| Main authors: | Meindert Commelin, Shaun Coutts, Jantiene Baartman, Isobel Wright, Antonio Ferreira, Gerard Velthof, Oene Oenema and Violette Geissen |
| FAIRWAYiS Editor: | Jane Brandt |
| Source document: | »Commelin, M. et al. 2020. Identification of most promising measures and practices: 1. Reduction of diffuse pesticide transport from agricultural land to groundwater and surface waters bymanagement practices. FAIRWAY Project Deliverable 4.3, 72 pp |
| Contents table |
| 1. Literature synthesis |
| 2. Meta-analysis |
| 3. Case studies |
To categorize the measures and to evaluate their effectiveness, the first aspect that needs to be taken into account is the main transport pathway of the pollution process (Rittenburg et al., 2015). Pesticides are mainly transported by air, water and soil. The main transport agent is water, and in some cases also soil particles can transport pesticides, when these in turn are carried by water (Reichenberger, Bach, Skitschak, & Frede, 2007). Reducing diffuse pollutant transport is linked to the transport routes shown in Figure 1.2. Pesticide transport to water bodies can be reduced by either decreasing the input of pesticides into the system (e.g. less or no application) or by influencing the hydrological flow paths and thus reduce the off-site transport of the pesticides. The pathways that are identified as main transport routes to groundwater and surface water are: overland flow, subsurface flow and drainage, leaching and drift.
Figure1.2
1. Literature synthesis
Results from major reviews since 2000 until present were synthesized (Y. Liu et al., 2015; Reichenberger et al., 2007; Rittenburg et al., 2015; Tang et al., 2012; Wauchope, 1978), including the reviews about specific measures in relation to pesticide pollution (Alletto, Coquet, Benoit, Heddadj, & Barriuso, 2010; Felsot et al., 2017; Krutz, Senseman, Zablotowicz, & Matocha, 2005). From these reviews and extra literature that was collected for the meta-analysis, a qualitative overview is made of the most used and studied measures to reduce pesticide pollution of groundwater and surface waters (Table 1.3).
Table 1.3: Synthesis of literature results: effectiveness and costs of key measures. Symbols are explained below the table
| Measure [source] | Effectiveness | Costs | Notes [source] | |
| Groundwater | Surface water | |||
| 1. Vegetated filter strips | + | +++ | €€ | Effectiveness depends on design, added ecological value (Arora, S. K. Mickelson, & J. L. Baker, 2003; Krutz et al., 2005; Rafael Muñoz-Carpena, Ritter, & Fox, 2019; Reichenberger et al., 2019) |
| 2. Constructed wetlands | + | +++ | €€€ | Effectiveness depends on local design. (Moore, Schulz, Cooper, Smith, & Rodgers, 2002; Stehle et al., 2011; Tournebize, Chaumont, & Mander, 2017; Vymazal & Březinová, 2015) |
| 3. Erosion reduction | - | +/- | ? | (Fawcett, Christensen, & Tierney, 1994; Sadeghi & Isensee, 2001) |
| 4. Tillage intensity | +/- | +/- | € | Effectiveness depends on local design (Alletto et al., 2010; Elias, Wang, & Jacinthe, 2018; Tang et al., 2012) |
| 5. Drainage optimization | ? | + | € | (Flury, 1996) |
| 6. Residue management/ Mulching | ? | + | € | (Alletto et al., 2010) |
| 7. Drift reduction | na* | ++ | €€ | High ecological value (Al Heidary, Douzals, Sinfort, & Vallet, 2014; De Snoo & De Wit, 1998; Felsot et al., 2017; Hilz & Vermeer, 2013; Otto, Loddo, Baldoin, & Zanin, 2015) |
| 8. Crop rotations | ++ | ++ | €€ | (Brown & Van Beinum, 2009; Rittenburg et al., 2015) |
| 9. Application rate reduction | + | + | € | (Reichenberger et al., 2007) |
| 10. Alternative pesticide | ? | ? | ? | Depends on choice (Reichenberger et al., 2007) |
| 11. Integrated Pest management | ++ | ++ | €€€ | (Gentz, Murdoch, & King, 2010; Reichenberger et al., 2007) |
NOTE: Symbols in the table indicate a scale from negative to positive with – is negative, +/- is neutral and +++ is very positive, this is a qualitative overview since quantitative data is not generally presented in the reviews. For the cost three categories were made, as follows: low (€), moderate (€€) and high (€€€). An ? indicates that no clear data is available and the evaluation of the measure is still unknown. * not available: this transport route does not exist.
1.1 Tillage practices
Runoff and infiltration processes on the field are strongly related to tillage practices, and thus tillage practices influence the transport pathways of pesticides. Alletto et al. (Alletto et al., 2010) extensively reviewed the effectiveness of tillage practices on both overland and leaching transport of pesticides. For both overland and leaching transport, changes in tillage practices were effective, but local design and application were very important for success (Alletto et al., 2010). Ghidey et al (2005) found that incorporation of applied pesticides below the upper 2-5 cm of the soil is one of the most effective ways to reduce overland flow of pesticides. A meta-analysis of papers after 1985 showed that no-till practices have a higher overland transport of pesticides compared to conventional tillage, including plowing (Elias et al., 2018).
The costs of changing tillage practices are generally low and practicability and feasibility of changing to other tillage practices is good (Reichenberger et al., 2007). However there is a risk that tillage practices will not remediate total pesticide pollution but only change the transport route, because infiltration (leaching) and overland transport are mutually exclusive (Rittenburg et al., 2015). Tillage alters the soil hydraulic properties and thereby the transport pathways of water and related solutes such as pesticides (Alletto et al., 2010). Conservation tillage (i.e. no-tillage or reduced tillage) increases retention/sorption of pesticides in the topsoil (Elias et al., 2018), particularly because of retarded degradation of soil organic matter compared to tillage, this decreases the availability of pesticides for biological degradation, leading to enhanced persistence in soils (Alletto et al., 2010).
1.2 Vegetated filter strips
A widely used measure to reduce pesticide pollution by overland transport are VFS. They are used to reduce the negative effects of overland flow, and are initially designed as erosion reduction measures. However they also affect pesticide transport. Most filter strips are located at the downstream end of a field, were runoff water leaves the field. VFS have been shown to be effective in reducing overland flow and soil erosion (Krutz et al., 2005; Lerch, Lin, Goyne, Kremer, & Anderson, 2017). They reduce pesticide loss by
- facilitating the deposition of particles which sorb pesticides,
- enhancing pesticide retention / sorption by increasing the time available for infiltration,
- sorbing dissolved-phase herbicides to the grass, grass thatch and soil surface, and
- reducing the volume of overland flow containing dissolved and particulate-associated pesticides [12,23,39].
Performance of the vegetated filter strips for pesticide trapping depends on the hydrological conditions (e.g. precipitation, infiltration and overland flow), the strip design; strip width, area ratio and type of vegetation cover (Krutz et al., 2005) and characteristics of the particles and pesticides (Tang et al., 2012). The effect of a buffer strip on ground water pollution by pesticides is mentioned as a potential risk of increased leaching, however no data was found during the search for this study. In the past two decades several models, both empirical and mechanistic, have been developed to predict the retention capacity of a VFS. The VFSMOD model by Munoz-Carpena (1999) is further developed and shows to perform well in combination with either empirical or mechanistic pesticide retention equations (Reichenberger et al., 2019). The empirical, revised Sabbagh equation (Sabbagh, Fox, Kamanzi, Roepke, & Tang, 2009) is based on an extensive dataset with experiments from the last decades:
[3]
with ΔPthe pesticide trapping efficiency (%) , ΔQ the infiltration in the buffer (% of total inflow), ΔE the sediment trapping in the buffer (% of total inflow), Fph the solid-dissolved distribution (%) of the pesticide and C the percentage organic matter in the incoming sediment. The equation performs well (R2 of 0.82) (Reichenberger et al., 2019).
However, while the empirical model performed well, it does not really explain the process of VFS effectiveness. Therefore, Reichenberger (2019) proposed a mass-balance equation as a more mechanistic approach:
[4]
Where Vi is the incoming water (L), Ei the incoming sediment (kg) and Kd (L/kg) the sorption coefficient. This model also performs well against empirical data (R2=0.77), and it is regarded as a good predictor for the effectiveness of a VFS.
If well designed and adjusted to local conditions, vegetated buffers are very effective measures, as indicated by the above formula (Reichenberger et al., 2019). The costs are estimated to be moderate, including implementation and maintenance costs and loss of productivity on the area of the field that is used as buffer (Rittenburg et al., 2015).
1.3 Constructed wetlands
Constructed wetlands are less studied than vegetative filter strips but if well designed, maintained and implemented they can be very effective with rates of pesticide reduction up to 100% (Tournebize et al., 2017; Vymazal & Březinová, 2015). A meta-analysis of the existing data until 2011 showed that the main influential parameters for effectiveness are pesticide characteristics, vegetation type and coverage (Stehle et al., 2011). However the costs are high, and they can take a relatively large surface of productive land to be installed.
1.4 Subsurface flow and leaching
For situations/locations with mainly subsurface flow, reduction of pesticide loss to surface and groundwater is challenging because altering the pathway of water flow is difficult. Source input management (i.e. Integrated Pest Management) is possible and pipe drainage may decrease the overland flow volume. However, drains may create subsurface flow paths and do not necessarily reduce overall pollutant transport (Rittenburg et al., 2015). Locations with deeply drained soils and thus a risk of leaching to groundwater benefit most from input control measures and increased residence time in the mixing layer (0-5 cm from the soil surface) to enhance degradation of the pollutant (Alletto et al., 2010).
1.5 Spray drift reduction
Spray drift is a pollution pathway that is different from the other pathways in the sense that no water (flow) is involved. Preventing drift is mainly done by reducing the transport route from the spraying device to offsite areas including open water bodies. Input control is the most effective measure to reduce drift pollution, because if less or no pesticide is sprayed there is less potential pollution. Buffer zones and application technology are effective measures to reduce drift after spraying (Felsot et al., 2017; Hilz & Vermeer, 2013). In addition, no spray zones and windbreaks often have a high ecological value, and within the EU are rewarded within the Common Agriculture Policy (Reichenberger et al., 2007). Optimizing droplet size and speed, in combination with applications during the correct meteorological conditions, greatly reduce drift risk (Al Heidary et al., 2014; De Cock, Massinon, Salah, & Lebeau, 2017). Mechanical drift reduction consists of a broad spectrum of technologies to reduce drift by changing spraying nozzles (Al Heidary et al., 2014) or ventilator design (Otto et al., 2015).
2. Meta-analysis
The quantitative dataset collected for this paper was sufficient for a meta-analysis on two different tillage practices; tillage and vegetated filter strips (VFS). For VFS a more detailed analysis was conducted, to understand which variables influence effectiveness.
2.1 Tillage practices
Conventional tillage uses cultivation as the major means of seedbed preparation and weed control. It typically includes a sequence of soil tillage, such as ploughing and harrowing, to produce a fine seedbed, and to incorporate the plant residue from the previous crop into the soil.
In case of no-till the soil is often only disturbed once, where the new crop is sown directly into the harvested field. No-till management is associated with higher amounts of organic matter on the soil surface (Alletto et al., 2010).
For the general analysis the data is separated into two groups; pesticide transport to groundwater and to surface waters, indicated by “leaching” and “overland” in Figure 1.4, respectively.
Figure 1.4
It was expected that no-till management would reduce pesticide transport, because on its performance in terms of erosion. However, the results show for both transport pathways that conventional tillage results in less pesticide pollution. Overall, the effect of tillage was significant with a reduction of 51%. The effectiveness is higher for leaching to groundwater than for overland transport; 55% and 50% pollution respectively. However the effect was not significant for leaching due to the high variation in the dataset.
A meta-regression analysis was done, to evaluate the effect of pesticide adsorption coefficient (Koc ) on the effect size. However all studies reported data for pesticides in the moderate class (Koc=75-500). The meta-regression did not give any significant effect, but this might be strongly related to the absence of low or high sorbing pesticides in the dataset.
2.2 Vegetated filter strips
VFS are a measure to reduce overland transport of pesticides. In the dataset 38 comparisons were found. Recent papers used more data (Reichenberger et al., 2019), however the statistical constraints of the meta-analysis strongly reduced the available data.
The general effectiveness of VFS is good with all data points showing a reduction of pesticide transport (Figure 1.5). On average the reduction of pollution is 53%, with a 95% confidence interval of 39% - 65%. Variables that might influence the effectiveness of VFS are pesticide type, strip dimension and area to buffer ratio. Buffer area to source area classes are; low >0.08, moderate 0.08 – 0.04 and high <0.04. Figure 1.5 shows no significant effect between buffer to upstream area ratio. This is in accordance with the empirical and mechanistic models presented by Sabbagh (5) and Reichenberger (6) where VFS pesticide removal is influenced by the reduction of water and sediment in the strip and the chemical adsorption properties of the pesticide involved. Based on this relation, there are optimal dimension for a certain location (Rafael Muñoz-Carpena et al., 2019).
Figure 1.5
Figure 1.6 shows the response ratios for a subgroup analysis on characteristics of involved pesticides. The data is categorized based on the Koc and solubility values of the pesticides in three groups, according to the values presented in table 2. In the lowest category for adsorption (Koc) only one study was included, so no statistics can be shown for this group. Although the visualized trend is not significant the data suggests that a higher sorbtivity will result in the most effective trapping of pesticides by VFS. For solubility there does not seem to be a clear relation. This is expected because VFS tend to reduce the sediment concentration in runoff water by trapping the sediment. The larger the fraction of pesticide connected to sediment, the more effective the measure will be. This result corresponds with the mechanistic and empirical models by Sabbagh and Reichenberger (Reichenberger et al., 2019), where the relation between trapping of pesticide and Kd is in the same direction.
Figure 1.6
3. Case studies
Table 1.5 shows the types of measures implemented in 7 of the FAIRWAY the case studies including evaluation factors as graded by local experts. The measures have been evaluated for the costs and effectiveness for reducing pollution of groundwater and surface water.
Table 1.5: Particular applied measures studied within the FAIRWAY case studies, with indicated properties based on expert judgement by experts working in the case study.
| Measure | Involved Countries5 | Effectiveness | Costs | |
| Groundwater | Surface water | |||
| Safe pesticide cleaning and storage facilities | NL, NIR | +/- | + | ? |
| Safe storage unit for pesticides | NIR | ? | + | ? |
| Vegetated filter strips | FR, SI | ? | +++ | €€ |
| Crop rotation improvement | FR | ++ | ? | €€€ |
| Input reduction | FR, UK | ++ | ++ | €€€ |
| Network engagement1 | UK | ? | + | ? |
| Alternative (pesticide or mechanical) | UK, NIR | ? | + | ? |
| Integrated Pest Management2 | UK, DK | +++ | + | €€ |
| Obligatory reduced input | PT, DK, SI | +++ | +++ | € |
| Bio filters/beds | NIR | ? | ++ | ? |
| Economic/Tax management3 | DK | +++ | ? | €€ |
| Drift reduction | NL | ? | ++ | €4 |
NOTE: Symbols in the table indicate a scale from negative to positive with – is negative, +/- is neutral and +++ is very positive.
Costs are categorized as; low (€), moderate (€€) and high (€€€). No data as ‘?’.
1Network engagement: embedding information and communication at all levels stimulate change of practice.
2Intergrated Pest Management, is a holistic method to reduce pesticide use, by using alternative mechanical and biologic pest management in combination with adjusted cropping and resource management.
3These measure increase the price of pesticides, as an extra incentive to look for alternative crop management methods.
4Low cost or on the long term even benefits due to reduced use of pesticides
5Abbreviations of countries are: NL, The Netherlands; NIR, Northern Ireland; FR, France; SI, Slovenia; UK, United Kingdom; DK, Denmark; PT, Portugal.
There is a clear distinction between cases where groundwater or surface water is the main area of pollution. No clear mechanical measures are available to reduce leaching risk of pesticides to groundwater. In these cases (Denmark, United Kingdom, Slovenia and Portugal) the main approach is to reduce the input into the system, or the change to an alternative pesticide with lower pollution risks. The reduced input is often enforced by laws or policy, and sometimes subsidies are used to make implementation feasible. This type of measure will reach many land managers because they are enforced at a national or sub-national level.
Reducing input of pesticides often requires a broader ‘redesign’ of the farming system. The effectiveness of crop rotation and integrated pest management (IPM) depends a lot on the design and on climatic and farm specific conditions, so this type of measure can be effective but has a low adoptability because it needs changes in farming system (Balderacchi & Guardo, 2008).
In cases of overland transport and surface water pollution there is beside diffuse sources, also a lot of attention for point source pesticide pollution, in these cases pesticides will reach surface waters by drainage from the storage/cleaning areas or from accidental spills away from the handling area. This can lead to high concentration in surface waters and there are good working measures to reduce yard associated problems. Examples are wash and load basins (Northern Ireland and the Netherlands) or biobeds/filters that degrade the pesticides before it reaches the surface water.
Diffuse source measures implemented in the cases in Europe are vegetated filter strip (VFS) for sloping agricultural areas (France and Slovenia) and drift reducing measures (NL). Both measures are well studied and also discussed in earlier sections.
In terms of costs and application by farmers the general trend is that measures are not cost effective, so they are either enforced or subsidized to stimulate implementation. One exception are drift reducing measures, which can be cost effective due to decreased pesticide use (NL).
Note: For full references to papers quoted in this article see