THE CONTRIBUTION OF NEIGHBOURING COUNTRIES TO PESTICIDE LEVELS IN DUTCH SURFACE WATERS

Transcription

1 Comm. Appl. Biol. Sci, Ghent University, 76/2, THE CONTRIBUTION OF NEIGHBOURING COUNTRIES TO PESTICIDE LEVELS IN DUTCH SURFACE WATERS M. van t Zelfde, W.L.M. Tamis, M.G. Vijver & G.R. de Snoo Department of Conservation Biology, Institute of Environmental Sciences, Leiden University (CML), P.O. Box 9518, NL-2300 RA Leiden, The Netherlands Corresponding author, zelfde@cml.leidenuniv.nl SUMMARY Compared with other European countries, Dutch consumption of pesticides is high, particularly in agriculture, with many of the compounds found in surface waters in high concentrations and various standards being exceeded. Surface water quality is routinely monitored and the data obtained are published in the Dutch Pesticides Atlas. One important mechanism for reducing pesticide levels in surface waters is authorisation policy, which proceeds on the assumption that the pollution concerned has taken place in the Netherlands. The country straddles the delta of several major European rivers, however, and as river basins do not respect national borders some of the water quality problems will derive from neighbouring countries. Against this background the general question addressed in this article is the following: To what extent do countries neighbouring on the Netherlands contribute to pesticide pollution of Dutch surface waters? To answer this question, data from the Pesticides Atlas for the were used. Border zones with Belgium and Germany were defined and the data for these zones compared with Dutch data. In the analyses, due allowance was also made for authorised and non-authorised compounds and for differences between flowing and stagnant waters. Monitoring efforts in the border zones and in the Netherlands were also characterised, showing that efforts in the former are similar to those in the rest of the country. In the border zone with Belgium the relative number of non-authorised pesticides exceeding the standards is clearly higher than in the rest of the Netherlands. These exceedances are observed mainly in flowing waters. In contrast, there is no difference in the relative number of standard-exceeding measurements between the border zones and the rest of the Netherlands. In the boundary zones the array of standard-exceeding compounds clearly deviates from that in the rest of the Netherlands, with compounds authorised in the neighbouring countries but not in the Netherlands, such as flufenacet, featuring prominently. The share of the neighbouring countries in the total number of exceedances in the Netherlands is roughly proportional to the relative area of the border zones. Although there is a certain influx of pesticides from across national borders, the magnitude of the problem appears to be limited. Keywords: pesticides, surface waters, Netherlands, Belgium, Germany, exceedance, foreign contribution, water quality standards INTRODUCTION Compared with other European countries, per-hectare pesticide consumption in the Netherlands is relatively high: 2.7 kg/ha on average (Eurostat, 2011). The compounds are used mainly in agriculture, but also in public parks and greenery, on paved surfaces and in a variety of private applications. Some of these pesticides end up in surface waters. Surface water quality, including pesticide occurrence, is monitored routinely by regional water authorities and the Ministry of Infrastructure and Environment. Since 1993 the monitoring data have been collated at Leiden University and published in the form of a Pesticides Atlas accessible on the inter-

2 2 net: (De Snoo et al., 2006). This atlas, covering the , contains information on the occurrence of approximately 500 different pesticides (and metabolites thereof) as measured at around 700 Dutch surfacewater monitoring stations (Van t Zelfde et al., 2010; Vijver et al., 2008; Van t Zelfde & Vijver, 2008). These measurements show that between 1997/1998 and 2009 the percentage of pesticides exceeding the Maximum Permissible Concentration (MPC) in the Netherlands declined by 50-70%. Nonetheless, many pesticides are still encountered in high concentrations, with frequent exceedance of national standards like MPC as well as the European standards laid down in the Water Framework Directive (AA-EQS and MAC-EQS). The Dutch government has pledged to improve surface water quality. One of the main instruments available for this purpose is authorisation policy, which prescribes which pesticides may be used on which crops, when, where and under what circumstances. A pesticide application may thus be either authorised or nonauthorised (possibly after previously being admitted). As a matter of course, Dutch pesticide policy is geared to pesticide users in the Netherlands and the assumption generally adopted is that any contamination of surface waters derives from activities within the country. As its name already says, though, the Netherlands is a lowlying country and straddles the delta of the Rhine, Maas, Scheldt and Eems river basins. Besides the major rivers, these basins also comprise numerous smaller waterways that flow into the Netherlands from its two higher-lying neighbours: Belgium and Germany. Some of the pollution burden of Dutch surface waters may therefore derive from the transboundary waters coming in from these two adjacent countries and their respective hinterlands (in particular, France). In these countries not only are different pesticides authorised than in the Netherlands, but the nature and scale of applications may also differ. As a result, compounds that are not (or no longer) authorised in the Netherlands may still find their way into this country. This study seeks to answer the question to what extent the countries bordering on the Netherlands contribute to the pesticide loads encountered in Dutch surface waters. More specifically, the following issues are addressed: - At how many sites in the vicinity of Dutch borders has monitoring taken place and what percentages of standard-exceeding pesticides and measurements are observed there? - In these border zones, are more non-dutch-authorised pesticides and standards exceedance thereof observed than in waters further from the border and in the rest of the Netherlands? - In what respects do the numbers, nature and concentrations of standardexceeding compounds in waters in the border zone differ from those in waters further from the border and in the rest van the Netherlands? - What is the maximum contribution of neighbouring countries to pesticide standards exceedances in the Netherlands? To answer these questions an analysis was made of pesticides occurrence in Dutch surface waters in the vicinity of the Belgian and German borders ( border zones ), waters further from the border ( reference zones ) and waters in the rest of the Netherlands. In some of the analyses a distinction was also made between flowing waters, possibly entering from neighbouring countries, and stagnant waters. For the latter it was assumed that these are burdened solely by local, Dutch activities.

3 Comm. Appl. Biol. Sci, Ghent University, 76/2, MATERIALS AND METHODS Pesticide data This study makes use of monitoring data incorporated in the Dutch Pesticides Atlas (Van t Zelfde et al., 2010). For each of the water quality monitoring sites in the Netherlands this database contains monitoring results and derived products on a range of pesticides and metabolites measured in Dutch surface waters in the For comparing the situation in the border zones with that in the rest of the Netherlands, monitoring data spanning 5 years ( ) were taken from the Atlas. Using a multi-year time series allows a larger number of monitoring sites to be included in the further analysis. Breakdown of the Netherlands (and monitoring sites) into zones For the purpose of the study a 5-km-wide border zone was defined along the entire length of the Netherlands border with its two neighbours, Belgium and Germany, with on the Dutch side of this zone a 5-km-wide reference zone. Throughout the analysis a distinction was made between Germany and Belgium because pesticide use and policy in the two countries differ. As an illustration, Figure 1 shows a section of the Dutch-Belgian border, indicating the various zones and the monitoring sites within them. The chosen width of the zones is such that there are sufficient monitoring sites in each zone for robust conclusions to be drawn about any differences in monitoring data and exceedances of standards. The rest of the Netherlands, beyond the reference zones, is also defined as a separate zone. Figure 1: Section of the 5-km-wide border zone and reference zone along the border with Belgium showing the monitoring sites in each zone.

4 4 Table 1 shows the number of monitoring sites in operation in the of study ( ). As can be seen, the percentage share of monitoring sites in each of the zones is approximately the same as their percentage share in terms of physical area. Table 1: Distribution of monitoring sites of pesticides over the various zones, with their respective percentage areas Zone No. of monitoring sites % of monitoring sites Area of zone (km 2 ) % of total area Belgian border zone Belgian reference zone German border zone German reference zone Rest of the Netherlands Total Flowing versus stagnant waters The two principal pathways by which legally applied pesticides can enter the Netherlands from neighbouring countries are via air and water, in the latter case mainly via waterways. A distinction was therefore made between monitoring sites on flowing and stagnant waters, using the variable water class given in the Pesticides Atlas for each of the monitoring sites. Reference standard To assess standards-exceedance the legislative standard in force in the Netherlands was taken: the Maximum Tolerable Risk (MPC), because this has already been set for virtually all the pesticides monitored, in contrast to the standards of the new European Water Quality Directive, AA-EQS and MAC-EQS. The MPC standard is derived from ecotoxicological data. Pesticide authorisation For all the pesticides monitored it was established whether or not they are authorised in the Netherlands or were so in the past. Because the analysis concerned a multi-year time series, three categories of pesticides were distinguished: 1) those authorised in the entire, 2) those authorised in part of the, 3) those non-unauthorised in the entire (Ctgb 2011). For non-authorised pesticides with a high degree of standards-exceedance in one or both border zones it was determined whether these were authorised in the neighbouring countries during the under review.

5 Comm. Appl. Biol. Sci, Ghent University, 76/2, Calculations In the Pesticides Atlas most of the results are calculated as an annual figure per monitoring site. To derive a 5-year value for each of the zones distinguished, in this study we first aggregated the data for each site over the 5-year, then aggregated these data for the zone as a whole. In both these aggregation steps the values for the various compounds were averaged. Most of the calculations described below were carried out for all the compounds together as well as for three authorisation categories : authorised in entire, authorised in part of and non-authorised in entire. The following calculations were performed: - Percentage of standard-exceeding compounds per zone. At each monitoring site, for each compound the 90-percentile of the monitoring data in each year was first compared with the MPC standard. The results were then aggregated. - Percentage of standard-exceeding compounds per water class per zone. The same procedure as above, but now distinguishing between monitoring sites on stagnant and flowing waters. This was only done in the Belgian border zone and reference zone because more standard-exceeding compounds were observed there than in the German border zone. - Percentage of standard-exceeding measurements per zone. At each monitoring site the percentage of measurements of all compounds exceeding the standard (for the compound concerned) in each year was determined. The results were then aggregated. - Highest standard-exceeding compounds per zone. For each zone the standard-exceeding compounds were ranked using the Standard Exceeding Index, SEI (Van t Zelfde et al., 2010). This measure is based on two factors: the percentage of standard-exceeding monitoring sites per zone and the degree of standard-exceedance per site. This was first calculated per zone per year and then for the 5-year. Below, the compounds with the greatest degree of standards exceedance are discussed. - Inter-zone comparison of the SEI for each compound. In addition to an interzonal comparison of the ranking of problematic compounds, the calculated SEIs can also be compared per compound and per zone. This was done for the five highest-ranking non-dutch-authorised standard-exceeding compounds in the Belgian border zone, the German border zone and the rest of the Netherlands. RESULTS Percentage of standard-exceeding compounds per zone In Table 2 the average percentage of standard-exceeding compounds in the various zones is reported for each of the three authorisation categories (entire, part of the, none of the ). In the Belgian border zone the percentage of such compounds is higher for pesticides not authorised for use in (part or all of) the than in the reference zone and far higher than in the rest of the Nether-

6 6 lands. For the German border zone these percentages are similar to those in the reference zone and in the rest van the Netherlands. Table 2. Percentage of standard-exceeding compounds per zone Category authorised, entire authorised, part of non-authorised, entire Belgium Germany Border zone Reference zone Border zone Reference zone Rest of Netherlands Flowing versus stagnant waters For a closer look at the situation along the Belgian border, in Table 3 the data of Table 2 are broken down further according to whether the standards-exceeding compounds were observed in flowing or stagnant waters. As can be seen, in the Belgian border zone it is mainly pesticides authorised only in part of the or not at all that are encountered relatively more often in flowing waters than in stagnant waters. In the reference zone this effect is clearly less pronounced. In the border zone the percentage of authorised compounds in flowing waters is lower than in stagnant waters. Table 3. Percentage of standard-exceeding compounds in flowing and stagnant waters Category Belgian border zone Belgian reference zone Flowing Stagnant Flowing Stagnant authorised, entire authorised, part of non-authorised, entire Percentage of standard-exceeding measurements Table 4 shows the average percentage of standard-exceeding measurements in the various zones for the three authorisation classes. For the Belgian border zone this percentage is similar to or slightly higher for the compounds non-authorised in part or all of the compared with the reference zone and slightly higher compared with the rest of the Netherlands. For the German border zone no such effect can be observed and the percentages are similar to the reference zone and the rest of the Netherlands. The percentage of standards-exceeding measurements, for authorised compounds, in the Belgian reference zone is substantially higher than in the other zones.

7 Comm. Appl. Biol. Sci, Ghent University, 76/2, Table 4: Percentage of standard-exceeding measurements per zone Category authorised, entire authorised, part of non-authorised, entire Belgium Germany Border zone Reference zone Border zone Reference zone Rest of Netherlands Highest standard-exceeding compound per zone To quantify the degree of standards-exceedance use was made of the Standard Exceedance Index (SEI, see Materials and Methods). Table 5 provides a breakdown of the 25 highest standard-exceeding compounds by authorisation category and zone. As can be seen, the split is approximately the same for all the zones. Table 5: Breakdown of the 25 highest standards-exceeding compounds in each zone by authorisation category Category authorised, entire authorised, part of non-authorised, entire Belgium Border zone Reference zone Germany Border zone Reference zone Rest of Netherlands To illustrate the interzonal differences in the main standard-exceeding compounds observed, Table 6 lists the ten highest standard-exceeding compounds in each zone. As can be seen, imidacloprid is a problematic compound in every zone and, in line with Table 5, there are no major interzonal differences in the number of compounds in each of the authorisation categories. In the border zones, nonauthorised compounds rank higher than in the rest of the Netherlands. The compounds observed in the Belgian and German zones differ from the list for the rest of the Netherlands. Some of the non-dutch-authorised compounds occurring in the Belgian and German border zones in Table 6 are authorised in these neighbouring countries (Fytoweb 2011; BVL 2011). A case in point is flufenacet, which features in the top ten for both the Belgian and German border zone, but not in that for the rest of the Netherlands.

8 Table 6: Top ten standard-exceeding compounds per zone. Legend: (p) = Dutchauthorised in part of, (n)=non-dutch-authorised; no code = authorised; Compounds: diethyltoluamide=deet, trichloorfon=dep 8 Belgium Germany Border zone Reference zone Border zone Reference zone Rest of Netherlands 1 imidacloprid imidacloprid imidacloprid terbuthylazin, desethylterbuthylazin, desethyl- 2 flufenacet (n) chloorpyrifos terbuthylazin, imidacloprid Imidacloprid desethyl- 3 DDT, 44 (n) bromacil (n) ETU ETU captafol (n) 4 diethyltoluamide diethyltoluamide methiocarb tricyhexatin (p) captan 5 aldicarbsulfoxide dichloorvos propoxur azoxystrobin ETU 6 difenoconazool captan flufenacet (n) terburtryn (n) aldicarbsulfoxide 7 chloorpyrifos isoproturon (p) flurtamon (n) diethyltoluamide pirimifos-methyl 8 diuron (n) monolinuron (n) terbuthylazin malathion (p) triflumuron (n) 9 isoproturon (p) spiromesifen DDT, 44 (n) DDE, 44 (n) trichloorfon (n) 10 kresoxim-methyl kresoxim-methyl chloorpyrifos propoxur methoxychloor (n) Value of Standard Exceedance Index Using the Standard Exceedance Index, an interzonal comparison of individual compounds can be made. In Figure 2 the SEI of the five highest-ranking non-dutchauthorised compounds found in the Belgian zone is compared across the various zones. As can be seen, all five of these compounds have a far higher SEI in the Belgian border zone than in the rest of the Netherlands. Flufenacet, for example, has a high SEI in both border zones, while encountered nowhere else in the rest of the Netherlands. Also noteworthy is that DDT has a far higher SEI in the two border zones compared with the rest of the Netherlands, even though this compound has been banned in Europe for several decades. Although the difference in the SEI of the five highest-ranking non-dutch-authorised compounds is less pronounced in the German border zone, the same basic picture emerges here, the main difference lying in the precise list of pesticides encountered here. The five highest-ranking non-dutch-authorised compounds occurring in the rest of the Netherlands occur infrequently in the two border zones, if at all. These are clearly compounds that are problematic at the national level.

9 Comm. Appl. Biol. Sci, Ghent University, 76/2, Figure 2: Interzonal comparison of the SEI of the five non-dutch-authorised compounds ranking highest in the Belgian border zone Contribution to standards exceedance in the Netherlands Table 7 shows the maximum contribution of the neighbouring countries to pesticides occurrence in Dutch surface waters, calculated as the average percentage of standard-exceeding measurements, corrected for the number of monitoring sites in the zone in question relative to the total number of sites. As can be seen, the maximum contribution in the two border zones is lower than to be expected on the basis of the area of the zones. In the German border zone it is in fact far lower. Tabel 7: Degree of contribution to standards exceedance per zone relative to percentage area. The maximum percentage is based on area and relative share in standard-exceeding measurements Zone Percentage area of zone Maximum percentage contribution Belgian border zone Belgian reference zone German border zone German reference zone Rest of Netherlands Total

10 10 CONCLUSIONS, DISCUSSION AND RECOMMENDATIONS The overall conclusion is that while there is a certain degree of pesticide influx from neighbouring countries, the contribution to the sum total of pesticide standards exceedance in the Netherlands appears to be limited. This conclusion holds both for the reference zones adjacent to the border zones and for the rest of the Netherlands, as the percentage of standard-exceeding measurements in the border zones and in the rest of the Netherlands is similar. Given the limited area of the border zones (approx. 5%), the share in total standards exceedance is approximately the same as the percentage area. However, the percentage of non-dutchauthorised pesticides exceeding the standard in the Belgian border zone is clearly higher than in the rest of the Netherlands or in the German border zone. These cases of standards exceedance are observed particularly in flowing rather than stagnant waters. The relative number of non-dutch-authorised compounds contributing most to standards exceedance is similar in the border zones and the rest of the Netherlands. To some extent the array of standard-exceeding compounds in the Belgian and German border zones differs from that found in the rest of the Netherlands, although imidacloprid is the main pesticide exceeding the standard in all the border and reference zones. In addition, several non-dutch-authorised compounds have a higher degree of standards exceedance in the border zones than in the rest of the Netherlands. This holds both for compounds authorised in the neighbouring countries like flufenacet and for non-authorised compounds like DDT. These conclusions are obviously influenced by the assumptions made in this study and the methodology adopted to establish the contributions of the neighbouring countries. Some of the choices made in the methodology were pragmatic, though realistic, as in the case of the five-km width of the border zones, for example. One key assumption is that the compounds observed in the border zones derive from across the border, entering the Netherlands via waterways. It will be clear, however, that (agricultural) activities in these zones will also play their part in this respect. Moreover, no effort was made to establish any (illegal) contribution of use in the border zones of non-dutch-authorised compounds obtained across the border. Furthermore, the adopted aggregation procedure, per monitoring site and per year, then averaged per site over the five-year and ultimately averaged for each entire zone, is of influence on the results. A different aggregation method would have yielded higher or lower figures. However, this would not have affected the relative differences between the zones. A final methodological issue is that there was no separate analysis of the monitoring data for the major rivers, which, while transporting a much greater volume of water than the smaller transboundary waters, have relatively far fewer monitoring sites. It should be added, though, that the Netherlands Association of River Waterworks, RIWA, reports regularly on water quality at the entry points of these rivers to the Netherlands and their data show that various pesticides are also present in the incoming river water (RIWA 2011). The percentage of standard-exceeding measurements in the Belgian border zone is similar to that seen in the rest van the Netherlands, although a higher percentage of compounds was observed in this zone compared with the rest of the Netherlands. This holds in particular for the contribution of non-authorised compounds to standards exceedance. This means that the number of standard-exceeding measurements per compound is lower in this border zone. Another result to emerge is

11 Comm. Appl. Biol. Sci, Ghent University, 76/2, that the list of standard-exceeding pesticides found in the border zones differs in certain respects from that in the rest van the Netherlands. This appears to be due partly to differences in authorisation policy between the Netherlands on the one hand and Germany and Belgium on the other. A case in point is flufenacet, use of which is authorised in Belgium and Germany but not in the Netherlands and which is clearly a greater problem in the border zones than within the bulk of the Netherlands. Another example is DDT, which shows a higher degree of standards exceedance in the Belgian border zone than in the rest of the Netherlands or in the German border zone. Noteworthy, finally, is that the main standard-exceeding compounds in the rest of the Netherlands also include a fair number of non-dutchauthorised pesticides. One complication in this respect is that while certain compounds may no longer be authorised, in such cases existing stocks may still be legally applied for a of two or more years. These conclusions lead to the following recommendations. There is a need to further substantiate the assumptions and methodology used in this study to establish the contribution of neighbouring countries to pesticide contamination of Dutch surface waters, based for example on mass-balance or drift-deposition calculations (Duyzer & Vonk 2003). In addition, monitoring sites located in the neighbouring countries should also be included for purposes of comparison. Pesticide influx via the major rivers should also be separately considered and the data compared with the figures for the rest of the border zones. Finally, this study represents a firstpass analysis of transboundary pesticide problems, and the aggregation procedure, the manner in which percentages were calculated and the results themselves all deserve further quantitative analysis. More nuance could also be introduced with respect to the contribution of non-authorised compounds, by making due allowance for the s in force for using up existing stocks. The main recommendation, however, is that there should be greater harmonisation of national pesticide authorisation policy among neighbouring countries, certainly where waterways run from one country into the next. In European authorisation policy, too, this is an issue that deserves greater attention. ACKNOWLEDGEMENTS We thank Corine van Griethuyzen (Ctgb) for her information on pesticide authorisation in the Netherlands and abroad. Thanks are also due to Kees Musters (CML) for his critical contribution to data analysis and interpretation. We are also very grateful to Nigel Harle of Gronseveld for his editing of the Dutch manuscript and his translation of the text into English. REFERENCES BVL - Bundesamt für Verbraucherschutz und Lebensmittelsicherheit. (2011). ctionproducts/02_onlinedatabase/plantprotectionproducts_onlinedb_node.html;jsessionid= 06FDC04BD4D88F718E8853D1F0EC3190.1_cid103

12 12 Ctgb.(2011). _schema=portal Duyzer J.H. & Vonk A.W. (2003). Atmospheric deposition of pesticides, PAHs and PCBs in the Netherlands (translation of R2002/606) rep. No. R2003/255., TNO, Delft. Eurostat.(2011). Fytoweb. (2011). RIWA. (2009). Jaarrapport 2010, de Rijn, RIWA, Nieuwegein. Snoo G.R. de, Tamis W.L.M, Vijver M., Musters C.J.M. & Zelfde M. van t. (2006). Risk mapping of pesticides: the Dutch atlas of pesticide concentrations in surface waters: Comm. Appl. Biol. Sci. 71(2a): Vijver M.G., Van t Zelfde M., Tamis W.L.M., Musters C.J.M. &De Snoo G.R. (2008). Spatial and Temporal Analysis of Pesticides Concentrations in Surface Water: Pesticides Atlas. Journal of Environmental Science and Health Part B. 43: Zelfde M. van 't, Tamis W.L.M., Vijver M.G. (2010). Technische rapportage van de update van de bestrijdingsmiddelenatlas met gegevens van het jaar Leiden: Institute of Environmental Sciences (CML). Zelfde M. van 't, Musters C.J.M., Tamis W.L.M. & Vijver M.G. (2010). Technische rapportage van project: Bestrijdingsmiddelenatlas Kader Richtlijn Water (KRW) proof. Leiden: Institute of Environmental Sciences (CML). Zelfde, M., van 't & Vijver, M.G. (2008). Technische rapportage van de update van de bestrijdingsmiddelenatlas met gegevens van de jaren 2005/2006. Leiden: Institute of Environmental Sciences (CML).

13 Comm. Appl. Biol. Sci, Ghent University, 76/2, Legend figure 1