1. Discussion

This study combines several approaches to review the effectiveness of agronomical measures to reduce pesticide pollution to water resources. The combination of literature synthesis, meta-analysis and evaluation by experts gives a unique overview of the performance of selected measures.

The meta-analysis performed agrees with the results of measure specific analyses (Elias et al., 2018; Reichenberger et al., 2019). Across a wide range of studies vegetative buffer strips and conventional tillage significantly reduced the transport of pesticides into water resources. This study reinforces the meta-analysis by combining them with a literature synthesis and specific case studies. These additional approaches also provide extra context and an unique overview of the effectiveness of agronomical measures to reduce pesticide pollution in water resources.

In the literature synthesis and the meta-analysis we focused on on-site measures to reduce pesticide transport, the case studies did implement some of these measures, but also other regulatory and system wide measures are used in these cases. In countries where leaching of pesticides to groundwater is the main threat (e.g. DK, NL, SL) reducing the pesticide input is the most used measure. This can be applied through regulations of changing the farm system. The results from the literature synthesis and the data collection are in agreement with this finding, because these studies show no effective on-site measures to reduce leaching. The on-site measures that were presented in the literature are only scarcely applied in practice. For example VFS are suitable for use in France and Slovenia, but not much applied there. The questionnaire results point out that this can mainly be explained by a lack of enforcement (»Annex 1). The case studies show that point source pollution management is an important measure to reduce pesticide pollution. In several case studies, reduction of point source pollution is promoted, with clearly defined measures, and positive results for surface water pollution. The contribution of point source pollution to total pesticide pollution of surface and groundwater sources is unclear. A modelling study in Germany estimated it to be very low (Bach, Huber, & Frede, 2001), while estimates from UK are much higher with contributions up to 40% (CPA, 2010).

The main goal of the meta-analysis is to understand the transport potential of pesticides on agricultural land and the effectiveness of measures to decrease this. Covering the entire spectrum of possible pesticide types and important physical characteristics related to transport is important. However, the studies in the dataset do not cover this entire spectrum. For example, for sorbtivity of pesticides, the systematic search resulted in sufficient data only on both ends of the range, very mobile and very immobile. In addition, mainly pesticides with a short half-life time were studied, which will persist in the system for a shorter period, leaving a knowledge gap at the transport and movement of more persistent pesticides.

The current analysis is carried out using mainly load (mass per area per year) as indicator for pollution. However, the concentration is used as indicator for many regulations. As also shown by Elias (2018) results can differ substantially when using concentrations instead of loads. Especially for event based transport, a concentration might become too high with still an overall low load on annual basis. For this analysis the choice for loads was made, because it represents better how much pesticide is lost in total.

When comparing the results of the literature synthesis with the meta-analysis many similarities emerge. Literature is clear about the potential effectiveness of vegetative filter strips (VFS) (Krutz et al., 2005; R Muñoz-Carpena, Fox, Ritter, Perez-Ovilla, & Rodea-Palomares, 2018; Reichenberger et al., 2019), although they have to be designed to match local conditions. The calculated pollution reduction of VFS is high (53% - 39 – 64 CI 95%), indicating that these are very effective measures to decrease pesticide pollution. For VFS recent model developments show that adapting the dimensions to local conditions is possible (Rafael Muñoz-Carpena et al., 2019).

Tillage practices are clearly related to transport pathways of pesticides and are therefore extensively studied (Alletto et al., 2010). A recent meta-analysis on tillage effects on pesticide pollution has created more insight into this relation (Elias et al., 2018). Where earlier sources expected that erosion reduction would also reduce pesticide pollution (Alletto et al., 2010; Reichenberger et al., 2007; Rittenburg et al., 2015), it is shown by Elias et al. (2018) that no-till systems do not reduce pesticide loads, and may make them worse. Our study, comes to the same conclusion, showing that conventional tillage often leads to less pesticide transport, using a slightly different set of papers for the meta-analysis. This reinforces the result and shows it is general enough to be replicated by an independent study. Substances with higher solubility and lower sorbtivity tend to be transported more under no till management. This process is linked to the influence of no-till management on soil properties like organic matter and pH (Alletto et al., 2010). The relation between organic matter content and pH and the phase distribution of pesticides influences the transport by overland runoff. Under no-till management the soil mixing layer is more shallow than in cultivated soils, which might result in more potential transport during overland flow. To prevent this, incorporating pesticides into the 2 – 5 cm layer will enhance their degradation (Ghidey et al., 2005) thereby reducing the risk on overland transport.

An effective way of reducing pollution by pesticides is by input control or by redesigning the farming system. Input control and redesigning farming systems are farm level measures. In the reviewed literature the main focus is on diffuse pollution, or pesticide transport from the field. Although point source pollution also occurs it is identified as less complex to control (Bach et al., 2001). Reduced input and redesigning the farming system is often referred to as ‘Good Agricultural Practices’ (GAPs) or ‘Best Management Practices’ (Rittenburg et al., 2015), which are agricultural management practices aiming at minimizing off site movement of pesticides to surface waters. Examples of such practices include band spraying on row crops, application restrictions for vulnerable soils and/or wet climates and keeping a certain distance from adjacent water bodies when spraying (Tang et al., 2012). Also the timing of pesticide application (with regards to e.g. forecast of heavy rainfall) or an integrated approach to pest management is important to reduce pesticide pollution (Gentz et al., 2010).

The systematic search for articles resulted in only a limited amount of usable data sources. This is mainly due to the requirement of statistical data (mean, SD and n) to be able use statistical models for the analysis. This problem with quality of presented data is commonly mentioned in recent meta-analysis studies (Elias et al., 2018; Valkama, Usva, Saarinen, & Uusi-Kämppä, 2019). For most measures within the data set, the amount of data was not sufficient for a thorough meta-analysis. Based on this, we would like to stress the need to always present the mean, SD and sample size of data with experimental results.

Within the case studies examples are given of national laws or regulations which restrict or prohibit the use of pesticides. Such measures are effective on higher policy levels and not reviewed or studied in detail in this study. However, this might be a promising approach to reduce pesticide pollution of drinking water sources. It is evident that there is no single strategy to reduce pesticide losses. When aiming at transport reduction, site-specific plans that are well managed may provide greatest success (Rittenburg et al., 2015).

In the context of preserving the quality of water resources, both surface and ground water, the reduction of pesticide transport is of vital importance (Hildebrandt et al., 2008; Rittenburg et al., 2015). In this study we show that proper management on the field can contribute to reduced pollution from overland transport, but that for transport to groundwater no readily usable agronomical measures are available. To achieve reduction of pesticide pollution in water sources, measures should also include farm system redesign, reduced inputs and regional or national approaches to facilitate a sustainable farming system.

For overland transport well-studied measures are available, such as vegetated filter strips. These measures can strongly decrease pesticide transport. However, to study the effectiveness of these measures, or of agricultural management in general in more detail, high quality and well documented experimental studies are of great value. While this paper discussed the effectiveness of single measures, the contribution of each flow path to total pollution is not specified. This needs further attention in the future, to provide the possibilities for optimal strategies and management to reduce pesticide pollution to ground and surface water resources.

2. Conclusions

The main conclusions of our study are

• The driving factors for diffuse pesticide pollution are (i) the amount and type of used pesticides, (ii) water facilitated transport through or over the soil, (iii) erosion of sediment that causes transport of sorbed particles, and (iv) spray drift during application.
• Vegetated filter strips are the most clear measure to reduce overland transport and pollution by pesticides. Models are available to calculate dimensions and predict effectiveness for pesticide reduction.
• Tillage practices are extensively studied in relation to off-site transport of pesticides. The analysis shows that no-till does not provide less off-site transport than conventional tillage, and suggests even higher pollution in no-till systems under specific circumstances.
• On-site measures against diffuse pollution comprise only a small part of the available approaches to reduce pesticide pollution. To obtain a sustainable system, input reduction, farm system redesign, point source mitigation and policy measures are essential to be taken into account.

Notes:

For full references to papers quoted in this article see »References

Download the full report for Annex 1