SETAC Spotlight: Moving towards field specific risk assessment: A drainflow risk assessment case study in the UK

The registration of plant protection products in Great Britain includes an assessment of risk to surface water via agricultural drains using the MACRO preferential flow model. The assessment considers the number of years in which the regulatory acceptable concentration (RAC) is exceeded for 20 pedo-climatic scenarios (5 drained soils and 4 climates – dry, intermediate, wet and very wet). Depending on the species driving the RAC up to 3 (aquatic invertebrates and fish) or 18 (aquatic plants and algae) out of the 30 years simulated may exceed the RAC in any given pedo-climatic combination. As such, even if a small number of pedo-climatic combinations fail the risk assessment no product registration is possible even though the simulations indicate that there are many pedo-climatic combinations which demonstrate safe usage along with all undrained soils. This presentation outlines a spatially distributed drainflow modelling (SDDM) approach that would allow vulnerability maps that identify areas that may pose a risk to edge of field waterbodies to be developed and also inform a decision support tool that would allow farmers to comply with a spatially varying label restriction precluding application to specific soil classes that are drained. This approach is closely aligned with the increasing prevalence of precision agriculture, digital pesticide labels and field specific mitigations to achieve required product label mitigation levels.

Soil classes previously defined for landscape modelling in a Welsh Assembly Government (183/2007/08) and CRD projects (PS2238; PS2245) were used to define a broader suite of drained soils (see Table 1) than is currently used within the scenario-based risk assessment. The spatial extent of these was defined by the 1 km resolution NATMAP dataset in England and Wales and the 1:250k soils dataset in Scotland. The most recent 30 years of weather data for each MARS 25 km resolution grid square covering GB were processed to produce MACRO daily weather files. The extent of winter and spring cereal cropped area within GB was defined using the 2 km resolution EDINA 2010 landuse dataset. MACRO v4.4.3 simulations were carried out for each pedo-climatic-crop combination for a range of substances with varying environmental fate properties. In keeping with the standard scenario risk assessment, simulated daily fluxes of water and solute from a 1 ha field were diluted into an edge of field waterbody comprising 30,000 L with a flow rate of 0.35 m/s. Annual statistics of daily mass/water flux, ditch concentration and RAC exceedance were compiled.

An assessment of these annual statistics in conjunction with previous in-field/field-level level risk indicators, e.g. CRD project PS2245 and UKWIR projects 15/DW/14/11 & 16/DW/14/14, was conducted to define an approach that is closely aligned with the current regulatory process to facilitate CRD acceptance whilst also being user friendly to facilitate farmers’ and spray contractors’ ability to comply with a spatially varying label restriction.

Of the approaches considered, the number of daily exceedances of the RAC in each year was preferred as this is closely aligned with the scenario based regulatory approach. While the soils data is expressed at 1 km resolution, this is derived from soil association maps and as such each 1 km grid square in many cases has multiple drained soils classes. Through considering the soils class that produced the most RAC exceedances and the class that produced the least, 6 potential vulnerability classes were established (see Figure 2).

Figure 2: Vulnerability map illustrating the 6 risk classes defined from the soil classes in each 1 km grid square producing the most and least annual RAC exceedances for an active substance and where (a) up to 3 or (b) up to 18 in 30 years may exceed the RAC. Areas in green have predominantly undrained soil whilst those with no vulnerability class have no arable land.

Vulnerability maps using these classes demonstrate that (a) there are many areas (light cream colour) where all drained soil classes would support safe usage; (b) there is a small area (dark blue) where no usage should take place as all drained soils classes exceed the RAC threshold and (c) there is a wide swathe of agricultural land (intermediate blues) where one or more soil class suggests risk and further assessment would be required. Drawing on these results, previous field scale risk indices and accounting for dataset and regulatory uncertainty, a decision tree is outlined to inform user decision making. The manner in which this decision tree might be built into farm software and decision support tools is also considered.

A spatially distributed drainflow modelling approach has been established which would facilitate user compliance with spatially varying label restrictions. This is closely aligned with the existing regulatory modelling and decision-making processes to facilitate regulatory adoption. Moving to a field-based risk assessment approach would allow GB growers access to a more diverse suite of plant protection products whilst maintaining protection of the environment.

Acknowledgement – The authors thank Bayer AG, Crop Science for sponsoring this project.

If you would like to discuss any of the topics raised in this article, feel free to contact us.

This slide-deck is available for free download.

You can find all of the other posters that CEA presented at SETAC Seville here. You can also find all of our publications from previous conferences and links to journal articles we have authored on our library page.

Enjoyed this article? Avoid missing out on our future news articles by signing up – you can select which topics you are interested in and can unsubscribe at any time.