|Main authors:||Meindert Commelin, Jantiene Baartman, Piet Groenendijk, Oene Oenema, Susanne Klages, Isobel Wright, Tommy Dalgaard, Morten Graversgaard, Jenny Rowbottom, Irina Calciu, Sonja Schimmelpfennig, Nicola Surdyk, Antonio Ferreira, Violette Geissen|
|FAIRWAYiS Editor:||Jane Brandt|
|Source document:||»Commelin, M. et al. 2018. Review of measures to decrease pesticide pollution of drinking water sources. FAIRWAY Project Deliverable 4.2, 79 pp|
Here we present the results of the systematic literature search and data analyses. In total 37 sources about pesticide measures were reviewed, 4 studies were excluded because of study design of data incompleteness. The 33 studies that were analysed contain 104 experimental comparisons on the effectiveness of pesticide measures. Of these 8 cover groundwater pollution, 88 contain data about surface water pollution (of which 36 are specific about drift reduction) and 9 experiments cover both ground and surface water pollution. Table 5 shows the topics covered by the collected sources, groundwater and surface water pollution have been separated, because transport mechanisms in this two cases differ a lot.
Table 5: Summary of database content for pesticide mitigation aimed at decreasing pesticide pollution of groundwater and surface waters (Status 1 October 2018)
|Pathway||Pesticide category||# experiments||Type of measures|
|Groundwater (11)||Herbicides, haulm destructors and moss killers||9||
|Surface water (61)||Herbicides, haulm destructors and moss killers||51||
|Insecticides and acaricides||2||
|Fungicides and bactericides||2||
The analysis of the effect of the measures is done for each of the three pathways separately. The pesticide categories have been combined during this analysis.
The literature search and data screening took more time than initially expected and the number of studies included in the database is less than initially expected. The 37 studies included in the database have been conducted mostly in the EU-28, but some studies originate from other continents. Most of the studies from EU originated from western Europe. Most studies dealt with pesticide overland transport reduction with buffers and tillage methods. As such, the database is not well balanced with a proper distribution of studies across measures. Our preliminary conclusion is that more literature sources need to be added to the database to allow a more balanced and complete meta-analysis.
Based on the aforementioned conclusion, the decision was taken to apply the method of meta-analysis on the available data and present here a first quantitative analyses of the measures in the database. A full meta-analysis of an updated database is given in »Most promising measures to decrease nitrate pollution of groundwater and surface waters.
|1. Summary overview of the effectiveness of measures|
|2. Drift reduction measures|
|3. Measures to reduce groundwater pollution|
|4. Surface water|
The effectiveness of measures was derived from the response ratio (RR), which is the pesticide pollution under a treatment measure divided by the pollution of the reference treatment (control treatment). The latter is usually current practice or conventional practice. The ratio may vary from 0 to more than 1; a value smaller than 1 indicates that the treatment measure decreases the pesticide loss relative to the reference treatment. A ratio of 1 means no effect, and a ratio above 1 indicates a negative effect. Instead of a relative comparison of pesticide loss, the response ratio was sometimes derived from a comparison of pesticide concentration in waterbodies or from the pesticide content in the soil between treatments, depending on the availability of the data in the reviewed publications (»Review methodology).
Table 6 provides an overview of the response ratio RR of some key treatment measures. The overall mean RR ranged from 0.2 to 5.6, indicating a wide range of effectiveness of the measures. Most measures had an RR in the range of 0.2-0.5. Treatments related to tillage methods did not result in effective decrease of pollution, in fact for both ground and surface water the mean response ratio was above 1, which means the loss of pesticides increases. This overview suggests that vegetative buffers, reducing overland transport, and mechanical drift reduction are the most effective measures.
Table 6: Summary of response ratios for each measure.
|Pathway||Measure||Response Ratio (RR)||n*|
|Drift||Mechanical drift reduction||0.26||27|
|No spray zones||0.30||9|
|IPM/ input control||0.22||4|
|Surface water||IPM/ input control||0.41||9|
*number of experiments
This first analysis shows that treatment measures greatly differ in their effectiveness and that there is a large variability in effectiveness within a set of treatment measures. A few additional comments have to be made here. Firstly, the number of studies/comparisons differed greatly between treatment measures; some of the treatment measures (e.g. tillage methods and vegetative buffers for surface water) have a much greater experimental basis than others (e.g., IPM and input control). Secondly, the mean response ratios have as yet not been corrected for the number of measurements and variance within studies. Third, the effectiveness of the treatment measures has not been analysed taking into account different environmental and socio-economic conditions. These aspects need to be taken into account while further analysing a (larger) database.
Drift pollution is a specific pathway of pesticides to surface water, or offsite areas. When pesticides are applied by spray application, a part of the spray liquid can be transported through the air to other locations. Two main approaches to reduce this transport are:
- technical modifications in the spraying technique to reduce the potential of drift transport and
- no spray zones between an application field and open water sources.
These two approaches are both analysed in more detail.
Mechanical drift reduction
Mechanical measures to reduce spray drift are often related to the nozzle type of the sprayer, the height of the spraying boom above the surface and the driving speed. Adjustments always have to optimize both the reduction of spray drift as well as a uniform and efficient application of the pesticide to the field/crop. Figure 14 shows the effect ratios for the gathered experiments, it is clear that all tests show a decrease in pollution; all effect ratios are below 1.
The mean effect ratio of drift reduction measures is 0.26 with the 95% Confidence Interval (CI) ranging from 0.0 to 0.53, showing a strong significant effect with the control treatment. This shows that technical drift reduction technologies are very effective to reduce pollution of off-site locations. In the reviewed experiments these were often streams and open water bodies bordering the agricultural fields.
No spray zones
Another option to reduce pollution by spray drift is to create no spray zones between the open water sources and agricultural fields. These areas will function as a buffer for the occurring spray drift and prevent the pollution of the open water.
Based on the results in Figure 15, no spray zones seem effective to reduce pesticide pollution. The 95% CI ranges from 0.0 to 0.76 with a mean effect ratio of 0.30. In this analysis the width of the buffer zone is not taken into account, because of the low number of study results. However it is likely that the width of the no spray zone is a co-variable to explain effectiveness, where broader zones will be more effective.
Pollution of groundwater occurs mainly through transport of pesticides with leaching of water deeper into the soil. To reduce this, the amount of input in the system can be changed by for example integrated management or reduction of applied pesticides. Besides that the soil management has a large influence on infiltration and leaching of pesticides, optimizing the tillage method can be a measure to reduce pollution of groundwater by pesticides.
In the database one experiment studied the effect of increased vegetative cover on leaching to groundwater, this study showed a positive effect with a response ratio of 0.68. However no further analysis can be done with only one data point.
The mean response ratio of tillage methods to reduce leaching towards groundwater is very high. As Figure 16 shows it is far above 1 indicating an increase in pollution by changing the tillage method.
The mean effect ratio of these 6 experiments is 5.6, and the distribution of the results is very large. To get a better insight in the actual effect more studies should be added to this analysis. Besides that, the effectiveness is influenced by the chosen reference point. For tillage methods, conventional tillage in used as a reference, in the studies by Hall et al., (1991) and Watt et al., (1996) the treatments are either mulch tillage or no tillage, these land management methods are well known for their erosion reducing effect and an increase in infiltration. The reviewed data agrees with that because an increase in infiltration may also lead to higher leaching of pesticides to groundwater. However, this also means that when the reference is changed e.g. there is a switch from a no tillage system to conventional tillage, this data suggest a decrease in pesticide leaching.
IPM and input control
Managing the amount of pesticides that are applied to the system is an effective way of reducing pollution. When the input is reduced this will very likely also show in a reduction in pollution. However, when reducing the input of pesticides the productivity of the system should be kept at a good level, otherwise the efficiency of these measures is still low.
Because there are only a few experiments in this database, input reduction and IPM are combined. The difference between both is that IPM is a more holistic approach taking into account the needed adjustments of the whole agricultural system, where input control experiments often only test the effect of a reduced pesticide application. Figure 17 shows the effectiveness of experiments with input control to reduce leaching of pesticides.
The mean effect ratio is 0.22 which indicates a strong effect of input control on pesticide leaching to groundwater, however there is a very low number of reviewed experiments (n=4), which should increase to obtain a better insight in the effectiveness of input reduction. For input control the amount of reduction will very likely be related to the effectiveness so this should be taken into account as a co-variable.
Most studies within the collected database focus on transport of pesticides to surface water. The main pathway for this pollution is via overland transport during and after intense rainfall events. Many measures aim at reducing the runoff potential by delaying runoff time or increasing infiltration. Besides that also measures for input control are used. Some studies differentiate between transport by water and adsorbed transport with eroded sediment. However in this analysis both processes are combined and total overland transport is used.
Input control and IPM
As for pollution to groundwater also surface water pollution can be reduced by decreasing the input of pesticides into the system. Figure 18 shows the distribution of response ratios for input control measures to reduce surface water pollution.
The mean response ratio 0.41 which tends towards an effective measure, however the variation in the results is large and there is no significant difference with the reference studies. As described for grondwater, a strong relation is expected between the input of pesticides and the pollution on agricultural fields. The mean response ratio for surface water is a bit higher than for groundwater which indicates less effect. To improve the analysis the number of experiments will be increased and the amount of input reduction will be used as a co-variable to explain the effectiveness of the measure.
Overland transport reduction
If the pesticide is applied to the crop/field, it should remain on the field without being transported to other locations. For many herbicides overland transport is a main pathway towards open water bodies. Several measures are reviewed to minimize the transport along this pathway.
Vegetated buffer strips are used to decrease the flow velocity of the runoff water, in this way the potential infiltration is increased, moreover this vegetative buffers also function as erosion reducing measures. In Figure 19 the effectiveness of buffer strip experiments is shown. All experiments show a decreasing effect on the transport of pesticides to groundwater.
The mean effect ratio of applying buffers at the edge of a field is 0.28 with a 95% CI ranging from 0 to 0.70, which means there is a significant effect. To understand the influence of buffers better, the buffer width or the buffer to source area ratio are good co-variables to explain effectiveness. However the gathered studies do not consistently include the required data for the meta analysis. Reichenberger et al. (2007) also attempted to review the effect of buffers quantitatively, however they did not come to a final effectiveness, due to the high heterogeneity within the data.
Changing the tillage method influences the infiltration characteristics of the soil and by altering the surface roughness also the potential runoff of water. The reference treatment for tillage methods is fixed at conventional ploughing, as in the analysis for groundwater pollution. The results of the tillage studies is mainly spread around 1 with a group of outliers to the negative effect side (fFgure 20). This result was also found for tillage methods and groundwater pollution. However more data is needed to apply a co-variable analysis to understand the wide spread of results. Often the results depend on the agricultural system in which the tillage measure is applied. The mean effect ratio is 2.5 but this is strongly affected by the outliers.
Within the reviewed database also a group of measure was collected which aimed to reduce the amount of erosion on the field, and through that indirectly the transport of pesticides. These measure will be mainly effective when erosion and runoff occur a lot.
The mean effect ratio of erosion reducing measure is 0.52 (Figure 21), however this does not give a significant difference with the reference. This is mainly because of the low number of experiments included in the analysis.
Within the dataset also experiments with drainage and crop type changes were included. These were too few for a statistical analysis and to show quantitative results. For drainage there is a trend that shows that with less drainage, or a wider spacing of the drainage pipes there is less transport through the drainage system. Changing land use influences the dynamics within a system, often the type and amount of applied pesticides will change and the hydrologic conditions will be altered. To unravel and understand the effect of landuse changes, more data with well-defined co-variables is needed. For example the change from arable land to forest is not comparable with the change from a monocrop orchard to an intercropping system.
Note: For full references to papers quoted in this article see