Site- and use-specific risk assessment for diffuse-source pesticide inputs into German surface waters

Transcription

1 4th European Modelling Workshop, Nov 2005, Paris Site- and use-specific risk assessment for diffuse-source pesticide inputs into German surface waters as part of the Information System on Integrated Plant Production (ISIP) M. Bach, S. Reichenberger, J.G. Höhn, H.-G. Frede, ILR, University Gießen

2 Overview 1. Information System on Integrated Plant Production (ISIP) 2. Materials and Methods 3. Selected Results 4. Discussion 5. Conclusions and Outlook

3 1. The Information System on Integrated Plant Production (ISIP) Online information system for farmers and plant protection advisors in Germany ( Responsible bodies: German agricultural chambers and federal states Pest infestation surveys, up-to-date weather data, individual forecasts and recommendations on pesticide applications Task of University Giessen: development of a risk assessment component for ISIP Look-up tables

4 2.1 Models 2. Materials and Methods Drainage simulations: MACRO 4.3b Preferential flow model Mechanistic or physically based, while still having a reasonable number of model parameters Reasonable validation status Official drainage model in the European pesticide registration procedure (FOCUS surface water scenarios) Runoff modelling: PRZM 3.21β Official runoff modell in the European registration procedure Relatively easy to parameterize Needs only little computation time

5 2.2 Modelling Approach Soil and climate scenarios Drainage (MACRO 4.3b): 8 soils, 19 soil/climate combinations, 1 representative weather year per climate scenario Runoff (PRZM 3.21β): 5 soils, 8 climate scenarios, 20 different weather years ( ) for each climate scenario Pesticide use scenarios MACRO: 109 different active ingredients (a.i.) with 229 uses registered in Germany PRZM: 185 a.i.; 446 uses

6 2.3 Simulations MACRO Simulation period 3 a (7 a), application in first year only Target variables: cumulative drainflow and pesticide loss via drainage maximum daily pesticide loss for the whole simulation period and for each calendary month (+ corresponding drainflow volume) PRZM Simulation period 2 a, application in first year only Target variables: cumulative runoff, soil loss, pesticide runoff and erosion losses maximum daily pesticide runoff loss for the whole simulation period and for each calendary month (+ corresponding runoff volume); analogously for erosion

7 2.4 Predicted Environmental Conc. Only acute exposure (initial PEC sw ) Erosion inputs are not considered Conservative estimate revealed: It can be justified to consider only runoff inputs for the calculation of initial PEC sw Consideration of regional variables (climate, land use, cropping percentage, river discharge flow duration curves, etc.)

8 2.4 PEC sw Calculation The PEC refer to the outlet of a small catchment (ca. 10 km 2 ; approximately covering the area of a smaller municipality) The PEC sw calculation basically dilute the max. daily pesticide input in the sum of drainflow or runoff volume and the stream discharge (75 th percentile of long-term month-specific discharge) on that day PRZM: The 90 th percentile of maximum daily loss (which corresponds to the weather year of the 20 with the 2 nd highest max. daily loss) is used for PEC sw calculation

9 2.5 Risk Assessment For each test species: ecotoxicological endpoint (EC 50, LC 50 ) required safety factor (TER of 10 or 100) Maximum tolerable concentration : highest concentration at which for all test species the required TER is met The higher of both calculated PEC sw (drainage or runoff) is compared with the maximum tolerable concentration. PEC sw > maximum tolerable concentration 'Danger to the aquatic ecology can't be excluded'

10 3. Selected Results Relative drainage losses are highest for weakly sorbing compounds Relative runoff losses are highest for moderately to rel. strongly sorbing (K oc ca L kg -1 ), relatively persistent compounds For most uses, PEC sw due to runoff are on average higher than PEC sw due to drainage For both input paths, the annual variability of the weather constitutes an enormous source of uncertainty

11 PEC sw : Example Isoproturon (IPU) Compound properties: normalized Freundlich coefficient K oc = 71 L kg -1 Freundlich exponent m = 0.88 lab degradation half-life DT 50,lab = 11 d max. tolerable concentration: 1.3 µg L -1

12 Drainage PEC sw for IPU in winter cereals, autumn application Results Exceedance only on clayey sites (soil class 8) Lower PEC sw for sandy soils and for drier regions Existing interdiction of IPU application on clayey soils is meaningful PECsw [µg/l] 0 > > 13 no data

13 Runoff PEC sw for IPU in winter cereals, autumn application Results PEC are strongly spatially variable (mainly due to the effect of soil, climate and area proportion cropped with wc) Exceedance of the max. tolerable concentration (1.3 µg L -1 ) for large areas PECsw [µg/l] 0 > > 13 no data Regionally differentiated exposure and risk assessment necessary

14 Cumulative distribution of PEC sw for isoproturon over all municipality-soil-combinations (winter cereals, autumn application) 1.0 kumulative Cumulative relative frequency Häufigkeit th percentile PEC sw due to runoff inputs max. tolerable concentration PEC sw due to drainage inputs PEC sw (µg L -1 )

15 Limitations Discussion So far no risk assessment for long-term exposure So far no erosion inputs considered Annual variability of the weather not sufficiently accounted for in the MACRO modelling remedy by switching to the faster MACRO 5 Validation status of the MACRO and PRZM is questionable for many soil/climate/application combinations Input data (e.g. market shares) have to be updated on a regular basis Possibilities Site- and use-specific online risk assessment for every farm in Germany ISIP: farmers have the possibility to include the risk to the aquatic environment in their decision which pesticide to apply

16 Regulatory relevant items The spatial differentiated approach allows to carry out probabilistic exposure assessments and to create risk maps for Germany (for a given use of a given substance). The overall results of MACRO 4.3b and PRZM 3.21 calculations for the entire area of Germany state, that PEC sw (initial) exceed max. tolerable concentrations for many situations (soils, climates, compounds). This outcome contradicts (partly) with the current risk judgement of the German regulatory practice ("no risk" for surface waters by pesticide losses via runoff and tile drainage). The scenario-based (soil, climate, agro-region, river basin specific) PEC sw calculation developed here can provide a viable basis for improving the regulatory practice in Germany.

17 Many Thanks... - to Stefan Reichenberger (main author, PhD thesis in preparation) -to Deutsche Bundesstiftung Umwelt DBU (German Environmental Foundation) for funding - for your attention! More information: martin.bach@agrar.uni-giessen.de stefan.reichenberger@agrar.uni-giessen.de