areas of agricultural production, crops grown and the proportions of home grown foods contributing to the overall family diet was important. The agricultural structure consists of many small farms frequently producing food for their own consumption. Similarly foods grown by allotment owners and those with household gardens form an important component of the total family diet. Obviously, food grown in areas identified as at risk from contamination is a cause for concern. 2.3.1 FARM PRODUCTION A confidential survey of all individual farms above 1 hectare in size was undertaken for the Gmina to provide information on typical agricultural production patterns at a farm scale. The results of this survey provided the essential base of information needed to aid the formulation of strategic planning and individual advice. The address location of each farm surveyed was known and overlaid in the GIS with the determined land contamination classification. Those farms producing foods susceptible to contamination on land identified as within an area of high contamination were therefore identified and subsequently targeted for specific advice. The risk of such potentially contaminated foods entering the local food markets was therefore identified and reduced. 2.3.2 ALLOTMENT AND GARDEN PRODUCERS Some vegetables grown on allotments for family consumption may represent an important source of the total intake of lead and cadmium for some families. A similar survey to that undertaken for local farms was implemented to ascertain the range of vegetables grown in each of the 24 blocks (with over 2,000 growers) of allotments in the Gmina. The results of this survey aided estimations of the amount of potentially contaminated foods consumed in the daily diet via this type of production. Data on the location and area of each of these allotment blocks were implemented in the GIS and overlaid with the land classification results. Those allotment blocks which fell within areas of high contamination were again immediately identified and targeted for advice. Growers were informed of the dangers of food production on highly contaminated soil, and advised of techniques for production and food preparation to reduce the risk of eating contaminated foods. Private gardeners producing food for their own consumption are similarly at risk. In order that similar advice could be conveyed to these producers, built up residential areas falling within a contaminated area were identified and targeted with general advice via local publications. 3. RESULTS 3.1 LAND CLASSIFICATION The initial utilisation of the land classification scheme for lead and cadmium concentrations identified several hot spots. Lead concentrations were found to be high in the town of Tarnowskie Gory (where much historic ore mining has been undertaken with many mine spoil heaps) and also in close proximity to the HUTA zinc smelter in Miasteczko Slaskie. Small cadmium hot spots occurred in the region of the smelter and also in small areas to the south of Tarnowskie Gory town. The modelled atmospheric deposition concentrations for lead and cadmium clearly revealed that deposition occurred more extensively to the north east of the HUTA zinc smelter (the likely effect of prevailing winds). Lead and cadmium concentrations were modelled down to 50mg/m2/a-1. Figure 1 shows the area over which lead is deposited from the HUTA zinc smelter. The results from the soil concentrations and atmospheric depositions for lead and cadmium were combined and adjusted in accordance with the land classification scheme. The majority of land in the Gmina was classified as being within classes 1 or 2. A significant area of land around the town of Tarnowskie Gory fell into class 4 which is fringed by land classified as class 3. In the vicinity of the HUTA smelter the effects of atmospheric deposition are clear as a large area (similar in pattern to the modelled atmospheric dispersions) falls in class 3. A corridor of land extending from this area to the south east boundary of the Gmina is similarly classified as class 3. Small patches of land within each of these areas reach class 4, as can be seen in Figure 2. 3.2 AGRICULTURAL PRODUCTION Of the 542 farms surveyed the address point of only 54 farms were found to be located on land classified as class 4. 68 farm address points located on land in class 3. Over 77% of farm address points fall on land that is in either class 1 or 2. Only 7 allotment blocks are located on land classified as being in either class 3 or 4. This affects only 38% (33.04 hectares) of land that is under allotment cultivation. Approximately 700 allotment parcels of differing sizes are affected. The remaining 87.39 hectares of land (with over 1,600 allotment parcels) under allotment cultivation is located on land falling into either class 1 or 2. 4. DISCUSSION Due to the variable nature of the soil analysis results (extremely high values adjacent to low values) the heavy metal contamination of the soil may not be as extensive as has been currently classified. The inclusion of further soil analysis data (once collected) would refine the classification considerable and is likely to result in a more patchwork pattern of land reaching class 3 or 4. It is likely that contamination is more localised and in close proximity to sites of historic mining activity as is the region of the smelter. Although the employment of the land classification scheme for the pilot project did not utilise the road traffic component it is expected that in time road traffic is likely to rise above the critical volume. The inclusion of the traffic component of the classification can be quickly employed as the location of all the Gminas roads have been implemented in the GIS. Emissions from the HUTA zinc smelter have been significantly reduced since 1990 with the introduction of improved pollution measures (fitting of filters, etc.). Since 1985 there has been over 100 tonnes/year reduction in the lead emissions and over 0.8 tonnes/year reduction in cadmium. Dust emissions have been reduced by over 80% to only 72 tonnes/year. It is likely that these reductions will help reduce future health risks due to heavy metals in the environment. However, further reductions are required in some areas in order to meet standards set in existing Polish regulations. Legally binding Polish regulations giving critical concentrations for the deposition and concentration of atmospheric contaminants are currently below existing EU guidelines (where given). In time it is likely that the atmospheric deposition of heavy metals from the smelter will have a reduced effect on the land classification implementation as levels are reduced. Future model results can be easily fed into the GIS and the atmospheric consideration in the classification scheme re-employed. As a result of the farm survey the home address of each farmer was known. Unfortunately due to practical reasons the actual location of land farmed by each of these farmer was not. Although the address point of 21% of these farms over 1 hectare in size were found to be located on land classified as class 3 or 4 it can not be assumed that all or any of the land associated with each of these farmer addresses also falls on land of the same classification. Each of these farms located on land in class 3 or 4 were all in built up areas, Tarnowskie Gory town and near the smelter. It is therefore likely that the land associated with these addresses is located away from these built up areas and therefore falls on land of a 'cleaner' class (1 or 2). In order to obtain results on the exact total area of farmed land at risk of heavy metal contamination, farm boundary information should be included in the GIS. It was unfortunate that the duration and resources available within the pilot project did not allow the inclusion of this data. The utilisation of GIS technology played an underpinning role throughout the entire project, particularly in the implementation of the land classification scheme for Tarnowskie Gory Gmina. GIS enabled the various spatial datasets to be drawn together in a logical and structured fashion from a wide range of original data formats (including scanned basemaps, survey information, original historic maps, graphs and spreadsheets). The ability to interpolate soil sample locations across a spatial extent and overlay these with spatial distance functions and atmospheric model results enabled the employment of the soil contamination classification quickly and easily. Thus the spatial extent of Tarnowskie Gory Gmina was classified into one of the four classes defined by the land classification scheme. Agricultural areas and allotments at risk of producing contaminated foods were instantly identified by using GIS functions to overlay these land classification results with farm survey results and allotment locations. Vast areas of non-agricultural land, forest and industrial parks, could immediately be excluded therefore enabling more detailed analysis to be deployed on the more likely areas of interest. Figure 3 outlines the key role GIS played in uniting the wide range of aspects involved in the programme. Although the cost of measures to control heavy metal contamination are frequently high this may be partially or completely offset by related savings in health care, work efficiency and other benefits to the local economy. Any excessive contamination with toxic substances can cause short or long term illness, or lead to premature death of humans or other animals. The direct and indirect costs of control measures must be balanced against those of health protection. 5. CONCLUSION From the implementation of the land classification scheme in the Tarnowskie Gory Gmina of Katowice Voivodeship, Poland the following can be concluded: * The majority of land (79%) in Tarnowskie Gory Gmina falls into either class 1 or 2. The majority of land in the Gmina is therefore essentially free from heavy metal contamination. No restriction on the production of food crops is required in these areas. * Over 77% address locations of farms over 1 hectare in size in the Gmina are located on land free from heavy metal contamination. * Over 70% of the allotment blocks in the Gmina are located on land free from heavy metal contamination. * 21% of the land in Tarnowskie Gory Gmina falls into either class 3 or 4. These areas are therefore considered as contaminated due to heavy metals. The implementation of restrictive food production activities and livestock grazing is required. * 23% of farms over 1 hectare in size in the Gmina have home address points located on land which is contaminated due to heavy metals. It is not known whether or not land under agricultural production associated with these addresses is located on land similarly contaminated. Further information on the exact location of the areas farmed by these farmers is required so that advice on changes in agricultural practices can be more specifically targeted. * Less than 30% of the allotment block in the Gmina are located on land considered to be contaminated. Food production activities on these allotments must be restricted in accordance with the targeted agricultural advice. 6. ACKNOWLEDGEMENTS Financial support for this work from the European Union (EU) is gratefully acknowledged. Thanks are due to the whole ADAS HASKONING Team for all information relating to the soil classification scheme and also to various Polish Associates and Institutes. 7. REFERENCES ADAS HASKONING (1994): General Study and Pilot Project in Katowice Voivodeship on the Elimination of Food Production in Areas Polluted with Toxic Substances. Technical Proposal Submission to the European Union. ADAS HASKONING (1995): Collation and Evaluation of Environmental Data - Katowice Voivodeship. Report to the European Union. Gzyl, (1992): Heavy Metal Content in Soil and Plant Samples from Allotments and from the Vicinity of the Motor Roads in Katowice District. Institute of Environmental Protection, Katowice.