evidence project final reportrandd.defra.gov.uk/document.aspx?document=10599_ps2238...evid4 evidence...

EVID4 Evidence Project Final Report (Rev. 06/11) of 53

General Enquiries on the form should be made to:

Defra, Procurements and Commercial Function (Evidence Procurement Team) E-mail: [email protected]

Evidence Project Final Report

Note

In line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The Evidence Project Final Report is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website An Evidence Project Final Report must be completed for all projects.

This form is in Word format and the boxes may be expanded, as appropriate.

ACCESS TO INFORMATION

The information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code PS2238

2. Project title

Assessing the impact of minor PPP usage

3. Contractor organisation(s)

ADAS UK Ltd Pendeford House Pendeford Business Park Wolverhampton WV9 5AP

54. Total Defra project costs £ 148 515.00

(agreed fixed price)

5. Project: start date ................ 1 June 2010

end date ................. 31 October 2011

mailto:[email protected]

PS2238 - DRAFT REPORT 2

6. It is Defra’s intention to publish this form.

Please confirm your agreement to do so. ................................................................................... YES NO

(a) When preparing Evidence Project Final Reports contractors should bear in mind that Defra intends that they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.

Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the Evidence Project Final Report can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.

In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain


Executive Summary

7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

The potential aquatic exposure following the ‘minor usage’ of pesticides is not well understood, with the current perception being that (a) due to the scale of use or (b) there being a registered use on a major crop, that the risk to the environment is small. This project looked to inform this issue through the development of a risk assessment framework and tool that provided an objective way of assessing the relative risk of product use on minor crops versus usage of the same product on major crops at edge of field and catchment scales. The specific objectives of this study were to:

1. Collation of minor crop/usage and major crop extent datasets in order to inform a suite of representative minor uses for inclusion into the risk assessment tool. This likely spatial extent of crops/uses would be summarised within hydrometric catchments across England and Wales using different scenarios;

2. Development of a synchronous climate and river flow database to inform the dilution potential within each catchment;

3. Development of a FOCUS relevant field scale loss fate meta-model 4. Development of an edge of field water body fate meta-model 5. Development of a minor uses regulatory screening spreadsheet tool 6. Provision of a users workshop for the spreadsheet tool

Objective 1: Minor Crop/Usage Database In order to spatially distribute the risk assessment framework within the agricultural landscape it was necessary to compile catchment summaries of the major and minor crops that were to be included in the minor uses screening tool. These data were extracted from the ADAS 1 km resolution land use database using census and industry intelligence for the 2009 harvest year. Some crops required a refinement of this database, for example, the top fruit category was split into four sub-categories using cropping statistics and industry intelligence gathered by ADAS horticultural consultants. This database was expanded to include some minor crops, for example Hops/Nuts were sourced from the Rural Payments Agency while Christmas Trees were sourced from a range of data sources. In addition to the agricultural land uses, some amenity type land uses were also included namely golf courses and urban green spaces. In order to assess the potential minor use of plant protection products on major arable crops, an assessment of minor diseases, pests and weeds was undertaken using the best available data in conjunction with expert opinion from leading national specialists within ADAS. This analysis indicated:

Diseases – given the difficulty in predicting infection, treatment is typically prophylactic with minor diseases largely treated through incidental control;

Pests – minor pests are by their nature poorly documented and as such a complete assessment is difficult;

Weeds – most weeds are fairly ubiquitous with the degree of infestation being controlled by local factors

In addition, there would probably be limited instances where a Notifier would seek to register a compound used extensively on a major crop for a minor use on another major crop. As such we recommend that default pest/disease/weed parameters not be included in the minor uses tool. Rather the facility to include regionally variable treatment proportions (as has already been included in the beta version) be used along with statistics supplied by the Notifier for the minor use on the other major crop. Objective 2: Catchment Flow Meta-Model This meta-model was developed using streamflow data from the National River Flow Archive (NRFA) for 315 catchments in England and Wales where these flows are reported to be natural or have been naturalised and are longer than 10 years in record length, accounting for record completeness. Each catchment was assigned to the modelling climate region in which it is located. Synchronous periods of record of weather and stream flow were extracted from these datasets for each catchment. Streamflows associated with spray drift, runoff and drainflow events were extracted, adjusted for low flow periods in stressed water catchments and used in the construction of a series of regression models to predict these flows using catchment descriptor and flow variables. These models were typically very robust except for


one summer month drainflow model which was replaced by the next summer month model. Objective 3: Field Loss Meta-Model In order to populate the database of lookup values of predicted maximum fluxes of pesticide and associated water/sediment lost to surface water bodies via surface runoff and drain flow, it was necessary to parameterise and run millions of pesticide fate model simulations in accordance with standard regulatory guidance for a pre-defined set of substance-crop-pedo-climatic combinations. This was undertaken for surface runoff using the FOCUSSW PRZM v1.1.1 model and for drainflow using the FOCUSSW MACRO v4.4.2 model for 19 soils, 6 climate regions, 22 model crops, 12 monthly applications and 36 compounds. Objective 4: Edge of Field Fate Meta-Model An edge of field meta-model for predicting environmental concentrations of pesticides in edge of field surface water bodies was developed based on the Step 3 approach of the STEPS1-2-3-4 model of Klein (2007a, 2007b), which has undergone some validation. The edge of field water body meta-model was validated against standard FOCUSsw runs using test substances 1-sw through 7-sw for scenarios D2 ditch and R1 stream. The correlation between the meta-model and the FOCUS TOXSWA model results is high (R

2 = 0.9993).

Objective 5: Development of the Minor Uses Screening Tool The risk assessment tool was conceptualised around a suite of meta-models which would in essence “look up” appropriate values, but be underpinned by standard regulatory models/approaches and a suite of spatial databases. The first of these meta-models is the field loss meta-model which predicts the maximum flux of compound, the associated fluxes of water and sediment as well as the number of days after application that this would occur, using user inputs of compound properties, crop treated and application timings and amounts. The second is the edge of field fate meta-model which assesses the likely predicted environmental concentration in edge of field water bodies given these fluxes of compound, water and sediment. Comparison of these concentrations with a user input regulatory acceptable ecotoxicological concentration (RAC) allows for the completion of a risk assessment and the classification of all hectares of treated crop into two risk classes, ‘at risk’ or ‘not at risk’. This model also predicts the downstream loss of compound from edge of field water bodies. The third is the catchment flow meta-model which predicts the likely volume of catchment runoff into which the mass of compound lost downstream may be diluted to calculate a catchment concentration. Comparison of these catchment concentrations with user input regulatory (e.g. drinking water standard) or ecotoxicological thresholds (e.g. Water Framework Directive environmental quality standard) and an associated uncertainty factor, allows for the completion of a catchment risk assessment and the classification of all catchments into one of four risk classes, ‘at risk’, ‘possibly at risk’, ‘possibly not at risk’ and ‘not at risk’. The tool itself was developed in C# with the underpinning data and queries stored in a Microsoft SQL database. The tool outputs the tabulated edge of field and catchment scale results to Microsoft Excel and the catchment scale results to a bespoke ADAS spatial data viewer. While any evaluation of a complex modelling framework such as the MUSTool is difficult, the evaluation of both the underlying meta-models as well as the outputs from the tool suggest that the tool is fit for release as a beta version. It is recommended that future phases of work include:

1. Additional exploration of the MUSTool be undertaken using more highly resolved spatio-temporal pesticide usage and monitoring data such as that collected by and for water company catchment compliance and risk assessments. ADAS have collected such pesticide usage data for three catchments for two separate water companies and are aware of a number of catchments with high spatio-temporal resolution monitoring data collected by these and other water companies, primarily for undertakings to OFWAT and the Drinking Water Inspectorate.

2. Refinement of the edge of field meta-model to account for sediment associated pesticide being included in the catchment scale PECs, especially for high KOC compounds, and the inclusion of a sediment PEC as one of the outputs.

3. Review and implementation of spatial clustering metrics to objectively assess the degree of clustering of the cropping and ‘at risk’ areas.

4. Expansion of the spatial viewer to include non-categorical datasets such that the edge of field catchment results may be viewed in the spatial viewer.


Objective 6: Stakeholder Workshop A stakeholder workshop was held on the 28

th November 2011 at the offices of the Chemicals Regulatory

Directorate in York. This workshop was very productive and identified a mechanism by which the tool could be used within the regulatory framework and a number of refinements to both the interface to remove ambiguity and the reporting tool to facilitate easier and fuller interpretation of the outputs. Knowledge Exchange The approach and outputs from this project have been presented at a number of meetings, namely: Hughes, G.O., 2012. Demonstration of the CRD Minor Uses Screening Tool. Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 12 June 2012, York. Hughes, G.O., 2012. Impact of Minor Uses: Development of the CRD Minor Uses Screening Tool. Voluntary Initiative Water Sub-Group Meeting, 6 March 2012, Peterborough. Hughes, G.O., 2011. Impact of Minor Uses: Development of the CRD Minor Uses Screening Tool. Proceedings of the Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 14 June 2011, Solihull. Hughes, G.O., 2010. Can scale of use be used in aquatic risk assessments? Proceedings of the Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 17 June 2010, Warwick.


Project Report to Defra

8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include:

the objectives as set out in the contract;

the extent to which the objectives set out in the contract have been met;

details of methods used and the results obtained, including statistical analysis (if appropriate);

a discussion of the results and their reliability;

the main implications of the findings;

possible future work; and

any action resulting from the research (e.g. IP, Knowledge Exchange).

Introduction The term ‘minor uses’ encompasses both the use of a product on a crop that is minor in extent or usage for a particular problem that is sporadic or minor in extent (CRD, 2009). The potential aquatic exposure following the ‘minor usage’ of pesticides is not well understood, with the current perception being that (a) due to the scale of use or (b) there being a registered use on a major crop, that the risk to the environment is small. However, minor crop production may be quite intensive, highly localised and have greater inputs. As such the localised environmental risk could be greater, depending on spatial and temporal vulnerability factors. Similarly, for many minor usages the number of products available for use is limited and as such the ability to choose alternate less harmful products is not always possible if the risk from one product is deemed unacceptable by standard risk assessment approaches. This project looked to inform this issue through the development of a risk assessment framework that considers not only the extent of minor usage, but also its location; as this is fundamental to determining the likelihood and potential number of water bodies exposed, as well as the opportunity for dilution at the catchment scale, thereby contextualising the edge-of-field regulatory modelling results. The specific objectives of this study were:

1. Collation of minor crop/usage and major crop extent datasets in order to inform a suite of representative minor uses for inclusion into the risk assessment tool. This likely spatial extent of crops/uses would be summarised within hydrometric catchments across England and Wales using different scenarios;

2. Development of a synchronous climate and river flow database to inform the dilution potential within each catchment;

3. Development of a FOCUS relevant field scale loss fate meta-model 4. Development of an edge of field water body fate meta-model 5. Development of a minor uses regulatory screening spreadsheet tool 6. Provision of a users workshop for the spreadsheet tool

Risk Assessment Framework A key output of the project was the development of a risk assessment screening tool that encapsulated the minor uses risk assessment framework and provided an objective way of assessing the relative risk of product use on minor crops versus usage of the same product on major crops, at edge of field and catchment scales. The risk assessment tool was required to have short run times and produce a matrix like output for easy interpretation by risk managers. As such, the risk assessment tool was conceptualised around a suite of meta-models which would in essence “look up” appropriate values, but be underpinned by standard regulatory models/approaches and a suite of spatial databases (see Figure 1). The first of these meta-models is the field loss meta-model which predicts the maximum flux of compound, the associated fluxes of water and sediment, as well as the number of days after application that this would occur, using user inputs of compound properties, crop treated and application timings and amounts. The second is the edge of field fate meta-model which assesses the likely predicted environmental concentration in edge of field water bodies given these fluxes of compound, water and sediment. Comparison of these concentrations with user input regulatory or ecotoxicological thresholds and associated safety/trigger factors allows for the completion of a risk assessment and the classification of all hectares of treated crop into three risk classes. This model also predicts the downstream loss of compound from edge of field water bodies. The third is the catchment flow meta-model and predicts the likely volume of catchment runoff into which the mass of compound lost downstream may be diluted, to calculate a catchment


concentration. Comparison of these catchment concentrations with user input regulatory or ecotoxicological thresholds and associated safety/trigger factors allows for the completion of a catchment risk assessment and the classification of all catchments into one of four risk classes. The tool itself was developed in C# with the underpinning data and queries stored in Microsoft SQL databases. The tool outputs the tabulated results to Microsoft Excel and the visual output to a bespoke ADAS spatial data viewer.

Figure 1 Schematic of the minor uses screening tool risk assessment framework

Field Loss Meta-Model In order to populate the database of lookup values of predicted maximum fluxes of pesticide and associated water/sediment lost to surface water bodies via surface runoff and drain flow, it was necessary to parameterise and run pesticide fate models in accordance with standard regulatory guidance for a defined set of substance-crop-pedo-climatic combinations.

Pesticide Fate Models The FOrum for Co-ordination of pesticide fate models and their Use (FOCUS) surface water (FOCUSSW) MACRO v4.4.2 model was be used to define drain flow fluxes, while the FOCUSSW PRZM v1.1.1 model was used to define surface runoff fluxes (FOCUS, 2001). Standard FOCUS defaults were used for a range of parameters (see Table 1).

Soils The NSRI Natmap1000 soil dataset (Soil Survey Staff, 1983) was analysed with a view to defining broad soil groups that have similar hydrological responses (i.e. HOST class), depths, textures and organic carbon profiles. A total of 19 soil groups were identified within England and Wales (see Figure 2). For each group the dominant representative soil series was selected and a soil profile for inclusion in the two fate models constructed. The soils parameters required by PRZM v1.1.1 and MACRO v4.4.2 (FOCUS, 2001) were derived from/according to HSE CRD guidance (Beulke et al., 2002), PRZM guidance documents (Carsel et al., 1984; Carsel et al., 1998) and MACRO guidance documents (Jarvis, 1994; Jarvis et al., 1997; Stenemo and Jarvis, 2002; Stenemo and Jarvis, 2003; Jarvis et al., 2007) using soil property data taken directly from SEISMIC v2.0.6 (Hallett et al., 1995). Individual MACRO parameters were derived using the methods outlined below:

CTEN – (Beulke et al., 2002) using horizon clay content from SEISMIC

TPORV, WILT, RESID, GAMMA – SEISMIC;


XMPOR – Fitted Mualem Van Genuchten (MVG) curve to water release data in SEISMIC using SEISMIC MVG parameters and determined XMPOR at CTEN (Beulke et al., 2002);

FRACMAC and ASCALE - (Beulke et al., 2002);

ZLAMB – Fitted Brooks Corey curve to water release data in SEISMIC as well as XMPOR and determined ZLAMB by optimising the fit in accordance with Beulke et al., (2002);

KSM – Calculated using variables described above and the equation provided in Beulke et al., (2002);

KSATMIN – using equations provided in Beulke et al., (2002) adjusted for soil structural development according to Stenemo and Jarvis (2003);

ZN - (Beulke et al., 2002) implemented within the Footprint framework to determine values between the end members and for exceptional values (Jarvis et al., 2007);

ZP and ZA - (Beulke et al., 2002);

WATEN - (Jarvis et al., 1997) The number of soil layers were defined according to the FOCUS Groundwater I (FOCUSGW) guidance (FOCUS, 2000) resulting in 22 layers for each soil profile. The degradation rate within each soil layer was calculated using the standard guidance from FOCUSSW (FOCUS, 2001) and modified for depth in accordance with the FOCUSGW (2000) guidance. The lower boundary condition for the soils was set according to that summarised in Table 3. Tile drainage systems appropriate to each soil class and climate region (see Table 2) were determined using drainage design principles (Castle et al., 1984) and climate region averages of one day design rainfall defined for agroclimatic areas of England and Wales (Smith and Trafford, 1976). These design spacings are in line with published values (Ragg et al., 1984) except for soils where mole drains would be expected in addition to the tile drains. The wider tile drain spacing and deeper drain depth values used for the Hanslope and Denchworth soils, as opposed to the narrower mole drain spacings and shallower mole drain depths used in FOCUSSW and the UK higher tier drainflow scenarios, were deemed appropriate as they more appropriately represent the landscape effects of differential mole drain practices and survival as well as the means by which “drains” are represented in MACRO. The 2012 Defra Farm practice survey indicates that: 17% of holdings (not land area) have never repeated their moling and 37% have not renewed their mole drains for more than 6 years (6% for more than 11 years); between 2 and 4% of drained arable land (the area within individual fields rather than the whole field) has been directly affected by poorly functioning drains in the last 3 years. PRZM typically predicts losses via surface runoff and shallow sub-surface pathways such that there might be an overlap between the outputs from the two models. In order to account for this, the hydrological group of each drained soil was adjusted in an attempt to remove the subsurface flow component and assess surface runoff alone (Centofanti et al., 2008; Reichenberger et al., 2008).

Figure 2 Illustration of the spatial extent of the 19 soil classes defined in England and

Wales for the MACRO and PRZM simulation modelling


Table 1 Model defaults used in MACRO and PRZM

Soil Parameter Value Dispersivity (cm) (DV) 5

Mixing depth (mm) (ZMIX) 0.1

Initial pesticide concentration (mg m-3

) (SOLINIT) 0

Pesticide concentration at bottom boundary (mg m-3

) (CONCIN) 0

Critical air content for transpiration reduction (m3 m

-3) (CRITAIR) 0.05

Excluded pore volume (m3 m

-3) (AEXC) 0

Site Parameter Value

Pesticide concentration in rainfall (mg m-3

) (CONC) 0

Rainfall correction factor (RAINCO) 1

Snowfall correction factor (SNOWCO) 1

Rainfall intensity (mm h-1

) (RINTEN) 2

Snowmelt factor (mm oC d

-1) (SNOWMF) 4.5

Albedo (ALBEDO) 0.25

Wind Measurement height (m) (ZMET) (ZWIND) 10 Reference temperature for degradation coefficients (ºC) (TREF) (TBASE) 20

Evaporation depth during fallow period (cm) (ANETD) 15

Surface condition after harvest (ICNAH) [set to residue] 3

Manning’s roughness coefficient (-) (MNGN) 0.1

Reflectivity of soil surface to long wave radiation (fraction) (EMMISS) 0.96

Substance Parameter Value

Crop uptake (-) (FSTAR) (UPTKF) 0

Q10 factor for degradation rate increase, temperature increases by 10 °C (-) (QFAC) 2.58

Exponent for moisture correction of degradation rate (EXPB) (MSEFF) 0.7

Reference soil moisture for moisture correction of degradation rate (-) (MSLAB) 100

Diffusion coefficient for the pesticide in air (cm2 d

-1) (DAIR) 4300

Henry’s Law constant of the pesticide (-) (HENRYK) 0

Enthalpy of vaporization of the pesticide (kcal mol-1

) (ENPY) 22.7

Freundlich exponent (-) (FRNDCF) (FREUND) 1

Vapour phase pesticide degradation rate constant (d-1

) (DGRATE) 0

Diffusion coefficient in free water (m2/s) (DIFF) 4.98E-10

Table 2 Agricultural drain depth (m) and spacing (m) for the 5 drained soil classes for each rainfall region

Soil Drain depth

(m)

Drain spacing range according to Ragg et

al. (1984) (m)

Model drain spacing (m)

SW SE CW CE NW NE

Blackwood 0.9 20 – 40 30 37 32 43 30 35

Hanslope 0.8 40 – 80† 16 19 16 22 15 18

Bromyard 0.8 15 – 30 26 32 27 37 25 30

Dunkeswick 0.8 15 – 40 11 13 11 15 11 11

Denchworth 0.8 20 – 40† 6 8 6 9 6 7 † Drain spacing reflects the additional use of mole drains and sub-soiling


Table 3 Soil class characteristics for the 19 soil classes, including lower boundary conditions, PRZM hydrological grouping, IRRIGUIDE soils class and the proportion of infiltration contributing to runoff

Representative Soil

Bottom Boundary Condition PRZM Soil Hydrologic

Group IRRIGUIDE Soil Class

Proportion of infiltration

contributing to runoff

Blewbury Unit Hydraulic Gradient A Silty Clay 0.01

Wick Unit Hydraulic Gradient A Sandy Loam 0.14

East Keswick Unit Hydraulic Gradient B Clay Loam 0.41

Newport Unit Hydraulic Gradient A Sandy Loam 0.14

Carstens Unit Hydraulic Gradient B Clay Loam 0

Arrow Zero Flow B Sandy Loam 0.23

Kexby Zero Flow A Sandy Loam 0.26

Teme Zero Flow BC Clay Loam 0.33

Rivington Unit Hydraulic Gradient B Sandy Loam 0

Malham Unit Hydraulic Gradient B Clay Loam 0

Peaty† Zero Flow D Silt Loam 0.45

Blackwood Zero Flow A Sandy Loam 0.03

Ardington Percolation rate - water table height B Clay Loam 0.24

Hanslope Percolation rate - water table height B Silty Clay 0.03

Bromyard Percolation rate - water table height B Silt Loam 0.04

Dunkeswick Percolation rate - water table height C Clay Loam 0.01

Andover Unit Hydraulic Gradient B Clay Loam 0

Denbigh Zero Flow BC Clay Loam 0.32

Denchworth Zero Flow C Clay 0.03

† Peaty not modelled in MACRO – assumed zero PPP leached to drains owing to high OC content

Weather The MACRO model was set up to run with 6 years of weather data for model spin-up and then a single representative sixteen month assessment period, as per for FOCUSSW modelling (FOCUS, 2001). The PRZM model was set up to run with 6 years of weather data for model spin-up and then 30 years (1961-1990) of weather data, akin to the FOCUSSW modelling (FOCUS, 2001). This weather data was supplied by the Met Office for 6 climate zones as illustrated in Figure 3. Our initial intention was to use the Met Office HADUKP rainfall dataset (Alexander and Jones, 2001) to drive the fate models in each of five regions (see Figure 3). However, this rainfall is overly influenced by the upland areas in three of the five zones such that the number of rain days and total rainfall are over-inflated with respect to the more dominant lowland agricultural areas. In addition, these zones are not completely consistent with the river basin district boundaries. As such six new zones were defined by amalgamating the Met Office HADUKP zones with the UK river basin districts to ensure a consistent climate across each basin, resulting in the North West (NW), North East (NE), Central West (CW), Central East (CE), South West (SW) and South East (SE). Within each of these six zones, using all 5 km grid squares, the percentile of the annual average rainfall (AAR) that covered 90% of the arable land was determined. An appropriate MORECS grid square was then selected to be representative of this percentile of the AAR and the zone as a whole. The representative year for each climate zone in the 1961-1990 period was selected using the FOCUSSW approach (FOCUS, 2001) and assessed to be the 50

th percentile year with respect to annual rainfall and drain

flow. No representative year was selected for the runoff modelling and all years (1961-1990) were modelled and the 50

th percentile annual values extracted from the results. While this differs from the FOCUSSW approach it was

felt to be more realistic following an attempt to select representative years in the vein of FOCUSSW.


(a) (b)

(c)

Figure 3 Illustration of the (a) HADUKP (http://hadobs.metoffice.com/hadukp/), (b) river basin districts (http://www.wfduk.org/implementation/RBDmapfiles/) and (c) the climate zones adopted by this study with the driver MORECS grid squares selected indicated

Soil Water Budget Evaluation The surface water flow components of the model outputs for a 30 year simulation (1961-1990 plus 6 year model spin-up) were evaluated against the Standard Percentage Runoff (SPR) derived by Boorman et al. (1995) for each of the representative soils. In order to derive an estimate of SPR from the model outputs in line with that of Boorman et al. (1995), a series of assumptions was required. Firstly, in order to remove the need for correction owing to antecendant moisture conditions, only flows during the period where soils are at or near field capacity as defined by Smith and Trafford (1976) were considered. In addition, all surface runoff from PRZM was assumed to contribute to the SPR and only MACRO drainflow that occurs on, or the day after, a rain day was considered to contribute to the quickflow encapsulated within the SPR definition. The average SPR derived for each of the six climate regions for the soil typologies versus those derived by Boorman et al. (1995) are plotted in Figure 4a.

http://hadobs.metoffice.com/hadukp/

http://www.wfduk.org/implementation/RBDmapfiles/


(a) (b)

Figure 4 (a) Plot of Standard Percentage Runoff (SPR) modelled by Boorman et al. (1995) versus SPR modelled within this project for the 19 soil classes averaged across all six regions. (b) Including optimisation of infiltration fraction such that runoff soils (blue diamonds) now lie on the 1:1 line (black line) and regional drain prevalence such that the slope of a line fitted to the drained soils (red squares) approximates the slope of the 1:1 line

When SPR from this project is compared with modelled SPR from Boorman et al. (1995), with their concomitant modelling uncertainties, the correlation for a number of soils appears to be quite weak. This is especially the case for the undrained soils that have been attributed SPRs by Boorman et al. (1995), that exceed that modelled by PRZM by up to several orders of magnitude. While this difference in part reflects a field versus catchment scale difference, where additional processes like variable source areas are not being accounted for, they also partly reflect a deficiency in the PRZM modelling approach where (1) permeable soils overlying an impermeable substrate and (2) permeable soils overlying a shallow groundwater, are not explicitly accounted for. In both cases newly infiltrated water would “travel” via (1) shallow soil and (2) shallow groundwater pathways, to contribute to the quickflow in the SPR. In the first case, it would have been best to model an impermeable substrate at the base of the soil profile as opposed to free drainage and model the lateral flows, however, this functionality within PRZM does not appear to work properly for the pesticide mass flux. The proportion of infiltration required to increase the SPR to that of the HOST SPR (see Figure 4b) was determined for each soil in each region and averaged (see Table 3). This value is used within the edge of field meta-model implemented within the Minor Uses Screening Tool (MUST). To explore if the differences in the SPR values for drained soils were a function of the proportion of agricultural drains within the agricultural landscape (not all soils requiring drainage are drained), these values were factored by the proportion of soils requiring drains that are likely to have surviving agricultural drains (ADAS, 2002). The results (see Figure 4b) suggest that drain prevalence is a factor, for while the SPRs do not lie on the 1:1 line, they are close and their relative values are correct such that the slope of the line fitted to these points is not significantly different (t-test) to that of the 1:1 line.

Crops The crops parameterised and used within the modelling were selected following a pragmatic assessment of the major and minor crops whose cropping extent could realistically be defined such that valid crop-pedo-climatic combinations could be defined in each catchment. Each crop was assigned a model crop (see Table 4 and Table 5), the parameters for which were taken from standard FOCUSSW scenarios representing UK conditions, namely the D2 surface water scenario (FOCUS, 2001) and the Okehampton ground water scenario (FOCUS, 2000), as well as the FOCUSSW UK scenarios (Price et al., 2006; Price et al., 2007) and expert assessment for crops not covered within these. The actual and model crops are presented in Table 4 and Table 5 while the model crop parameters are presented in

0

10

20

30

40

50

60

0 10 20 30 40 50 60 70 80 90

Modelled SPR

HO

ST

SP

R

0

10

20

30

40

50

60

0 10 20 30 40 50 60 70 80 90

Modelled SPR

HO

ST

SP

R


Table 6 and Table 7.

Table 4 Summary of the major crops extracted from the ADAS land use database for circa 2009 and their model crops

Major Crop Model Crop

Winter wheat Winter Cereal

Spring wheat Spring Cereal

Winter barley Winter Cereal

Spring barley Spring Cereal

Oats Winter Cereal

Mixed grain Winter Cereal

Potatoes Potatoes

Sugar beet Sugar Beet

Winter OSR Winter OSR

Spring OSR Spring OSR

Field Beans Spring Field Beans

Peas dry Legumes

Maize Maize

Root fodder crops Sugar Beet

Other crops for stock feeding Winter Cereal

Temporary grassland Grass/Alfalfa

Permanent grassland Grass/Alfalfa

Rough grassland Grass/Alfalfa


Table 5 Summary of the minor crops extracted from the ADAS land use database for circa 2009 and their model crops

Minor Crop Model Crop

Rye Winter Cereal

Triticale Winter Cereal

Linseed Winter Cereal

Apples Pome/Stone Fruit

Pears Pome/Stone Fruit

Plums Pome/Stone Fruit

Other top fruit Pome/Stone Fruit

Strawberries Small Fruit

Raspberries Cane Fruit

Blackcurrants Cane Fruit

Wine grapes Vines

Other small fruit Cane Fruit

Hops Hops

Christmas trees Small Trees

Bulbs and flowers grown in the open Bulb Flower

Asparagus Bulb Veg

Beans Legumes

Beetroot Sugar Beet

Field celery Leafy Veg

Lettuce & leaf vegetables Leafy Veg

Peas for processing Legumes

Rhubarb (natural) Leafy Veg

Salad onions Bulb Veg

Swede & turnips Root Veg

Brussel sprouts Leafy Veg

Cabbage Leafy Veg

Calabrese Leafy Veg

Cauliflowers Leafy Veg

Other brassicas Leafy Veg

Carrots Root Veg

Parsnips Root Veg

Bulb onions Bulb Veg

Leeks Leafy Veg

Courgettes & pumpkins Fruit Veg

Field grown herbs Leafy Veg


Table 6 Example of the model crop parameters employed in MACRO for the major crops

Variable WCereal SCereal WOSR SOSR Sugar Beet

Potatoes Field

Beans Legumes Maize

Radiation attenuation factor (ATTEN)

0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6

Transpiration adaptability factor (BETA)

0.1 0.2 0.1 0.1 0.2 0.5 0.5 0.5 0.2

Max. Interception capacity (mm) (CANCAP)

3 2 3 2 2 2 2 2 3

Leaf development factor, growth (CFORM)

2 2 2 1.7 1.7 1.7 1.7 2 1.7

Leaf development factor, senescence (DFORM)

0.2 0.3 0.2 0.2 1 0.3 0.3 0.3 0.3

Max. Crop height (m) (HMAX)

0.8 0.8 0.7 0.7 0.6 0.6 0.6 0.6 1.8

Green leaf area index at harvest (LAIHARV)

2 2 2 2 5 2 3 2 2

Maximum leaf area index (LAIMAX)

6 4 5 4 5 4 4 4 5

Leaf area index on ZDATEMIN (LAIMIN)

1 0.01 1 0.01 0.01 0.01 1 0.01 0.01

Root depth on ZDATEMIN (m) (ROOTINIT)

0.2 0.01 0.2 0.01 0.01 0.01 0.1 0.01 0.01

Root depth (m) (ROOTMAX)

0.75 0.75 1.1 1.1 0.8 0.5 0.8 0.8 1.1

% roots in top 25% of root depth (RPIN)

60 60 60 60 67 75 67 67 67

Min. stomatal resistance (s m

-1)

(RSMIN)

50 50 40 40 40 40 40 40 60

Ratio evaporation of intercepted water to transpiration (ZALP)

1 1 1 1 1 1 1 1 1.5

Crop height on specified day (m) (ZHMIN)

0.2 0.01 0.2 0.01 0.01 0.01 0.1 0.01 0.01

Max. leaf area development day (IDMAX)

181 179 166 196 196 196 180 181 222

Emergence day (IDSTART)

298 105 253 120 100 120 316 120 125

Harvest day (IHARV) 219 227 196 232 298 244 248 227 263

Intermediate crop development day (ZDATEMIN)

94 106 94 121 101 121 94 121 126

Proportion extractable micropore water exhausted before reduction in transpiration (FAWC)

0.8 0.65 0.8 0.8 0.65 0.5 0.5 0.5 0.65


Table 7 Example of the model crop parameters employed in PRZM for the major crops

Model Crop WCereal SCereal WOSR SOSR Sugar Beet

Potatoes Field

Beans Legumes Maize

Pan evap factor, (PFAC) 0.84 0.92 0.78 0.9 0.93 0.94 0.89 0.96 0.94

Canopy interception (CINTCP)

0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.3

Maximum rooting depth (AMXDR) (cm)

110 110 110 110 80 50 80 80 110

Maximum coverage (COVMAX)

90 90 90 90 90 80 80 85 90

Maximum height (HTMAX)

100 110 140 140 40 100 150 100 250

Emergence date (1) 2510 1504 1009 3004 1004 3004 1211 3004 0505

Intermediate crop date (ZDATEMIN) (2)

0404 1604 0104 0105 1104 0105 0104 0105 0605

50% crop cover date (3) 1705 3005 1504 3005 3005 1506 3004 3005 1506

Mature Date (4) 3006 2806 1506 1507 1507 1507 2906 3006 1008

Harvest Date (5) 0708 1508 1507 2008 2510 0109 0509 1508 2009

Fallow Date (6) 0109 0109 0108 0109 0111 1509 1509 0109 0110

USLEC (1) 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6

USLEC (2) 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

USLEC (3) 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3

USLEC (4) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

USLEC (5) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

USLEC (6) 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Irrigation

The crops that are typically irrigated with overhead techniques, for example sprinklers or rain guns, in each of the

6 climate zones, were defined using expert opinion and survey data (Weatherhead, 2007) and are summarised in

Table 8. Given that trickle irrigation is designed to improve application efficiency and, ergo, unlikely to impact on

drain flow and surface runoff, it was not included in this approach. For each crop in each climate zone the

irrigation requirements were defined using IRRIGUIDE v4.4.2 (Silgram, 2005) for 5 soil classes, namely a sandy

loam, clay loam, silty clay, clay and silt loam. Crop parameterisations and development dates were taken from

the MACRO and PRZM parameterisations. Critical soil moisture deficits, which would trigger irrigation and the

amount applied (see Table 9), were taken from the literature (Bailey, 1990; Birkenshaw and Bailey, 2003; Perry

and Atwood, 2003; Buckley et al., 2005; Tiffin et al., 2007). IRRIGUIDE uses Met Office HAGRO data for its

simulations, but this is not available for pre-1986 time periods. The MORECS weather data used in the MACRO

and PRZM simulations was adapted to run IRRIGUIDE for the 1961 to 1990 period. In order to calculate the dew

point temperature (Td), saturated vapour pressure (Es) was calculated from mean daily temperature (T) using the

August–Roche–Magnus approximation (see equation 1 - Allen et al., 1998), which along with the observed

vapour pressure, allowed calculation of the relative humidity (RH), which in turn was used to calculate the dew

point temperature (Td) (see equation 2 - Lawrence, 2005). Soil temperature at 30 cm was calculated using

MACRO v4.4.2 and the MORECS weather data. Where this temperature was less than 0°C the state of ground

was set to frozen, otherwise it was set to the default of moist.

3.237

27.17exp*6108.0

T

TEs Equation 1


T

TRH

T

TRH

Td

7.237

*27.17

100ln27.17

7.237

*27.17

100ln*7.237

Equation 2

Table 8 Summary of the crops that are typically irrigated with overhead irrigation (e.g. sprinklers or rain guns) in each climate zone

Crop SW SE CW CE NW NE

Potatoes Y Y Y Y Y Y

Maize N Y N N N N

Sugar beet N N N Y N N

Apples1 N N N Y N N

Blackcurrants2 N N N Y N N

Strawberries3 N N N Y N N

Carrots4

Y Y Y Y Y Y

Onions5 Y Y Y Y Y Y

Lettuce6 Y Y Y Y Y Y

Courgettes7 Y Y Y Y Y Y

Representative of the following crops: (1) top fruit, (2) cane fruit, (3) soft fruit, (4) root veg, (5) bulb veg, (6) leafy veg and (7) fruit veg Climate Zones: SW – South West, SE – South East, CW – Central West, CE – Central East, NW – North West, NE – North East

Table 9 Summary of the soil moisture deficits triggering irrigation and irrigation amounts for each crop irrigated using overhead irrigation

Crop Irrigation Factor IRRIGUIDE Soil Class

sandy loam clay loam silty clay clay silt loam

Main Crop Potato

Irrigation SMD (mm) 28 30 28 28 38

Irrigation Amount (mm) 21 23 21 21 29

Sugar beet Irrigation SMD (mm) 50 100 100 100 100


Maize Irrigation SMD (mm) 50 50 50 50 75


Apples1



Blackcurrants2



Strawberries3



Carrots4



Onions5



Lettuce6



Courgettes7


Irrigation Amount (mm) 20 20 20 20 30 Representative of these crops: (1) top fruit, (2) cane fruit, (3) soft fruit, (4) root veg, (5) bulb veg, (6) leafy veg and (7) fruit veg


Substances

The combinations of substance properties (KOC, DT50 and 1/n) were selected following an analysis of the

FOOTPRINT Pesticide Properties Database (see Table 10) which contains records for all substances registered

in the European Union (http://sitem.herts.ac.uk/aeru/footprint/en/index.htm) using ranges informed by literature

(Holman et al., 2004; Stenemo et al., 2007) and expert knowledge. A Q10 of 2.58, which corresponds to an

activation energy of 65.4 KJ/mol, was used. Owing to the large number of possible combinations, a pragmatic

selection of a range of KOC and DT50 combinations was made resulting in 36 combinations being selected (see

Table 10). The Henry’s Law Constant was set to zero, which assumes no volatilisation, and a 1/n of 1 was used,

which assumes linear sorption. A standardised 1 kg/ha application was made directly to the soil surface, ignoring

crop interception. Monthly application windows appropriate to each crop were used with the date of application

being defined by a standalone version of the FOCUS PAT applicator (FOCUS, 2001).

Table 10 Summary of the number of substances included in the FOOTPRINT pesticide properties database broken down by KOC, DT50 and 1/n. Light grey cells are combinations selected for use in the pesticide fate models

KOC$

DT50 10 30 100 300 1000 >1000† Total 1/n

$ Count

1 2 3 6 6 14 13 44 0.8 26

5 4 9 18 7 12 34 84 0.9 77

10 3 11 19 13 12 21 79 1 114

50 12 16 34 50 45 115 272

100 3 5 12 16 27 41 104

>100‡ 4 11 13 20 48 96

Total 28 44 100 105 130 272 679 217 † A value of 15000 was used to represent this group ‡ A value of 350 was used to represent this group $ For those substances that had values in the database

http://sitem.herts.ac.uk/aeru/footprint/en/index.htm


Edge of Field Environmental Fate Modelling

This section describes the development and parameterisation of an edge of field water body pesticide fate meta-model that is compatible with the rapid screening approach being developed, yet is representative of the outputs that would be generated by the TOXSWA model (Adriaanse, 1996), which is used in the FOCUSsw assessments.

Meta-Model Description An edge of field meta-model for predicting environmental concentrations of pesticides in edge of field surface water bodies was developed based on the Step 3 approach of the STEPS1-2-3-4 model of Klein (2007a, 2007b), which has undergone some validation. This 1-D model approach is much simpler than that modelled by the 2-D TOXSWA model yet retains the linkages with the FOCUSsw scenario inputs and the edge of field model parameterisations (e.g. with regard to the upstream catchment contributions). This simplified system comprises a water layer of fixed depth underlain by an upper and lower 5 cm thick sediment layer. Hourly fluxes of water and dissolved pesticide from spray drift, surface runoff and drain flow are added to the water layer (a), degraded (b), subject to outflow from the system (c) and allowed to interact with the upper sediment layer (d) as described in the equations below:

XswsedtInflowVsw

Vsw

CwaterDTtMASSswtMASSaqtMASSsw

)(*

50

)2ln(exp*)1()()(

(a) (b) (c) (d) Where: MASSsw(t) The mass of PPP in the surface water body at time t or t-1 MASSaq(t) The mass of aqueous phase PPP added by spray drift, surface runoff and drain flow DT50Cwater PPP half life in water corrected for temperature using Q10-approach Vsw Volume of the surface water body (set to 100 m length, 1 m width, depth of 0.3 m for D2 and 0.41

for R1 scenarios) Inflow Flow into the surface water body from baseflow, surface runoff and drain flow according to the

FOCUSsw scenario definition and approach Xswsed Mass transfer between the water and the upper sediment layer

DEPsw

tMASSsw

OCKoc

BDsedDEPsed

tMASSsed

LENdiff

CdiffXswsed

)(

*

*

)(

*24/

Where: Cdiff PPP diffusion coefficient in water (set to 4.3*10

-5 m

2/d)

LENdiff Diffusion path length (set to 0.003 m) MASSsed Mass of the PPP in the sediment at time t DEPsed Depth of the upper sediment layer (set to 5 cm) BDsed Bulk density of the sediment (set to 0.8 kg/L) Koc Organic carbon sorption coefficient (L/kg) OC Organic carbon content of the sediment (set to 0.05) MASSsw Mass of the PPP in the surface water body at time t DEPsw Depth of the water body (set to 0.3 m for D2 and 0.41 for R1 scenarios) In addition, hourly fluxes of PPP associated with sediment from surface runoff are added to the water layer. These additions are subject to outflow from the system (a), added directly to the upper sediment layer, degraded (b) and allowed to interact with the lower sediment layer (c) as described in the equations below:

XsedsedlCsedDT

tMASSsedtInflowVsw

VswtMASSerodetMASSsed

50

)2ln(exp*)1(

)(*)()(

(a) (b) (c) Where: MASSsed(t) The mass of PPP in the upper sediment layer of the surface water body at time t or t-1 MASSerode(t) The mass of particulate phase PPP added by surface runoff


DT50Csed PPP half life in sediment corrected for temperature using Q10-approach Xsedsedl Mass transfer between the upper sediment layer and the lower sediment layer

OCKoc

BDsedDEPsed

tMASSsed

BDsedDEPsedl

tMASSsedl

LENdiffsed

CdiffXsedsedl

*

*

)1(

*

)1(

*24/

Where: Cdiff PPP diffusion coefficient in water (set to 4.3*10

-5 m

2/d)

LENdiffsed Diffusion path length (set to 0.003 m) MASSsedl Mass of the PPP in the lower sediment layer at time t MASSsed Mass of the PPP in the upper sediment layer at time t DEPsedl Depth of the lower sediment layer (set to 5 cm) DEPsed Depth of the upper sediment layer (set to 5 cm) BDsed Bulk density of the sediment (kg/L) Koc Organic carbon sorption coefficient (L/kg) OC Organic carbon content of the sediment (fraction)

Meta-Model Evaluation The edge of field water body meta-model was validated against standard FOCUSsw runs using test substances 1-sw through 7-sw for scenarios D2 ditch and R1 stream. The maximum predicted environmental concentrations of the 14 simulations conducted using the same .m2t and .p2t files are illustrated in Figure 5 where it will be noted that the correlation between the meta-model and the TOXSWA model (Adriaanse, 1996) results is high (R

2 =

0.9993). The relationship is very close to the one-to-one line (slope of 0.9577) with a small underestimation of the TOXSWA concentrations, primarily on the higher concentrations (>50 μg/L). Given their relevance to the UK (Price et al., 2006; Price et al., 2007) and the performance of the meta-model, the D2 and R1 scenarios were chosen as the basis for ditch and stream edge of field risk assessments, respectively.

y = 0.9577x

R2 = 0.99930

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120 140 160

Toxswa PEC (microg/L)

Meta

Mo

del

PE

C (

mic

rog

/L)

Figure 5 Comparison of the maximum predicted environmental concentrations modelled using TOXSWA and the edge of field meta-model. The one-to-one relationship line is illustrated in red


Catchment Scale Stream Flow Meta-Modelling This section describes the development and parameterisation of a catchment scale daily stream flow volume meta-model that would predict flows available for dilution of compounds mobilised and delivered to surface water in spray drift, surface runoff and drainflow events.

Meta-Model Description This meta-model was developed using streamflow data from the National River Flow Archive (NRFA) for 315 catchments in England and Wales (see Figure 6) where these flows are reported to be natural or have been naturalised and are longer than 10 years in record length, accounting for record completeness. Each catchment was assigned to the modelling climate region in which it is located, the assumption being that the 50 year (1961-2010) representative weather file used in each of these regions adequately represents the weather that drove the hydrology in each catchment in these regions. Synchronous periods of record of weather and stream flow were extracted from these datasets for each catchment. Stream flows for Spray Drift Events The stream flow on the date of application was extracted for monthly application dates defined using a stand alone version of the FOCUS PAT applicator. It was not practical to model 50 years of runoff and drain flow in each catchment for every crop-pedo combination used in the fate meta-modelling to define the date of the largest runoff and drainflow event such that corresponding flows could be extracted. As such, a simple set of rules needed to be established to extract appropriate flows from the synchronous rainfall and stream flow records. Steam flows for Runoff Events Analysis of a suite of PRZM runs for a winter cereal on a range of soils indicated that the largest monthly pesticide flux is typically highly correlated with the pesticide flux associated with the largest monthly runoff event (see Figure 7a). As such, catchment stream flows associated with the peak runoff event in each month were extracted from the synchronous records and the median value calculated over the period of catchment record. Stream flows for Drain Flow Events Analysis of a suite of MACRO runs for a winter cereal on a denchworth soil indicated that peak fluxes of pesticide were associated with peak fluxes of drain flow (see Figure 7b). As such, the 50 years of modelled drain flow for this single crop-soil combination in each region was added to the synchronous record. As such, catchment stream flows associated with the peak drain flow event in each month were extracted from the synchronous records and the median value calculated over the period of catchment record. Flows from each “local” catchment were totalled to calculate the total flow for downstream “total” catchments fed from upstream local catchments.

Figure 6 Illustration of the catchments, selected in green, and the associated climate

regions used to train the catchment streamflow meta-model


(a)

R2 = 0.999

0

0.01

0.02

0.03

0.04

0.05

0.06

0 0.01 0.02 0.03 0.04 0.05 0.06

Runoff Flux in Peak Event (mg)

Maxim

um

Ru

no

ff F

lux (

mg

)

(b)

R2 = 0.9992

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Daily Drainflow Flux (mm)

Dail

y P

PP

Flu

x (

mg

)

Figure 7 Example plots of the relationship between (a) Runoff - the largest monthly pesticide flux and the pesticide flux associated with the largest monthly runoff event (b) Drain flow – the daily pesticide and water fluxes.

Catchment Descriptors Catchment descriptors were required for (1) the interpolation of Q90 and Qmean values to all catchments and (2) the development of the streamflow meta-model (see section 0). The descriptors considered, along with a description of what they comprise and the data from which they stem, is provided in Table 11. These are largely the same as those used within the Revised Flood Estimation approach (Institute of Hydrology, 1999; Kjeldsen et al., 2005) albeit they are not identical, having been developed for this assessment.


Table 11: Summary of the catchment descriptors considered in the development of the statistical meta-models

Descriptor Description

AREA Catchment area (km2)

BFIHOST Catchment weighted Base Flow Index (BFI) using the HOST classification system and the NATMAP1000 soils dataset

DPLBAR Mean of distances between each grid cell on OS panorama DEM (50 m resolution) and the catchment outlet (km).

SLPBAR Calculated as the mean slope of the catchment, using OS panorama DEM (50 m resolution)

LDP Longest drainage path (km), defined by recording the greatest distance from a catchment node to the defined outlet.

FCDAYS Number of days dominant soils at field capacity (FC) defined using median dates of return to and from FC for the period 1941-1970 (Smith and Trafford, 1976)

SAAR Average annual rainfall in the standard period (1961-90) (mm).

SPRHOST Catchment weighted standard percentage runoff (SPR) using the HOST classification system and the NATMAP1000 soils dataset

URBEXT Extent of urban and suburban land cover in 1995 using ITE landcover dataset (fraction).

DD Drainage density defined using the EA’s detailed river network

AGEXTENT Extent of agricultural land cover in 1995 using ITE landcover dataset (fraction).

GRASSEXTENT Extent of managed grassland land cover in 1995 using ITE landcover dataset (fraction).

ROUGHEXTENT Extent of rough grassland land cover in 1995 using ITE landcover dataset (fraction).

WOODEXTENT Extent of woodland land cover in 1995 using ITE landcover dataset (fraction).

PRECIP1to12 Average monthly precipitation in the standard period (1961-90) (mm)

RAINDAYS1to12 Average number of raindays in each month in the standard period (1961-90) (days)

HERPSYCHIC Hydrologically effective rainfall predicted using the PSYCHIC model (mm)

LAKEAREA Area of lakes using the EA’s lakes dataset (km2)

Qmean/Q95 Mean flow or flow exceeded 95% of the time, extracted from the National River Flow Archive (m

3/s)

Monthly Stream Flow Meta-Model Development and Evaluation The median monthly stream flows associated with drift/runoff/drain flow events were used along with the catchment descriptors in the construction of step-wise linear regression models. The step-wise linear regression models were developed in R with the most parsimonious model being selected through consideration of the Akaike Information Criterion. The models developed were typically robust (see Table 12) although correlations with drain flow in the summer were low, which is a function of minimal or no drainflow. The July drainflow model was discarded and replaced with the next best model (that for August). Given these models predict naturalised flows in catchments, catchments where there are pressures on low flow periods may have a lower flow associated with the spray drift flows and as such have a lower dilution potential. In order to account for these pressures, an additional flow reduction was implemented using catchment specific assessments undertaken as part of the WFD river basin characterisations for river abstractions and flow regulation (Environment Agency, 2009). Catchments classed as medium sensitivity have a 20% sensitivity threshold to alteration of flows between the Q95 (the flow exceed 95% of the time) and the Q70, the flow range believed to approximate the flow associated with spray drift during the summer months. Catchments at risk exceed this threshold by more than 50% (50% used in this study) while those probably at risk exceed this threshold by less than 50% (25% used in this study; See Figure 8) and those possibly at risk are licensed to exceed this threshold but don’t typically exceed the threshold (10% used in this study). These reduction factors were applied to the summer months (June, July and August) where low flows are typical and abstractions are likely to affect flows (Environment Agency, 2009).


Table 12: Linear regression model adjusted R2 values.

Month Adjusted R Squared

Drift Runoff Drainflow

Jan 0.759 0.893 0.971

Feb 0.764 0.874 0.812

Mar 0.880 0.939 0.928

Apr 0.895 0.940 0.972

May 0.925 0.927 0.901

Jun 0.900 0.841 0.804

Jul 0.934 0.858 0.247†

Aug 0.950 0.757 0.601†

Sep 0.872 0.742 0.680

Oct 0.773 0.726 0.820

Nov 0.798 0.785 0.905

Dec 0.836 0.885 0.888 † July flow model replaced with August flow model

Figure 8 Illustration of the WFD catchment risk categorisation for flow regulation and water abstraction (Environment Agency, 2009)


Minor Crops Database In order to spatially distribute the risk assessment framework within the agricultural landscape it was necessary to compile catchment summaries of the major and minor crops (Table 13 and Table 14, respectively) that were to be included in the minor uses screening tool. These data were extracted from the ADAS 1 km resolution land use database using census and industry intelligence for the 2009 harvest year. Some crops required a refinement of this database, for example, the top fruit category was split into four sub-categories using cropping statistics and industry intelligence gathered by ADAS horticultural consultants. This database was expanded to include some minor crops, for example Hops/Nuts were sourced from the Rural Payments Agency while Christmas Trees were sourced from a range of data sources. It should be noted that the screening tool has been built using the highest possible resolution of this data and as such contains disclosive census and industry intelligence data. This precludes distribution of the tool beyond CRD and other government agencies that would have access to the appropriate data licenses.

Table 13: Major crops, the soil groups on which they are grown and the data source used to compile the data.

Major Crop Name Soil group Data Source and Method

Winter wheat Any ADAS Land Use 2009 Database: A1*Regional Factor

Spring wheat Any ADAS Land Use 2009 Database: A1*Regional Factor

Winter barley Any ADAS Land Use 2009 Database: A2

Spring barley Any ADAS Land Use 2009 Database: A3

Oats Any ADAS Land Use 2009 Database: A4

Mixed grain Any ADAS Land Use 2009 Database: A5

Potatoes Any ADAS Land Use 2009 Database: A10+A11

Sugar beet Any ADAS Land Use 2009 Database: A12

Winter oilseed rape Any ADAS Land Use 2009 Database: A24

Spring oilseed rape Any ADAS Land Use 2009 Database: A25

Field beans Any ADAS Land Use 2009 Database: A21

Peas for harvesting dry Any ADAS Land Use 2009 Database: A22

Maize Any ADAS Land Use 2009 Database: A23

Root fodder crops Any ADAS Land Use 2009 Database: A19

Other crops for stock feeding Any ADAS Land Use 2009 Database: A18

Temporary grassland Any ADAS Land Use 2009 Database: G1

Permanent grassland Any ADAS Land Use 2009 Database: G2

Rough grassland Any ADAS Land Use 2009 Database: ROUGH

The outdoor crop areas compiled in each statistical region needed to be distributed onto the arable land within these regions. This combined a dasymetric-pycnophylactic method for disaggregating this data spatially (Comber et al., 2007) considered land cover data, for example orchards are explicitly identified, and dominant soil type. Permanent crops, like top fruit, small fruit and Christmas Trees, were distributed equally onto soils in the parishes in which they occur. Crops grown as part of a rotation, for instance the vegetable and cereal crops, were assigned as either having a soil preference or not. Most vegetables are grown on light, flat agricultural land to ensure that the land can be accessed early and late in the growing season and were the main crops assigned a soil preference. When distributing crops onto arable land within a parish, the vegetables were done first and preferably placed onto their highest preference soil. This process was iterative and accepted that vegetables are grown as part of a rotation and as such other crops, like cereals, needed to also occupy some of these lighter soils. All other crops were considered to be part of a rotation and as such allowed to compete with each other for all soil types.


Table 14: Minor crops, the soil groups on which they are grown and the data source used to compile the data

Minor Crop Name Soil Group Data Source and Method

Rye Any ADAS Land Use 2009 Database: A6

Triticale Any ADAS Land Use 2009 Database: A7

Other Oil Seeds Any ADAS Land Use 2009 Database: A27

Apples Any ADAS Land Use 2009 Database: C1 * Regional Factor

Pears Any ADAS Land Use 2009 Database: C1 * Regional Factor

Plums Any ADAS Land Use 2009 Database: C1 * Regional Factor

Other Topfruit Any ADAS Land Use 2009 Database: C1 * Regional Factor

Strawberries Any ADAS Land Use 2009 Database: C5

Raspberries Any ADAS Land Use 2009 Database: C6

Blackcurrants Any ADAS Land Use 2009 Database: C7

Wine grapes Any FSA Vineyards Register

Other small fruit Any ADAS Land Use 2009 Database: C11

Hops Any Rural Payments Agency Database 2009

Nuts Any Rural Payments Agency Database 2009

Christmas Trees Any ADAS Land Use 2009 Database: D6; Forestry Commission Inventory of woodlands; Christmas Tree Growers Association Survey Data 2009

Bulbs and Flowers Grown in the Open

Any ADAS Land Use 2009 Database:: D99 Wales, D99-D6 England

Asparagus L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Beans L > M > H ADAS Land Use 2009 Database: B14

Beetroot L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Field Celery L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Lettuce & leafy veg L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Peas for processing L > M > H ADAS Land Use 2009 Database: B5

Rhubarb (Natural) L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Salad Onions L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Swede & Turnips L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Brussel Sprouts L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Cabbage L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Calabrese L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Cauliflowers L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Other Brassicas L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Carrots L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Parsnips L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Bulb Onions L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Leeks L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Field Grown Herbs L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Cucurbits L > M > H ADAS Land Use 2009 Database: B21 combined with industry intelligence

Golf courses Any AA Great Britain point locations combined with typical golf course sizes

Urban Green Space Any CORINE 2000 land cover dataset

L > M > H: Soil preference - Light, then medium and finally heavier soils

In addition to the agricultural land uses, some amenity type land uses were also included. Golf courses were included through consideration of their location and the fact that they are remarkably regular in size (sensu Fox et al., 2008): Tees: 1.1ha; Greens: 1.3ha; Fairways: 14ha, Rough: 28ha; Total: 44.4ha (http://www.epa.gov/oppefed1/models/water/golf_course_adjustment_factors.htm; British and International Greenkeepers Association – Pers. Comm.) The categories ‘Green Urban’ and ‘Sports and Leisure’ land were extracted from the CORINE land cover dataset (categories 1.4.1 and 1.4.2). In both cases the soils on which these land uses occur were summarised. Catchment summaries of the area of each major and minor land use grown on each of the 19 soil classes were extracted.

http://www.epa.gov/oppefed1/models/water/golf_course_adjustment_factors.htm


Assessing Minor Usage on Major Arable Crops In order to assess the potential minor use of plant protection products on major arable crops an assessment of minor diseases, pests and weeds was undertaken using the best available data in conjunction with expert opinion from leading national specialists within ADAS.

Minor diseases of major crops Fungicides are applied to crops across England and Wales routinely for a range of diseases that have the potential to reduce yields, regardless of the perceived risk from a pathogen in a particular region. For example, Septoria tritici is routinely treated in winter wheat by specifically targeted fungicide applications, as is Sclerotinia sclerotiorum (stem rot) in oilseed rape and Phytophthora infestans (late blight) on potato. Fungicides often have broad-spectrum activity and, therefore, non-target or minor diseases are controlled. Many of these routinely applied fungicides will also control diseases considered to have minor importance, including Ascochyta sp. on wheat and Mycosphaerella spp. on oilseed rape, therefore, additional fungicides to control these diseases, often considered minor, are not required. It can be difficult to pinpoint disease risk by region as the presence of disease in a particular area can vary depending on the climatic and cultural conditions experienced from season to season as well as on farm history of specific diseases. The objective of this review was to identify whether the scale of use of fungicides has the potential to vary from region to region to control minor diseases and the potential size of the areas affected.

Methodology and Assumptions A range of sources were used for this review (e.g. HGCA encyclopaedias of oilseed rape and cereal diseases; HGCA Planting survey results 2010; www.cropmonitor.co.uk; www.ukagriculture.com; Gladders, P. – Pers. Comm.; Garthwaite et al., 2005; Garthwaite et al., 2008). All diseases of wheat, barley, oats, field beans, peas, potatoes and sugar beet were identified and categorised into major and minor importance. A major disease was considered to be one that is treated routinely and these were excluded. Diseases controlled via the compulsory certification scheme (e.g. ergot on wheat) were also excluded. Most crop pathogens are found across the whole of England and Wales, with some areas having a higher incidence than others due to climatic, cultural or other cropping factors. The potential areas that could be affected were taken from the arable pesticide disease survey from 2008, which gives specific regional areas of crops by Government Office Regions (GORs). This gives precise regional data from 2008 for all crops except maize (taken from the 2010 Pesticide survey) and sugar beet (taken from http://www.ukagriculture.com/crops/sugar_beet_farming.cfm). It was difficult to quantify the proportion of crops affected each year by these diseases as they either occur sporadically or can be controlled by routine fungicides applied to control major diseases. Expert opinion was used to determine the actual areas within these regions at risk as the use of resistant cultivars or mitigation using cultural methods will also affect disease occurrence (see Table 15).

Table 15: Diseases of minor extent in England and Wales

Major crop Minor problem Potential crop

area affected (%) Model Regions

Affected

Barley Halo spot 2 SW

Maize Maize eyespot 7 CW

OSR Clubroot 9 CW, CE

OSR Ring spot 6 SE, CE

OSR White leaf spot 17 SE, SW, CW

Potatoes Alternaria sp. 19 CE, NE

Potatoes Pink rot 16 CW

Barley Barney patch 3 CE

Wheat Flag smut <1 CE, SE

Wheat and Barley Leaf and glume blotch 8 SW

http://www.ukagriculture.com/crops/sugar_beet_farming.cfm


Results Varietal resistance, cultural control and the treatment of major crop diseases all have a role in reducing the occurrence of minor diseases in crops in England and Wales. It can be concluded from this study that very few arable crops in England and Wales require additional fungicide treatments for minor diseases other than those applied for major diseases. Minor diseases such as halo spot on barley, ring spot and white leaf spot on oilseed rape can all be controlled to a degree by application of fungicides applied routinely for major diseases of these crops. Tan spot on wheat is increasing in incidence, with 19.5% of crops surveyed for CropMonitor (www.cropmonitor.co.uk) found to have symptoms. However, it can be controlled by fungicides applied for Septoria tritici. Similarly, yellow and brown rust on wheat are generally associated with Eastern regions, however, T0 and T3 fungicides will also be applied to susceptible varieties in other regions. Ploughing crops stubbles and long rotations are alternatives to fungicide use in many instances and could be used to control maize eyespot and clubroot of oilseed rape. Disease epidemics are very variable from year to year and the development of damaging epidemics is difficult to predict. Broad-spectrum fungicides and fungicide mixtures are therefore widely used so that there is satisfactory control of disease complexes. This includes many minor diseases and explains the lack of targeting of disease specific products for these pathogens.

Minor pests of major crops Insect pests are typically highly mobile and as a result have the potential to occur in most parts of the country. As such, the distribution of insect pests is typically most closely associated with the distribution of host plants. Non-resident pests such as Silver-Y typically arrive in the East of the country each year and the first generation in the UK is, therefore, usually localised in the East. However, this strong flier rapidly spreads across the rest of the country in subsequent generations.

Methodology and Assumptions Two key sources of information were used in the identification of minor pests and any geographic relationship in their distribution. The Pesticide Usage Survey Report 224 – Arable Crops in Great Britain 2008 was used to provide information on total crop areas, crop areas by region, insecticide treated area, key active ingredients applied and, where available, information on the pests targeted. The Pesticide Usage Survey Reports provide only a ‘broad brush’ summary of insecticide usage and, as may be expected, focus on just the key pests. In order to obtain information on minor pests, ADAS entomologists Mike Lole and Steve Ellis were consulted. Steve and Mike have, in total, more than 50 years of arable entomology experience and are involved with producing the ADAS Crop Action publication and have extensive experience in pest identification and evaluation clinics. For the purposes of this study a minor pest was considered to be a pest which is controlled on less than 10,000 ha nationally. As far as possible, incidental pest control was excluded. However, the data available on the intended target pest is poor and so a certain amount of judgement has been made in determining the area specifically treated for each pest. Similarly, most pests have the potential to be damaging every year and have been recorded as such, but in reality certain pests may not require control for several years (e.g. saddle gall midge). Each major crop was considered in turn and potential pests identified and confirmed through expert judgement.

Results The resulting table of major crops, pests and areas affected are summarised in Table 16. However, by its very nature incidents of minor pests are poorly recorded and so it has not been possible, with confidence, to confirm that all or even most minor pests have been identified. Changing farming/cropping practice and climate change is likely to result in some new pests while others may decline in importance.

http://www.cropmonitor.co.uk/


Table 16: Pests of minor extent in England and Wales

Major crop Minor problem Potential crop

area affected (%) Model Regions

Affected

Barley (spring & winter) Saddle gall midge <1 SE, CE, NE

Field beans Black bean aphid 2.8 SW, SE, CE, NE

Oats Cereal fly e.g. Frit <1 SE, CE, CW, NE

Oats Bean seed fly <1 SE, CE, CW, NE

Oats Stem nematode <1 SE, CE, CW, NE

OSR Mealy cabbage aphid 2 All

OSR Turnip sawfly 1 SW, SE, CE

OSR Rape winter stem weevil 1 NE

Peas dry Bean seed fly 2 SW, SE, CE, NE

Peas dry Field thrips 2 SW, SE, CE, NE

Peas dry Pea moth 2 SW, SE, CE, NE

Potato Two-spotted spider mite 1 All

Potato Chafer larvae 2 CE

Root fodder crops Turnip sawfly 4 NW, CW, SW

Sugar beet Noctuid caterpillars 3 CE

Wheat (spring & winter) Bibionid larvae <1 All

Wheat (spring & winter) Cereal leaf beetle <1 All

Wheat (spring & winter) Gout fly <1 All

Wheat (spring & winter) Saddle gall midge <1 CE, NE

Wheat (spring & winter) Cereal ground beetle <1 SW, SE, CE

Minor weeds of major crops A weed is a plant in the wrong place. When present in a crop, a weed can compete strongly for light and nutrients. It can often grow and reproduce aggressively and/or harbour and spread pests or pathogens which infect or degrade the quality of crops. Weeds have limited mobility, but can spread as seed or vegetative offcuts via animals, humans and machinery. They are also restricted due to their preference for specific soil types, altitude and temperature regimes. Non native weeds do appear near airports and sea ports where overseas goods are unloaded, or as garden escapees. Generally their spread is limited, but species such as Japanese knotweed, rhododendrons and Himalayan balm have become major problems in the UK since their introduction as garden species.

Methodology and Assumptions For the purposes of this project we have identified minor weeds in major crops. A minor weed definition for this project is one or more of the following:

Plant that is not specifically treated with current herbicides on a regular basis, but can cause economic damage in specific situations

Plant that appears in a small proportion (<10% of area) of specific crops e.g. thistles in grassland or brome in spring sown cereals

Plants that are present over a wide area in low populations that may not currently be economically damaging

Plants that are currently controlled incidentally by other chemicals, but could become a problem if management changed

Weed information was used from the following key sources and confirmed with ADAS weed specialists:

1. Herbicide product labels. The UK pesticide guide database (www.plantprotection.co.uk) was searched for herbicide target weeds. This method identified the major weeds of the major crops as they are itemised on the herbicide product label.

2. The encyclopaedia of arable weeds (www.hgca.com/awe) this classifies individual weeds as to their economic importance on a scale of 1-4 with 1 being the most injurious weeds to yield and quality and 4 being the least. Weeds within classifications 1-3 were selected.

http://www.plantprotection.co.uk/

http://www.hgca.com/awe


3. The National Biodiversity Network (www.nbn.org.uk/) which provides gridmaps for the distribution of each weed species.

4. Pest, Disease and Weed Incidence Information annual reports (www.pesticides.gov.uk/farmers_growers.asp?id=861) This information provides a guide to the significance of weeds in crops, and where particular control problems have been identified in a cropping year, (i.e. from autumn through to summer).

Results The resulting table of major crops, weeds and areas affected are summarised in Table 17. This analysis indicates that most weeds are fairly ubiquitous with the degree of infestation being controlled by local factors.

Summary and Conclusions In order to assess the potential minor use of plant protection products on major arable crops, an assessment of minor diseases, pests and weeds was undertaken using the best available data in conjunction with expert opinion from leading national specialists within ADAS. This analysis indicated:

Diseases – given the difficulty in predicting infection, treatment is typically prophylactic with minor diseases largely treated through incidental control;

Pests – minor pests are by their nature poorly documented and as such a complete assessment is difficult;

Weeds – most weeds are fairly ubiquitous with the degree of infestation being controlled by local factors In addition, there would probably be limited instances where a Notifier would seek to register a compound used extensively on a major crop for a minor use on another major crop. As such we recommend that default pest/disease/weed parameters not be included in the minor uses tool. Rather the facility to include regionally variable treatment proportions (as has already been included in the beta version) be used along with statistics supplied by the Notifier for the minor use on the other major crop.

http://www.nbn.org.uk/

http://www.pesticides.gov.uk/farmers_growers.asp?id=861


Table 17: Weeds of minor extent in England and Wales

Common name % Area affected

winter sown cereals

spring sown cereals

winter sown break crops

spring sown break crops

spring sown root crops

maize Model Regions

Affected

Loose silky bent 5 Minor Minor Minor Minor All †

Meadow brome 5 Minor Minor Minor Minor All †

Rye brome 2 Minor Minor Minor Minor Minor Minor All †

Common hemp-nettle 9 Minor Minor Minor All

Henbit dead-nettle 5 Minor Minor Minor Minor Minor Minor All

Scented mayweed 67 Minor Minor Minor All

Awned canary-grass 5 Minor Minor Minor Minor Minor Minor All

Wild radish 30 Minor Minor Minor Minor All

Shepherd's-needle 3 Minor Minor Minor Minor Minor Minor All †

Small nettle 5 Minor Minor Minor Minor Minor All

Fool's parsley 29 Minor Minor Minor Minor All

Black bent 7 Minor All

Creeping bent 7 Minor Minor Minor Minor Minor Minor All

Barren brome 13 Minor Minor All

Cow parsley 1 Minor Minor Minor Minor Minor Minor All

Parsley-piert 12 Minor Minor Minor Minor Minor Minor All

False oat Grass/onion couch 2 Minor Minor Minor All

Common orache 13 Minor Minor All

Soft brome 2 Minor Minor Minor Minor Minor Minor All

Shepherd's-purse 23 Minor Minor All

Fat hen 13 Minor Minor Minor All

Creeping thistle 4 Minor Minor Minor All

Common couch 21 Minor Minor Minor Minor Minor Minor All

Field horsetail 4 Minor Minor Minor Minor Minor Minor All

Black-bindweed 48 Minor All

Common fumitory 17 Minor Minor Minor Minor Minor All

Cut-leaved crane's-bill 11 Minor Minor Minor Minor Minor All

Dove's-foot crane's-bill 11 Minor Minor Minor Minor Minor All

Red dead-nettle 47 Minor All


Common name % Area affected

winter sown cereals

spring sown cereals

winter sown break crops

spring sown break crops

spring sown root crops

maize Model Regions

Affected

Nipplewort 20 Minor Minor Minor Minor Minor Minor All

Italian rye-grass 14 Minor All

Field forget-me-not 11 Minor Minor Minor All

Common poppy 18 Minor Minor All

Pale persicaria 8 Minor Minor Minor Minor Minor Minor All

Redshank 8 Minor Minor Minor All

Rough-stalked meadow-grass

7 Minor Minor Minor Minor Minor All

Knot-grass 16 Minor Minor Minor All

Bracken 4 Minor Minor All

Creeping buttercup 12 Minor Minor Minor Minor Minor Minor All

Broad-leaved dock 12 Minor Minor Minor Minor Minor Minor All

Charlock 36 Minor Minor Minor Minor All

Prickly sowthistle 35 Minor Minor Minor Minor Minor Minor All

smooth sowthistle 35 Minor Minor Minor Minor Minor All

Dandelions 12 Minor Minor Minor Minor Minor Minor All

Scentless mayweed 67 Minor Minor Minor All

Common nettle 12 Minor Minor Minor Minor Minor Minor All

Common field-speedwell 72 Minor All

Field pansy 45 Minor Minor Minor Minor Minor All

† While this weed occurs in all regions it would be considered minor in some regions and dominant in others


Screening Tool Calculation Methodology The work flow of operations/calculations within the screening tool is as follows:

1. The user defines the date of the first application to the major and minor crop along with the number of applications, the application interval, the application rate and the application type.

2. The amount of spray drift for each application is calculated. 3. The interception for each application is calculated. 4. The number of applications in each month is determined and the effective application for that month

calculated. 5. The runoff and drainflow losses are calculated using the field loss meta-model. 6. The edge of field concentrations are calculated using the edge of field meta-model. 7. The edge of field risk assessments are conducted using the user entered thresholds and safety factors

and the edge of field concentrations; the area of crop in each risk class is tabulated. 8. The catchment loads are determined using output from the edge of field meta-model and the area of each

crop-pedo combination in each catchment. 9. The catchment concentrations are determined using the catchment loads and the catchment streamflow

meta-model. 10. The catchment risk assessments are conducted using the user entered catchment thresholds and safety

factors and the catchment concentrations; the number of catchments in each risk class are tabulated. 11. The area and catchment results are cross tabulated with Environment Agency ecological status and may

be exported (the detailed catchment results may be exported to a spatial viewer).

Spray Drift Calculation Spray drift is calculated using the simplified form of the FOCUS equations (owing to the distances to water bodies being lower than the hinge distances), distances from crops to water bodies, and percentile based on number of applications, are summarised in Table 18 and Table 19 (after FOCUS, 2001):

][*)1(*)(

1

1

1

2

12

BBzz

Bzz

ADrift

where Drift = mean percent drift loading across a water body that extends from a distance of z1 to z2 from the

edge of the treated field A, B = FOCUS regression parameters z1 = distance from edge of treated field to closest edge of water body (m) z2 = distance from edge of treated field to farthest edge of water body (m)

Interception The date of each spray event is determined from the date of application, the number of applications and the application interval. The application type is also used to define whether interception is applied (aerial and air blast spray). Interception (see Table 20) is calculated using the FOCUS approach being proportional to the crop growth stage based on the emergence, intermediate, maximum LAI and harvest dates as well as the LAI at each growth stage and the maximum interception at maximum LAI (FOCUS, 2001). The interception values are interpolated between these critical growth stages.

Applications – number and effective in each month The number of applications in each month is determined from the date of application, the number of applications and the application interval. Where more than one application is to be made in the same month the effective application rate (PPPapplic) vulnerable to the peak runoff/drainflow event is calculated across all applications accounting for single first order degradation based on the time difference (Tdiff) between the applications:

PPPnextTDT

LNEXPsPPPpreviouPPPapplic diff )(*

50

2*


Runoff and drainflow losses The runoff and drainflow losses (PPPloss) to surface water are calculated using the field loss meta-model for the effective compound application rate (PPPapplic), compound KOC and DT50 values for the crop and month of interest (FFMMloss).

FFMMlossPPPapplicPPPloss *

Table 18: Summary of distance between the crop and water bodies used in the spray drift evaluation (FOCUS, 2001)

Model Crop Z1 Ditch Z2 Ditch Z1 Stream Z2 Stream

Winter Cereal 1 2 1.5 2.5

Spring Cereal 1 2 1.5 2.5

Potatoes 1.3 2.3 1.8 2.8

Sugar Beet 1.3 2.3 1.8 2.8

Winter OSR 1 2 1.5 2.5

Spring OSR 1 2 1.5 2.5

Field Beans 1.3 2.3 1.8 2.8

Legumes 1.3 2.3 1.8 2.8

Maize 1.3 2.3 1.8 2.8

Bare Fallow 1 2 1.5 2.5

Grass 1 2 1.5 2.5

Top Fruit 3.5 4.5 4 5

Cane Fruit 3.5 4.5 4 5

Small Fruit 1 2 1.5 2.5

Vines 3.5 4.5 4 5

Hops 3.5 4.5 4 5

Small Trees 3.5 4.5 4 5

Bulb Flowers 1 2 1.5 2.5

Bulb Veg 1 2 1.5 2.5

Leafy Veg 1 2 1.5 2.5

Leafy Veg Late 1 2 1.5 2.5

Root Veg 1 2 1.5 2.5

Fruit Veg 1 2 1.5 2.5

PS2238 – Final Report 35

Table 19: Summary of the FOCUS spray drift regression equation parameters

Crop grouping Number of applications Percentile A B C D Hinge distance (m)

Arable and Veg Crops <50cm 1 90 2.759 -0.978 -- -- --








Hops 1 90 58.25 -1.004 8655 -2.835 15.3

Hops 2 82 66.24 -1.2 5555 -2.823 15.3

Hops 3 77 60.4 -1.213 4061 -2.763 15.1

Hops 4 74 58.56 -1.217 3670 -2.762 14.6

Hops 5 72 59.55 -1.248 2861 -2.704 14.3

Hops 6 70 60.14 -1.27 2954 -2.727 14.5

Hops 7 69 59.77 -1.281 3192 -2.767 14.6

Hops 8 67 53.2 -1.247 3010 -2.755 14.6

Vines, Late Applications and Veg >50cm 1 90 44.77 -1.564 -- -- --








Vines, Early Applications 1 90 15.79 -1.608 -- -- --








Pome/Stone Fruit, Late Applications 1 90 60.4 -1.225 210.7 -1.76 10.3

Pome/Stone Fruit, Late Applications 2 82 42 -1.131 298.8 -1.946 11.1



Crop grouping Number of applications Percentile A B C D Hinge distance (m)

Pome/Stone Fruit, Late Applications 4 74 36.27 -1.162 202 -1.877 11

Pome/Stone Fruit, Late Applications 5 72 34.59 -1.153 197.1 -1.88 11




Pome/Stone Fruit, Early Applications 1 90 66.7 -0.752 3868 -2.418 11.4





Pome/Stone Fruit, Early Applications 6 70 58.83 -0.864 7066 -2.732 13




Table 20: Summary of the interception dates and percentages (after FOCUS, 2001)

Model Crop Dates Interception %

Emergence Intermediate Maximum Harvest Emergence Intermediate Maximum Harvest

Winter Cereal 25/10/2009 04/04/2010 30/06/2010 07/08/2010 0 15 90 30

Spring Cereal 15/04/2010 16/04/2010 28/06/2010 15/08/2010 0 0 90 45

Potatoes 30/04/2010 01/05/2010 15/07/2010 01/09/2010 0 0 80 40

Sugar Beet 10/04/2010 11/04/2010 15/07/2010 25/10/2010 0 0 90 90

Winter OSR 10/09/2009 01/04/2010 15/06/2010 15/07/2010 0 18 90 36

Spring OSR 30/04/2010 01/05/2010 15/07/2010 20/08/2010 0 0 90 45

Field Beans 12/11/2010 01/04/2010 29/06/2010 05/09/2010 0 0 80 60

Legumes 30/04/2010 01/05/2010 30/06/2010 15/08/2010 0 0 85 43

Maize 05/05/2010 06/05/2010 10/08/2010 20/09/2010 0 0 90 36

Top Fruit 01/04/2010 01/04/2010 01/07/2010 30/09/2010 50 50 80 50

Small Fruit 01/04/2010 01/04/2010 20/04/2010 15/07/2010 50 50 60 60

Cane Fruit 15/03/2010 16/03/2010 15/05/2010 30/07/2010 50 50 80 80

Vines 01/04/2010 02/04/2010 02/08/2010 15/10/2010 40 40 85 40

Hops 01/03/2010 02/03/2010 01/07/2010 01/09/2010 0 0 90 90

Bulb Flower 15/11/2010 01/01/2010 01/03/2010 31/03/2010 0 0 60 60

Bulb Veg 08/04/2010 09/04/2010 15/07/2010 15/09/2010 0 0 60 60

Root Veg 21/05/2010 22/05/2010 01/09/2010 01/10/2010 0 0 80 80

Leafy Veg 27/04/2010 28/04/2010 23/06/2010 30/06/2010 0 0 90 90

Leafy Veg Late 13/07/2010 14/07/2010 25/08/2010 30/08/2010 0 0 90 90

Fruit Veg 22/05/2010 23/05/2010 15/07/2010 07/08/2010 0 0 90 90

Grass/Alfalfa 01/01/2010 02/01/2010 04/01/2010 31/12/2010 90 90 90 90

Small Trees 01/01/2010 02/01/2010 04/01/2010 31/12/2010 80 80 80 80

Bare Fallow 01/01/2010 02/01/2010 04/01/2010 31/12/2010 0 0 0 0


Edge of Field Predicted Environmental Concentrations The edge of field (EOF) predicted environmental concentrations (PECs) are calculated for each crop, soil and climate combination using the drift, runoff and drainflow losses of PPP, with the days after application being used to define when the runoff and drainflow losses are added to the EOF ditch/stream system. The drift losses of PPP are added on the day of application. The EOF PECs are calculated using the EOF fate meta-model. The D2 ditch system is used to represent drained land while the R1 stream is used for all other soils where surface runoff is the primary loss pathway.

Edge of Field Risk Assessment The edge of field risk assessment is conducted by dividing the regulatory acceptable concentration (RAC in μg/L) entered by the user by the maximum ditch/stream predicted environmental concentration (in μg/L) to derive a toxicity exposure ratio (TER). This ratio is classed into two ranges: at risk (TER < 1) and not at risk (TER >= 1). The total hectares of land associated with each of the EOF risk classes are tabulated for each catchment.

Catchment Load and Predicted Environmental Concentration The maximum daily loss of PPP downstream from the EOF water bodies is extracted for each event type and used to determine the maximum catchment load. These losses are weighted by the area of crop on each soil in each catchment to determine the total load. They may also be weighted by a range of additional weights:

1. Product usage weight – defaulted to 1, this factor allows for market share of a compound to be considered with statistics being taken from the Pesticide Usage Survey for, for example, the most popular product;

2. Crop proportion weight – defaulted to 1, this factor allows for a fraction of a crop to be treated, for example just mint within the herbs category if such information were collected or known;

3. Connectivity weight – defaulted to 1, this factor allows for the investigation of delivery of surface runoff to surface water bodies based on field proximity statistics;

4. Drainage prevalence weight – defaulted to 1, this factor allows for the investigation of agricultural drain presence, as not all soils requiring drains are actually drained. This factor is derived from an as yet unpublished ADAS dataset derived using application of a statistical model of surviving agricultural drains (Anthony et al., 2012)

The catchment predicted environmental concentration is determined from the catchment load of PPP and the respective catchment flow.

Catchment Risk Assessment The catchment risk assessment is conducted by dividing the catchment ecotoxicological threshold entered by the user (e.g. drinking water standard of 0.1 μg/L or an environmental quality standard in μg/L) by the maximum catchment predicted environmental concentration, to derive a toxicity exposure ratio (TER). This ratio is classed into four ranges using the TER and an uncertainty factor (UF) (sensu Whelan et al., 2005) resulting in: at risk (TER * UF < 1), possibly at risk (TER < 1 and TER * UF > 1 ), possibly not at risk (TER > 1 and TER/UF < 1) and not at risk (TER/UF > 1). This classification system allows the range of PECs to be summarised into ranges that take account of the ecotoxicological threshold and the uncertainties associated with the modelling approaches. The total number of catchments associated with each of the four catchment risk classes is tabulated.

Tabulate and Export Results The edge of field results, the number of hectares of major/minor crop in each risk class for each exposure pathway, are exported for each catchment and also summarised for all of England and Wales onto summary sheets. The summary results are produced in a number of useful ways to aid interpretation (see Figure 9), including (a) cross tabulated against the Environment Agencies Water Framework Directive (WFD) ecological status to inform where the minor use may be adding to existing WFD issues or creating new ones where there are currently none; (b) summarised regionally as a quick means of assessing spatial bias; (c) major crop areas cross tabulated against minor crop/usage areas as a means of indicating the degree to which the minor usage overlaps with the existing major usage and the extent of any new ‘at risk’ areas; (d) graphics indicating (i) the typical proportion of a catchment that ‘at risk’ areas represent as a means of assessing the importance of local hotspots and pollution and (ii) the difference in ‘at risk’ areas produced by the major crop and minor crop/usage in each catchment as a means of assessing if the risk within any one catchment is more likely to arise from the minor crop/usage. Similarly the catchment results, the number of catchments in each risk class, are exported for each catchment and summarised in useful ways to aid interpretation. The catchment results may also be exported to a


shapefile of the catchment boundaries which may then be viewed in a bespoke ADAS spatial data viewer (see Figure 10) in order to visually assess the potential for hot spots.

(a)

(b)

(c)

(d)

(e)

Figure 9 Illustration of the output displayed in the MS Excel summary sheet for each exposure pathway, including (a) area of major crop in each risk class cross tabulated by Environment Agency Water Framework Directive ecological status; (b) area of major crop in each risk class summarised by climate region; (c) graph of the numbers of catchments containing percentage categories of areas at risk; (d) cross tabulation of the areas at risk shared by major and minor crops and (e) graph of the numbers of catchments in each crop difference category (difference between the major and minor crops)


Figure 10 Illustration of the catchment level risk assessment results for the spray drift and runoff exposure pathways for a major crop as displayed in the Minor Uses Screening Tool spatial viewer

Screening Tool Output Evaluation The outputs from the Minor Uses Screening Tool (MUSTool) were evaluated using a suite of scenarios (see Table 21). The edge of field results were compared with standard FOCUS Surface Water Scenario (FOCUSSWS) runs and published values (e.g. the catchment results were compared with Environment Agency catchment water quality monitoring data). The four scenarios selected were based on 2 factors (1) the compound was selected where a number of detections have been recorded in the Environment Agency’s water quality database and (2) the compound is typically used on a limited range of crops such that the MUSTool catchment output reflects the majority of losses to surface water. Compound properties were extracted from EU Draft Assessment Reports, while typical application rates (as opposed to the label rate), timings and extents were extracted from pesticide usage statistics (see http://fera.defra.gov.uk/scienceResearch/science/lus/pesticideUsageFullReports.cfm) and reports (see http://www.fera.defra.gov.uk/scienceResearch/science/lus/pesticideUsage.cfm). EA monitoring data for the years 2003 through 2008 were supplied by Defra for this project and all locations where the compounds were detected above their limit of detection were extracted for further statistical analysis.

Table 21: Summary of the scenario characteristics used in the evaluation of the Minor Uses Screening Tool

Scenario Name 1 – IPU on

Winter Cereals

2 – MCPP on Spring and

Winter Cereals 3 – Atrazine on Maize

4 – Propyzamide on Winter OSR

Compound Isoproturon Mecoprop(-P) Atrazine Propyzamide

Crop Winter Cereals Spring and

Winter Cereals Maize

Winter Oilseed Rape

Koc (mL/g) 36 20 100 840

DT50 Soil (days) 33 @ 20 ºC 8 @ 20 ºC 60 @ 20 ºC 54 @ 20 ºC

DT50 Water (days) 61 @ 20 ºC 39 @ 20 ºC 80 @ 20 ºC 90 @ 20 ºC

DT50 Sediment (days) 1000 @ 20 ºC 1000 @ 20 ºC 1000 @ 20 ºC 1000 @ 20 ºC

1/n† 1 1 1 1

Uptake Factor† 0 0 0 0

Application Rate (g/ha) 1200 600 1100 750

Application Date Oct May/Oct May Nov

Crop Treated (%) 71% 23% 79% 31% † Standardised to reflect the MUSTool defaults

http://fera.defra.gov.uk/scienceResearch/science/lus/pesticideUsageFullReports.cfm

http://www.fera.defra.gov.uk/scienceResearch/science/lus/pesticideUsage.cfm


Edge of Field Results The results of the edge of field risk assessments for the four scenarios (see Table 21) are presented in this section. It should be borne in mind that the statistics presented are influenced by the number of data points included in the analysis and that for the FOCUSSW scenarios these may be limited in number (maximum of 6 for the drainage scenarios but typically less depending on the crop) and these are limited to “realistic worst” cases. The statistics for the MUSTool include all cases found in agricultural landscapes not just realistic worst cases. In addition, the published edge of field values are for information purposes only and indicate the kinds of ranges that the four compounds have been detected in the environment in drainflow and surface runoff, they are not directly comparable to the modelled results in that the application rate, timing and crop are not taken into account nor the location of sampling (e.g. in-field as opposed to edge of field for runoff values). Scenario 1 – Isoproturon, a mobile and moderately persistent herbicide, applied to winter and spring cereals. The drainflow results (see Figure 11) indicate that the MUSTool under predicts the maximum PEC (74.2 versus 126.1 μg/L) for the edge of field water bodies but produces a median value that exceeds that produced by the FOCUSSWS (16.2 versus 11.9 μg/L). The runoff results (see Figure 12) indicate that the MUSTool under predicts the maximum PEC (74.2 versus 126.1 μg/L) and median (17.5 versus 76.5 μg/L) of the FOCUSSWS PECs, albeit that there are only three relevant FOCUSSWS R scenarios. Scenario 2 – Mecoprop, a mobile and low persistence herbicide, applied to winter and spring cereals. The drainflow results (see Figure 13) indicate that the MUSTool under predicts the maximum PEC (33.1 versus 97.0 μg/L) for the edge of field water bodies, but produces a median value that is similar (1.5 versus 4.0 μg/L) to that produced by the FOCUSSWS. The runoff results (see Figure 14) indicate that the MUSTool predicts similar maximum (34.5 versus 38.5 μg/L) and median (1.1 versus 2.5 μg/L) PECs to that produced by FOCUSSWS. Scenario 3 – Atrazine, a mobile and moderately persistent herbicide, applied to maize. The drainflow results (see Figure 15) indicate that the MUSTool over predicts the maximum PEC (24.2 versus 7.1 μg/L) for the edge of field water bodies, but produces a median value that is similar (4.1 versus 6.6 μg/L) to that produced by the FOCUSSWS. The runoff results (see Figure 16) indicate that the MUSTool marginally under predicts the maximum PEC (36.7 versus 46.8 μg/L) and median (6.2 versus 20.3 μg/L) of the FOCUSSWS PECs. Scenario 4 – Propyzamide, a low mobility and moderately persistent herbicide, applied to winter oilseed rape. The drainflow results (see Figure 17) indicate that the MUSTool predicts similar maximum (4.8 versus 10.5 μg/L) and median (1.1 versus 4.6 μg/L) PECs to that produced by FOCUSSWS. The runoff results (see Figure 18) indicate that the MUSTool predicts similar maximum (7.5 versus 9.5 μg/L) and median (3.9 versus 9.3 μg/L) PECs to that produced by FOCUSSWS, albeit there are only two relevant R scenarios for this crop.


Median

25%-75%

Min-Max

D Conc

D WB Max PEC

FOCUS D Conc

FOCUS D WB Max PEC

Monitored D0

100

200

300

400

500

600

700

Concentr

ation (μg/L)

Figure 11 Box plot summary of the edge of field concentrations for scenario 1, isoproturon

in drainflow, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations

Median

25%-75%

Min-Max

R Conc

R WB Max PEC

FOCUS R Conc

FOCUS R WB Max PEC

Monitored R0

200

400

600

800

1000

1200

1400

1600

1800

2000

Concentr

ation (μg/L)

Figure 12 Box plot summary of the edge of field concentrations for scenario 1, isoproturon

in surface runoff, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations


Median

25%-75%

Min-Max

D Conc

D WB Max PEC

FOCUS D Conc

FOCUS D WB Max PEC

Monitored D0

50

100

150

200

250

300

350

400

450

Concentr

ation (μ

g/L

)

Figure 13 Box plot summary of the edge of field concentrations for scenario 2, mecoprop in

drainflow, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations

Median

25%-75%

Min-Max

R Conc

R WB Max PEC

FOCUS R Conc

FOCUS R WB Max PEC

Monitored R0

20

40

60

80

100

120

140

160

180

200

Concentr

ation (μ

g/L

)

Figure 14 Box plot summary of the edge of field concentrations for scenario 2, mecoprop in

surface runoff, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations


Median

25%-75%

Min-Max

D Conc

D WB Max PEC

FOCUS D Conc

FOCUS D WB Max PEC

Monitored D0

10

20

30

40

50

60

70

80

90

Concentr

ation (μg/L)

Figure 15 Box plot summary of the edge of field concentrations for scenario 3, atrazine in

drainflow, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations

Median

25%-75%

Min-Max

R Conc

R WB Max PEC

FOCUS R Conc

FOCUS R WB Max PEC

Monitored R0

20

40

60

80

100

120

140

160

180

200

220

240

260

Concentr

ation (μg/L)

Figure 16 Box plot summary of the edge of field concentrations for scenario 3, atrazine in

surface runoff, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations


Median

25%-75%

Min-Max

D Conc

D WB Max PEC

FOCUS D Conc

FOCUS D WB Max PEC

Monitored D0

10

20

30

40

50

Concentr

ation (μg/L)

Figure 17 Box plot summary of the edge of field concentrations for scenario 4, propyzamide

in drainflow, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations

Median

25%-75%

Min-Max

R Conc

R WB Max PEC

FOCUS R Conc

FOCUS R WB Max PEC

Monitored R0

10

20

30

40

50

Concentr

ation (μg/L)

Figure 18 Box plot summary of the edge of field concentrations for scenario 4, propyzamide

in surface runoff, for the MUSTool (first two boxes), FOCUS (second two boxes) and monitored (last box). Abbreviations – D Conc: Concentration in drain water; D WB: concentration in ditch/stream from drainflow; Max PEC: maximum predicted environmental concentration; Monitored D: monitored drainflow concentrations

Catchment Results


It was proposed that the catchment scale results of the MUSTool would be evaluated against Environment Agency catchment monitoring data. This evaluation exercise was included as a means to check the catchment results acknowledging that the correspondence between the two may be low owing to (1) the modelling process, including uncertainty in the models and the fact that only a single crop is being modelled at a time in the MUSTool when applications may be made to several crops, (2) the diffuse agricultural losses being just one source of pesticides lost to surface water and (3) the spatio-temporal nature of the Environment Agency monitoring dataset. The four scenarios (see Table 21) were implemented within the MUS Tool and the catchment results extracted. Both the monitoring data and the modelled catchment results were processed to produce distributions of catchment concentrations. Owing to different numbers of samples these were plotted on different axes to facilitate comparison of the shapes of the distributions. Scenario 1 The distribution of the modelled catchment scale concentrations closely mimic that of the unimodal monitored distribution (see Figure 19), especially for concentrations exceeding the drinking water standard of 0.1 μg/L. Scenario 2 The distribution of the modelled catchment scale concentrations for spring cereals on their own under-represents the unimodal monitored distribution (see Figure 20a), especially for concentrations exceeding the drinking water standard of 0.1 μg/L. However, when both spring and winter cereals are considered (the maximum catchment concentration from either crop) the modelled catchment scale concentrations more closely mimic that of the unimodal monitored distribution (see Figure 20b), especially for concentrations exceeding the drinking water standard of 0.1 μg/L, albeit the distribution is skewed towards lower concentrations by up to a factor of 10. Scenario 3 The distribution of the modelled catchment scale concentrations closely mimic that of the bimodal monitored distribution (see Figure 21), especially for concentrations exceeding the drinking water standard of 0.1 μg/L. Scenario 4 The distribution of the modelled catchment scale concentrations only mimic that of the first peak in the bimodal monitored distribution (see Figure 22) for concentrations below the drinking water standard of 0.1 μg/L. There is little correspondence for concentrations above the drinking water standard of 0.1 μg/L with the MUSTool predicting few catchment concentrations in this range. This is attributed to a disjunct between the monitoring regime and the modelling approach for high KOC compounds. The Environment Agency monitored concentrations are derived from whole sample analysis (water and sediment phase pesticides) while the MUSTool does not include sediment associated propyzamide.

Monitore

d F

requency M

odelle

d F

requency

0.005 0.01 0.02 0.03 0.04 0.05 0.1 0.25 0.5 1 2 5 >5

Catchment Concentrations (μg/L)

0

10

20

30

40

50

60

70

80

90

0

100

200

300

400

500

600

700

800

900

Monitored(L)

Modelled(R)

Figure 19 Frequency distributions of the monitored and modelled catchment concentrations

(μg/L) for isoproturon (scenario 1)


Monitore

d F

requency M

odelle

d F

requency

0.005 0.01 0.02 0.03 0.04 0.05 0.1 0.25 0.5 1 2 5 >5


0

10

20

30

40

50

60

70

80

90

0

10

20

30

40

50

Monitored(L)

Modelled(R)

(a)

Monitore

d F

requency M

odelle

d F

requency

0.005 0.01 0.02 0.03 0.04 0.05 0.1 0.25 0.5 1 2 5 >5


0

5

10

15

20

25

30

35

40

45

50

0

100

200

300

400

500

600

700

800

900

Monitored(L)

Modelled(R)

(b)

Figure 20 Frequency distributions of the monitored and modelled catchment concentrations (μg/L) for mecoprop-p (scenario 2). Modelled catchment concentrations are for applications to (a) spring cereals and (b) winter and spring cereals


Monitore

d F

requency M

odelle

d F

requency

0.05 0.01 0.02 0.03 0.04 0.05 0.1 0.25 0.5 1 >1


0

10

20

30

40

50

60

70

80

0

100

200

300

400

500

600

700

Monitored(L)

Modelled(R)


(μg/L) for atrazine (scenario 3)

Monitore

d F

requency M

odelle

d F

requency

0.005 0.01 0.02 0.03 0.04 0.05 0.1 0.25 0.5 1 2 5 >5


0

2

4

6

8

10

12

14

16

18

20

22

0

100

200

300

400

500

600

700

800

900

Monitored(L)

Modelled(R)


(μg/L) for propyzamide (scenario 4) The degree of correspondence between the modelled results for each of the four compounds and detections monitored in surface water is promising given all of the confounding factors, which include: (1) The MUSTool includes a number of sources of uncertainty, for example: the parameterisation of, and the pesticide fate models themselves; the meta-modelling of these and interpolation between the combinations modelled; the spatial datasets used to locate the modelled scenarios in the agricultural landscape and catchments; the aggregation of field scale losses to catchment scale (e.g. the extent of agricultural drains and surface runoff connectivity); the generalisation of agricultural practice in all regional catchments (e.g. likely rates,


application dates and proportion of crop treated). In addition, the tool only models one major crop at a time while in any given catchment diffuse pollution may arise from a number of crops treated with the compound (e.g. see Figure 20). (2) The MUSTool only considers diffuse losses of compound through runoff and drainage from agricultural fields as a function of good agricultural practice while in reality there may be a range of other direct or “point” sources, for example: sewage sludge applied to land (e.g. Ghanem et al., 2007), forestry usage (e.g. Willoughby et al., 2004), highway (e.g. Huang et al., 2004) and railway usage (e.g. Torstensson, 2004), municipal and residential urban usage (e.g. Ruban et al., 2005; Blanchoud et al., 2007) either running off directly into streams or being contributed via sewage treatment plant effluent, direct over-spraying of surface water bodies (Voluntary Initiative, 2012), trafficking of farm machinery through surface water bodies, cleaning of sprayers and farm machinery and handling of the pesticides during filling on hard standings connected to drains (Mason et al., 1999; Rose et al., 2004; Jaeken and Debaer, 2005). These other sources can contribute significantly to pesticide loads and concentrations in surface water, for example Kolpin et al. (2006) demonstrated that waste water treatment plant contributions of glyphosate and its degradation metabolite aminomethyl phosphonic acid (AMPA) accounted for a roughly two-fold increase in their frequency of detection. Blanchoud et al. (2007), by creating a detailed annual budget of pesticide contributions from agricultural, municipal and residential sources, showed that a similar contribution is made by urban pesticides running off of impervious surfaces into the Marne River when compared with agricultural losses of pesticides used on cultivated soils. Hanke et al. (2010) likewise found that ~60% of the load of glyphosate transported during key runoff events originated from sewage treatment effluent and combined/separated sewer over flows. UK and EU research (BE, DE, FR, SE) indicates that 20 to 90% of the total pesticide load leaving a surface water catchment can originate from handling operations on the farm steading and that specific stewardship programmes targeting this source can reduce these by 40 to 95% (Mason et al., 1999; Rose et al., 2004; Jaeken and Debaer, 2005). (3) The Environment Agency monitoring dataset may not be spatially complete with respect to pesticide monitoring (J. Kennedy – Q&A, CRD ERAR Meeting, 2010) and may not be temporally appropriate with around 12 samples being taken at monitoring sites per year and these not being targeted at specific pesticide transport events.

Summary and Conclusions The potential aquatic exposure following the ‘minor usage’ of pesticides is not well understood, with the current perception being that (a) due to the scale of use or (b) there being a registered use on a major crop, that the risk to the environment is small. This project looked to inform this issue through the development of a risk assessment framework and tool that provided an objective way of assessing the relative risk of product use on minor crops versus usage of the same product on major crops at edge of field and catchment scales. The risk assessment tool was conceptualised around a suite of meta-models which would in essence “look up” appropriate values, but be underpinned by standard regulatory models/approaches and a suite of spatial databases. The first of these meta-models is the field loss meta-model which predicts the maximum flux of compound, the associated fluxes of water and sediment as well as the number of days after application that this would occur, using user inputs of compound properties, crop treated and application timings and amounts. The second is the edge of field fate meta-model which assesses the likely predicted environmental concentration in edge of field water bodies given these fluxes of compound, water and sediment. Comparison of these concentrations with a user input regulatory acceptable ecotoxicological concentration (RAC) allows for the completion of a risk assessment and the classification of all hectares of treated crop into two risk classes, ‘at risk’ or ‘not at risk’. This model also predicts the downstream loss of compound from edge of field water bodies. The third is the catchment flow meta-model which predicts the likely volume of catchment runoff into which the mass of compound lost downstream may be diluted to calculate a catchment concentration. Comparison of these catchment concentrations with user input regulatory (e.g. drinking water standard) or ecotoxicological thresholds (e.g. Water Framework Directive environmental quality standard) and an associated uncertainty factor, allows for the completion of a catchment risk assessment and the classification of all catchments into one of four risk classes, ‘at risk’, ‘possibly at risk’, ‘possibly not at risk’ and ‘not at risk’. The tool itself was developed in C# with the underpinning data and queries stored in a Microsoft SQL database. The tool outputs the tabulated edge of field and catchment scale results to Microsoft Excel and the catchment scale results to a bespoke ADAS spatial data viewer. While any evaluation of a complex modelling framework such as the MUSTool is difficult, the evaluation of both the underlying meta-models as well as the outputs from the tool suggest that the tool is fit for release as a beta version. It is recommended that future phases of work include:

1. Additional exploration of the MUSTool be undertaken using more highly resolved spatio-temporal pesticide usage and monitoring data such as that collected by and for water company catchment compliance and risk assessments. ADAS have collected such pesticide usage data for three catchments for two separate water companies and are aware of a number of catchments with high spatio-temporal


resolution monitoring data collected by these and other water companies, primarily for undertakings to OFWAT and the Drinking Water Inspectorate.

2. Refinement of the edge of field meta-model to account for sediment associated pesticide being included in the catchment scale PECs, especially for high KOC compounds, and the inclusion of a sediment PEC as one of the outputs.

3. Review and implementation of spatial clustering metrics to objectively assess the degree of clustering of the cropping and ‘at risk’ areas.

4. Expansion of the spatial viewer to include non-categorical datasets such that the edge of field catchment results may be viewed in the spatial viewer.

Knowledge Exchange The approach and outputs from this project have been presented at a number of meetings, namely: Hughes, G.O., 2012. Demonstration of the CRD Minor Uses Screening Tool. Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 12 June 2012, York. Hughes, G.O., 2012. Impact of Minor Uses: Development of the CRD Minor Uses Screening Tool. Voluntary Initiative Water Sub-Group Meeting, 6 March 2012, Peterborough. Hughes, G.O., 2011. Impact of Minor Uses: Development of the CRD Minor Uses Screening Tool. Proceedings of the Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 14 June 2011, Solihull. Hughes, G.O., 2010. Can scale of use be used in aquatic risk assessments? Proceedings of the Chemicals Regulation Directorate Environmental Risk Assessment Research Meeting, 17 June 2010, Warwick.


References to published material

9. This section should be used to record links (hypertext links where possible) or references to other published material generated by, or relating to this project.

ADAS. 2002. Development of a database of agricultural drainage. Defra R&D Report No. ES0111. Adriaanse, P.I. 1996. Fate of pesticides in field ditches: the TOXSWA simulation model. , SC-DLO Report

90, DLO Winand Staring Centre for Integrated Land, Soil and Water Research, Wageningen, the Netherlands, 241 pp.

Alexander, L.V., and Jones, P.D. 2001. Updated precipitation series for the U.K. and discussion of recent extremes. Atmospheric Science Letters, doi:10.1006/asle.2001.0025.

Allen, R.G., Pereira, L.S., Raes, D., and Smith, M., 1998. Crop evapotranspiration Guidelines for computing crop water requirements. Food and Agriculture Organization of the United Nations, Rome.

Anthony, S.G., Wilson, L., Hodgkinson, R., Jordan, C., Higgins, A., Lilly, A., Baggaley, N., and Farewell, T. 2012. Agricultural Field Under Drainage Installation in the United Kingdom. Report for Defra Project AC0114, WP4 - Data Synthesis, Modelling and Management. . p. 118.

Bailey, R., 1990. Irrigated crops and their management. Farming Press Books, Ipswitch, UK. Beulke, S., Renaud, F., and Brown, C.D. 2002. Development of guidance on parameter estimation for the

preferntial flow model MACRO 4.2, Final report to DEFRA/PSD, project PL0538. 67pp. Birkenshaw, J., and Bailey, J., 2003. Irrigation Best Practice - A guide for vegetable growers. Department

for Environment Food and Rural Affairs, London, UK. Blanchoud, H., Moreau-Guigon, E., Farrugia, F., Chevreuil, M., and Mouchel, J.-M. 2007. Contribution by

urban and agricultural pesticide uses to water contamination at the scale of the Marne watershed. . Science of the Total Environment, 375: 168-179.

Boorman, D.B., Hollis, J.M., and Lilly, A. 1995. Hydrology of soil types: a hydrologically-based classification of the soils of the United Kingdom. Report no. 126. , Institute of Hydrology, Wallingford, UK. p. 146.

Buckley, D., Groves, S., Bailey, J., Peters, J., and Bradshaw, N., 2005. Irrigation Best Practice - A guide for potato growers. Department for Environment Food and Rural Affairs, London, UK.

Carsel, R.F., Smith, C.N., Mulkey, L.A., Dean, J.D., and Jowise, P. 1984. User's manual for the pesticide root zone model (PRZM): Release 1. , EPA/600/3-84/109. U.S. EPA Athens, GA., USA.

Carsel, R.F., Imhoff, J.C., Hummel, P.R., Cheplick, J.M., and Donigian, J., A.S. . 1998. PRZM-3, A Model for Predicting Pesticide and Nitrogen Fate in the Crop Root and Unsaturated Soil Zones: Users Manual for Release 3.0.

Castle, D.A., McCunnall, J., and Tring, I.M., 1984. Field Drainage, Principles and Practices. Batsford Academic, London, UK.

Centofanti, T., Hollis, J.M., Blenkinsop, S., Fowler, H.J., Truckell, I., Dubus, I.G., and Reichenberger, S. 2008. Development of agro-environmental scenarios to support pestcide risk assessment in Europe. The Science of the Total Environment, 407: 574-588.

Comber, A.J., Procter, C., and Anthony, S.G. 2007. A combined pycnophylactic-dasymetric method for disaggregating spatial data: the example of agricultural land use. Proceedings of the 2007 GISRUK conference, Maynooth, Ireland.

Environment Agency. 2009. Technical Method: Abstraction and flow regulation, River Basin Characterisation 2 Project, Environment Agency, Downloaded from http://www.environment-agency.gov.uk/static/documents/Research/ January 2011. p. 5.

FOCUS. 2000. FOCUS groundwater scenarios in the EU plant protection product review process. Report of the FOCUS Groundwater Scenarios Workgroup, EC Document Reference Sanco/321/2000. p. 197.

FOCUS. 2001. FOCUS Surface Water Scenarios in the EU Evaluation Process under 91/414/EEC. Report of the FOCUS Working Group on Surface Water Scenarios, EC Document Reference SANCO/4802/2001-rev.2. p. 245.

Fox, G.A., Thelin, G.P., Sabbagh, G.J., Fuchs, J.W., and Kelly, I.S. 2008. Estimating watershed level nonagricultural pesticide use from golf courses using geospatial methods. Journal of the American Water Resources Association, 44: 1363-1372.

Garthwaite, D.G., Thomas, M.R., Anderson, H.M., and Battersby, A. 2005. Pesticide usage survey report 210. Grassland and fodder crops in Great Britain 2005 (including aerial applications 2003 - 2005), FERA.

Garthwaite, D.G., Thomas, M.R., Parrish, G., Smith, L., and Barker, I. 2008. Pesticide usage survey report 224. Arable crops in Great Britain 2008 (including aerial applications), FERA.

Ghanem, A., Bados, P., Estaun, A.R., de Alencastro, L.F., Taibi, S., Einhorn, J., and Mougin, C. 2007. Concentrations and specific loads of glyphosate, diuron, atrazine, nonylphenol and metabolites thereof in French urban sewage sludge. Chemosphere, 69: 1368–1373.

Hallett, S.H., Thanigasalam, P., and Hollis, J.M. 1995. SEISMIC: a desktop information system for

http://www.environment-agency.gov.uk/static/documents/Research/

http://www.environment-agency.gov.uk/static/documents/Research/


assessing the fate and behaviour of pesticides in the environment. Computers and Electronics in Agriculture, 13: 227-242.

Hanke, I., Wittmer, I., Bischofberger, S., Stamm, C., and Singer, H. 2010. Relevance of urban glyphosate use for surface water quality. Chemosphere, 81: 422-429.

Holman, I.P., Dubus, I.G., Hollis, J.M., and Brown, C.D. 2004. Using a linked soil model emulator and unsaturated zone leaching model to account for preferential flow when assessing the spatially distributed risk of pesticide leaching to groundwater in England and Wales. Science of the Total Environment, 318: 73-88.

Huang, X., Pedersen, T., Fischer, M., White, R., and Young, T.M. 2004. Herbicide Runoff along Highways. 1. Field Observations Environ. Sci. Technol., 38: 3263-3271.

Institute of Hydrology. 1999. Flood Estimation Handbook. 5 volumes., Institute of Hydrology, Wallingford, UK.

Jaeken, P., and Debaer, C. 2005. Risk of water contamination by plant protection products during pre and post treatment operations. Annual Review Agricultural Engineering, 4: 93-114.

Jarvis, N.J. 1994. The MACRO model (Version 3.1). Technical Description and sample simulations. Reports and Dissertations, 19. Department of Soil Sciences, Swedish University of Agricultural Sciences, Uppsala, Sweden. p. 51.

Jarvis, N.J., Hollis, J.M., Nicholls, P.H., Mayr, T., and Evans, S.P. 1997. MACRO_DB: a decision-support tool to assess the fate and mobility of pesticides in soils. Environmental Modelling and Software, 12: 251-265.

Jarvis, N.J., Lindahl, A., Messing, I., Stenemo, F., Hollis, J., Reichenberger, S., and Dubus, I.G. 2007. Algorithm to completely parameterise MACRO from basic soil property data. Report DL21 of the FP6 EU-funded FOOTPRINT project www.eu-footprint.org, 18p.

Kjeldsen, T.R., Stewart, E.J., Packman, J.C., Bayliss, A.C., and Fowell, S. 2005. Revitalisation of the FSR/FEH rainfall-runoff method. , Final report to Defra/EA. CEH Wallingford.

Klein, M. 2007a. STEPS1-2-3-4 Draft User Manual, Fraunhofer Institut, Schmallenberg, Germany. p. 66. Klein, M., 2007b. Long term surface water simulations with STEPS-1-2-3-4. In: A.A.M. DelRe ed.,

Environmental fate and ecological effects of pesticides: Proceedings of the 13th Symposium on Pesticide Chemistry, Piacenza, Italy, pp. 950-957.

Kolpin, D.W., Thurman, E.M.L., E.A. , Meyer, M.T., Furlong, E.T., and Glassmeyer, S.T. 2006. Urban contributions of glyphosate and its degradate AMPA to streams in the United States. Science of the Total Environment, 354: 191-197.

Lawrence, M.G. 2005. The relationship between relative humidity and the dew point temperature in moist air: A simple conversion and applications. Bulletin of the American Meterological Society, 86: 225-233.

Mason, P., Foster, I., Carter, A., Walker, A., Higginbotham, S., Jones, R., and Hardy, I., 1999. Relative importance of point source contamination of surface waters: River Cherwell catchment monitoring study., XI Pesticide Chemistry Conference, Cremona, Italy, pp. 405-412.

Perry, M., and Atwood, J., 2003. Irrigation Best Practice - A guide for top and soft fruit growers. Department for Environment Food and Rural Affairs, London, UK.

Price, O.R., Hollis, J.M., and Mackay, N. 2006. Establishing the representativeness of FOCUS surface water scenarios for pesticide risk assessment in the UK landscape - Phase II. Final Report to the Pesticide Safety Directorate for project PS2220.

Price, O.R., Hollis, J.M., Ford, S., and Hughes, G.O. 2007. Establishing the Representativeness of Focus Surface Water Scenarios for Pesticide Risk Assessment in the UK Landscape - Phase III. Final Report to the Pesticide Safety Directorate for project PS2229. p. 126.

Ragg, J.M., Beard, G.R., George, H., Heaven, F.W., Hollis, J.M., Jones, R.J.A., Palmer, R.C., Reeve, M.J., Robson, J.D., and Whitfield, W.A.D., 1984. Soils and their use in midland and Western England. Bulletin No. 12. Soil Survey of England and Wales, Harpenden.

Reichenberger, S., Dubus, I.G., Boulahya, F., Hollis, J.M., and Jarvis, N.J. 2008. Database containing complete PRZM parameterisation for FOOTPRINT soil, climate and crop scenarios. Report DL20 of the FP6 EU-funded FOOTPRINT project. p. 32.

Rose, S., Basford, B., and Carter, A. 2004. Development of a design manual for agricultural pesticide handling and wash down areas, Final Report to Environment Agency, Project No. P2-200/PR. p. 150.

Ruban, V., Larrarte, F., Berthier, E., Favreau, L., Sauvourel, Y., Letellier, L., Mosisni, M.L., and Raimbault, G. 2005. Quantitative and qualitative hydrologic balance for a small suburban watershed in the Nantes region, France. Water Science and Technology, 51: 231-238.

Silgram, M. 2005. Scientific and techinical revision of the IRRIGUIDE water balance model., Final report for Defra project NT2517. p. 12.

Smith, L.P., and Trafford, B.D. 1976. Climate and Drainage, Reference Book 434 (formerly Technical Bulletin No. 34), HMSO, Ministry of Agriculture, Fisheries and Food, London, UK.

Soil Survey Staff. 1983. The National Soil Map of England and Wales, 1:250,000 scale (in six sheets), Ordnance Survey (Crown Copyright), Southampton.

Stenemo, F., and Jarvis, N.J. 2002. Guidance document and manual for the use of MACRO_DB v.2.0,

http://www.eu-footprint.org/


Unpublished report (http://www.mv.slu.se/bgf/defeng.htm). Stenemo, F., and Jarvis, N.J. 2003. Users guide to MACRO5.0, a model of water flow and solute transport

in macroporous soil., Department of Soil Sciences, Swedish University of Agricultural Sciences, Internal Report. 40pp.

Stenemo, F., Lindahl, A.M.L., Gardenas, A., and Jarvis, N. 2007. Meta-modeling of the pesticide fate model MACRO for groundwater exposure assessments using artificial neural networks. Journal of Contaminant Hydrology, 93: 270–283.

Tiffin, D., King, J., Perkins, S., and Basford, B., 2007. Irrigation Best Practice - A water management toolkit for field crop growers. Department for Environment Food and Rural Affairs, London, UK.

Torstensson, L. 2004. Herbicide Use On Railways. European Glyphosate Environmental Information Source: Reporting environmental impacts of glyphosate. Retrieved 2012, from http://www.egeis.org/home/monitoring/case_studies.php?article_id=109.

Voluntary Initiative. 2012. Keep Water Clean. Best Practice - Better Environment. from http://www.voluntaryinitiative.org.uk/content/water.aspx?m=19.

Weatherhead, E.K. 2007. Survey of irrigation of outdoor crops in 2005 - England and Wales, Cranfield University, Cranfield, UK. p. 30.

Whelan, M.J., Walter, C., Smith, B.G., and Pendlington, D. 2005. Pesticide Risk Management and Profiling Tool: 'PRoMPTing' a science-based approach to mitigating the risks of water quality impacts from pesticide use in agriculture. Proceedings of Proceedings of the OECD Workshop on Agriculture and Water: Sustainability, Markets and Policies, 14-18 November, Adelaide, Australia, p. 16.

Willoughby, I., Evans, H., Gibbs, J., Pepper, H., Gregory, S., Dewar, J., Nisbet, T., Pratt, J., McKay, H., Siddons, R., Mayle, B., Heritage, S., Ferris, R., and Trout, R. 2004. Reducing Pesticide Use in Forestry. Forestry Commission Practice Guide., Forestry Commission. p. 140.

http://www.mv.slu.se/bgf/defeng.htm)

http://www.egeis.org/home/monitoring/case_studies.php?article_id=109

http://www.voluntaryinitiative.org.uk/content/water.aspx?m=19

evidence project final reportrandd.defra.gov.uk/document.aspx?document=10599_ps2238...evid4 evidence...

Documents