Technical ReportDefra Projects WQ0106 and ES0205
Cost and Effectiveness of Policy Instruments for Reducing Diffuse Agricultural Pollution
Part One – Farm Scale Modelling
September 2006
Prepared for
Dr Soheila Amin-Hanjani
Defra
Catchment Sensitive Farming
and Water Quality Science
Head of Branch
Dr Steven Anthony
Principal Scientist
Environment Systems
ADAS UK Ltd
1. Introduction
This is the first of two reports that summarise technical aspects of modelling work for Defra’s Water Quality Division (Catchment Sensitive Farming Group) on estimating the cost and effectiveness of Policy Instruments for reducing diffuse agricultural pollution. This work follows on from (and compliments) the user-manual reported in ES0203 and this report document itself is part of reporting of ES0205.
This first report documents the methodology for estimating the cost and effectiveness of Policy instruments at farm scale. The second report documents extensions to the methodology for application at the catchment and national scale. The second report (that incorporates this one) will be included in the appendices of the final report to Defra on Project WQ0106, as this work is currently ongoing.
1.1Background
The Water Framework Directive (2000/60/EC) requires Member States to protect, enhance and restore all bodies of surface water (including rivers, lakes, estuaries and coastal waters) and groundwater with the aim of achieving good ecological status by 2015, subject to certain exemptions set out in the Directive. The Directive also requires Member States to put in place River Basin Management Plans for managing the water environment. These must include programmes of measures to address the pressures on water bodies. A key pressure is diffuse pollution from agricultural sources.
Defra’s Water Quality Division (Catchment Sensitive Farming Group) is working towards identifying practical on-farm methods for mitigating diffuse water pollution from agriculture (DWPA). Studies have been commissioned to review methods and to calculate their cost and effectiveness at farm scale. These include broad reviews of mitigation methods, such as the ‘Diffuse Pollution Inventory’ (Defra Project ES0203), and development of cost-curves for nitrate, phosphorus and other pollutants (Defra projects PE0203; NT2511; and ES0121) to prioritise method implementation. The combination of this work is a priority list of mitigation methods that variously control potential pollutant inputs, risk of mobilisation and delivery to water bodies from agricultural land (Table 1.1). The farm scale cost and effectiveness of individual methods has been determined by an expert group of ADAS and IGER scientists for phosphorus, nitrate and faecal indicator organisms and is summarised in a ‘User Manual’ that describes their implementation (Cuttle et al., 2006).
The next step for Defra was to develop Policy Packages (combinations of Policy Instruments) for inclusion into the programme of measures under the RBMPs that will maximise uptake of the methods by the farming community. Examples of Policy Instruments include codes of good agricultural practice, advisory schemes and European legislation. These Policy Packages would be put to public consultation and would require quantitative support for their national cost and effectiveness. The development of the Policy Packages therefore required a methodology to calculate the cost and effectiveness of groups of mitigation methods, taking account of the dependencies between them, and that could be scaled from the previous farm level assessments to give catchment and national effects. This would also require an
Table 1.1 Mitigation methods for the control of diffuse pollutant losses from agricultural land to water bodies (Cuttle et al., 2006).
Type / No. / MethodLand use / 1 / Convert arable land to extensive grassland
Soil Management / 2 / Establish cover crops in the autumn
3 / Cultivate land for crop establishment in spring rather than autumn
4 / Adopt minimal cultivation systems
5 / Cultivate compacted tillage soils
6 / Cultivate and drill across the slope
7 / Leave autumn seedbeds rough
8 / Avoid tramlines over winter
9 / Establish in-field grass buffer strips
10 / Loosen compacted soil layers in grassland fields
11 / Maintain and enhance soil organic matter levels
12 / Allow field drainage systems to deteriorate
Livestock Management / 13 / Reduce overall stocking rates on livestock farms
14 / Reduce the length of the grazing day or grazing season
15 / Reduce field stocking rates when soils are wet
16 / Move feed and water troughs at regular intervals
17 / Reduce dietary N and P intakes
18 / Adopt phase feeding of livestock
Fertiliser Management / 19 / Use a fertiliser recommendation system
20 / Integrate fertiliser and manure nutrient supply
21 / Reduce fertiliser application rates
22 / Do not apply phosphorus fertilisers to high phosphorus index soils
23 / Do not apply fertiliser to high-risk areas
24 / Avoid spreading fertiliser to fields at high-risk times
Manure Management / 25 / Increase the capacity of farm manure stores
26 / Minimise the volume of dirty water produced
27 / Adopt batch storage of slurry
28 / Adopt batch storage of solid manure
29 / Compost solid manure
30 / Change from slurry to a solid manure handling system
31 / Site solid manure heaps away from watercourses and field drains
32 / Site solid manure heaps on concrete and collect the effluent
33 / Do not apply manure to high-risk areas
34 / Do not spread farmyard manure to fields at high-risk times
35 / Do not spread slurry or poultry manure to fields at high-risk times
36 / Incorporate manure into the soil
37 / Transport manure to neighbouring farms
38 / Incinerate poultry litter
Connectivity
Management / 39 / Fence off rivers and streams from livestock
40 / Construct bridges for livestock crossing rivers and streams
41 / Re-site gateways away from high-risk areas
42 / Establish new hedges
43 / Establish riparian buffer strips
44 / Establish and maintain artificial (constructed) wetlands
integrated assessment of the effect of methods on all pollutants, to ensure that all potential impacts on water status (eutrophication, siltation, contamination) were addressed.
This report describes the technical development and application of this methodology to determine optimal Policy Packages and their cost and effectiveness. This report documents the methodology for a farm scale assessment. The methodology is extended for catchment and national scale assessment in a second document (incorporating this one) reported under Defra project WQ0106. The catchment scale predictions of the effects of a Policy Package on pollutant loss are integrated with models of river water quality in Defra projects WT0719CSF and WQ0106 (Module 3) to quantify the gap between the current status of water bodies and good status, and the impact of the Policy Packages in reducing this gap (Anthony and Lyons, 2006; Anthony and Collins, 2006).
1.2Objective
The general objective of the work was to provide a quantitative assessment of the farm and national scale cost and effectiveness of Policy Packages for reducing diffuse pollution. This would require an assessment of groups of mitigation methods for reducing diffuse nitrate, phosphorus, sediment and faecal indicator organism losses from agricultural land to surface waters. Cost and impact were to be assessed relative to current (2000) and future (2015) agricultural practices, taking account of the effects of existing instruments on method uptake and structural change in the agricultural sector.
The first objective of this project was therefore to provide baseline estimates of pollutant losses from representative farm types across England and Wales. Losses were to be calculated for nitrate, phosphorus, sediment and faecal indicator organisms (Sections 2 to 4).
The second objective of this project was to develop and implement a methodology for calculating the cost and effectiveness of groups of mitigation methods implemented on each of the model farms, to enable comparison of Policy Instruments at the farm scale (Sections 5 and 6).
The third objective of this project was to develop a methodology for scaling the farm scale pollutant losses, to provide national scale estimates of the cost and effectiveness of Policy Instruments. This would take account of forecast changes in the agricultural sector. This third objective is reported separately under Defra project WQ0106.
The project concerns only pollutant losses to fresh river waters and the geographical scope is confined to England and Wales.
2. Model Farm Systems
Analysis of the impact of Policy Instruments was begun at the farm scale, at which there was a more direct link to the results of field studies and economics data.
A number of model farm systems were defined for which to calculate the costs and effectiveness of individual mitigation methods implemented under potential Policy Instruments. The farm systems were defined to be representative of current practices, especially relating to the limits on the total nitrogen content of manures spread to land. Each was characterised by an area of arable or grassland, a number of livestock, and associated inputs of nutrients in fertiliser, excreta and managed manure.
The model farm systems were based on those defined by Defra project NT2010 (Appendix 10) and included: arable; beef; dairy; broiler and both indoor and out door breeding pig units. Table 2.1 lists the farm areas and numbers of animals. The farm systems were greatly simplified to represent only the fundamental pollutant pressures (e.g. nutrients input in fertiliser and manures) associated with a farm, using data that would be available nationally for the scaling of results. Each farm system was also reduced to a single dominant land use. Therefore, the dairy farm system did not include non-grass forage crops (e.g. maize), and a single crop type based on winter wheat represented arable cropping.
Table 2.1 Summary of model farm system field areas and animal numbers.
Farm System / Animal Count / Excreta(t yr-1) / Percent Managed / Field Area (ha)
Grass (Dairy)
Grass (Suckler Beef)
Breeding Pig (Indoor)
Breeding Pig (Outdoor)
Broiler
Arable / 270
220
1,330
2,536
150,000
0 / 5,040
1,850
2,125
3,568
2,550
0 / 60
60
100
0
100
n/a / 150
100
71
24
437
300
Rates of mineral fertiliser applied to the model farms were based on national average field rates for winter wheat, cut and grazed grass (BFSP, 2004). The percentages of the total application made in each calendar month were also obtained from the British Survey of Fertiliser Practice (2004; Table EW3.0). As field rates, the inputs would over-estimate the over-all national average fertiliser rate, taking account of the proportion of fields that do not receive fertiliser and the difference in rate between crop types. This would be taken account of in the scaling of the model farm pollutant losses to the national scale (Defra project WQ0106). The over-all rates would be specific to the farm type (BFSP Tables EW5.1 to EW5.4).
The quantities of excreta and managed manure produced on each farm were estimated from the numbers of livestock, length of housing period, and reference data for the daily production of excreta (Smith and Frost, 2000; Smith et al., 2000a). The total and available nitrogen and phosphorus content of the manures were estimated from standard tables (MAFF, 2001). The percentage of the manure volume applied in each calendar month was estimated from stratified survey data, and was specific to each combination of animal and manure type, and whether it was spread to arable or grassland (Smith et al., 2000b; Smith et al., 2001a, 2001b). Table 2.2 summarises the seasonal distribution of manure spreading.
The model farms did not include extensive (upland) systems. Pollutant losses from sheep, rough pasture and woodland or forestry would be accounted for in the national scaling of results (Defra project WQ0106) but the effect of mitigation methods would not be investigated.
Table 2.2 Summary of the seasonal distribution of managed manure
spreading to agricultural land on the mode farm systems.
Percentage of Total Manure VolumeManure Type / Nov to Jan / Feb to Apr / May to Jul / Aug to Oct
Pig Slurry to ArableLand / 12 / 24.9 / 10 / 53.1
Dairy Slurry to Grassland / 18.9 / 29.1 / 15.1 / 36.9
Beef Solid Manure to Grassland / 24 / 27.9 / 10 / 38.1
Broiler Litter to ArableLand / 6.9 / 12.9 / 3.1 / 77.1
2.1 Arable System
The model arable system was defined as an area of 300 ha of mixed combinable crops. The average field size was 8 ha. The system was represented by winter wheat and all the fields received an average annual ammonium nitrate fertiliser application of 207 kg ha-1 on clay loam and 197 kg ha-1 on sandy loam soils, and an average phosphate application of 65 kg ha-1 (BSFP, 2004). There was no grassland and no imported manure. The arable land was assumed to be fully tile-drained if on clay loam soil and completely undrained if on sandy loam.
2.2 Broiler System
The model broiler system was defined as 150,000 bird places. The total excreta production is 2,550 tonnes (Smith et al., 2000). The excreta is managed as solid manure (litter). The total nitrogen content of all the excreta is 74,250 kg and the total phosphate content is 61,875 kg annually. The litter is spread to neighbouring arable land in autumn. This land is considered to be part of the system. Given the maximum across-farm NVZ organic nitrogen application rate of 170 kg ha-1, the required arable field area is 436.7 ha. The arable land is assumed to be in conventional production, and is defined as for the arable system (Section 2.1) with a 25 kg ha-1 reduction in the nitrate fertiliser application rate to take account of the manure contribution (BSFP, 2004). It is assumed that 100% of the litter will be stored for an average of 2-3 months.
2.3 Breeding Pig (Indoor) System
The model breeding pig (indoor) system was defined as 290 dry sow, 60 farrowing sow, 585 first stage weaner and 565 second stage weaner places. The total excreta production is 2,125 tonnes annually (Smith et al., 2000). The excreta are managed as slurry, with slatted floors to the pig houses. The total nitrogen content of all the excreta is 11,986 kg and the total phosphate content is 14,875 kg annually. The slurry is spread to neighbouring arable land. This land is considered to be part of the system. Given the maximum across-farm NVZ nitrogen application rate of 170 kg ha-1, the arable field area required is 70.5 ha. The arable land is assumed to be in conventional production, and is defined as for the arable system (Section 2.1) with a 25 kg ha-1 reduction in the nitrate fertiliser application rate to take account of the manure contribution to nitrogen supply. It is assumed that 100% of the slurry will be stored for an average of 2-3 months.
2.4 Breeding Pig (Outdoor) System
The model breeding pig (outdoor) system was defined as 500 dry sows, 92 farrowing sows, and 1,944 first stage weaners. The total excreta production is 3,568 tonnes annually (Smith et al., 2000). The sows occupy an area of 24 ha and excrete 2,340 t annually across this free range dunging area (25 sows per ha). The weaners occupy an area of 0.1 ha (40 weaners per kennel plot of 23 m2). The quantity of excreta cleared from the kennels is negligible overall. There is no collection and storage of manure. The excreta has a total nitrogen content of 20,090 kg and a total phosphate content of 24,975 kg (Smith et al., 2000).
2.5 Dairy System
The model dairy system was defined as 150 adult dairy cows and 120 followers. The total cut and grazed grassland area was 150 ha. The average field size was 8 ha. The total excreta production is 5,040 tonnes annually (Smith et al., 2000). The animals are housed for 180 days each year. A total of 60% of excreta is deposited in the housing or parlour, and the rest at grazing. Excreta in housing, collecting yards and the parlour is managed as slurry and is stored for 2 to 4 months. All managed slurry was assumed to be spread across the grassland area. The total nitrogen content of all the excreta is 25,200 kg and the total phosphate content is 17,640 kg. The grassland area receives an average annual ammonium nitrate fertiliser application of 190 kg ha-1 and an average phosphate application of 32 kg ha-1 (BSFP, 2004). It was assumed that two thirds of the grassland area was tile-drained if on clay loam soil, and completely undrained if on sandy loam soil.
2.6 Beef System
The model beef system was defined as 80 adult cows, 70 calves and 70 yearlings. The total cut and grazed grassland area was 100ha. The average field size was 8 ha. The total excreta production is 2,288 tonnes annually (Smith et al., 2000). The animals are housed for 180 days each year and 50% of the excreta is managed as straw based manure and stored for 2 to 4 months. The total nitrogen content of all the excreta is 11,448 kg and the total phosphate content is 8,008 kg. The grassland area receives an average annual ammonium nitrate fertiliser application of 80 kg ha-1 and an average phosphate application of 29 kg ha-1 (BSFP, 2004). It was assumed that one third of the grassland area was tile-drained if on clay loam soil, and completely undrained if on sandy loam soil.
3. Climate and Soil Zones
Diffuse pollutant losses are sensitive to environment conditions. Key factors are rainfall and soil type. Rainfall frequency and intensity determines the volume of soil drainage and likelihood of surface runoff – and hence the solute leaching potential and soil erosion risk. Soil type (characterised by texture) determines the relative importance of surface runoff, preferential and matrix flow pathways, and hence the efficiency of leaching and the likelihood of retention (Haygarth et al., 2005). Mitigation methods that affect a specific mobilisation or delivery process (e.g. riparian buffer zones impact on soil erosion loss by surface runoff) will have geographically variable effectiveness.
Baseline pollutant losses from each model farm system were therefore calculated for combinations of soil texture (clay loam or sandy loam) and climate condition to represent the range of pollutant losses across England and Wales in response to environment factors. Locations in the south west (SX150450), southern (ST000150) and Anglian (TF580300) regions were selected to give representative climate inputs to the pollutant models (Table 3.1). Modelled soil drainage at these sites was in the range 170 to 620 mm and was comparable to the range across the intensive agricultural area of England and Wales (Figure 3.1). The clay loam and sandy loam textures were selected as representative of the majority of soils in England and Wales (Figure 3.2). Climate and soil zones were defined for the extrapolation of the field scale model results (Table 3.2 and Figure 3.3).
Table 3.1 Summary of climate statistics at locations selected to be representative of conditions across the intensive agricultural area of England and Wales.