general enquiries on this form should be made...
TRANSCRIPT
General enquiries on this form should be made to:
General enquiries on this form should be made to:
Defra, Procurements and Contracts Division (Science R&D Team)
Telephone No.0207 238 5734E-mail:[email protected]
SID 5Research 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 SID 5 (Research 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. A SID 5 must be completed for all projects.
· This form is in Word format and the boxes may be expanded or reduced, 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
WQ0106 Module 6
2.Project title
Quantitative Assessment of Scenarios for Managing Trade-Off Between Economics, Environment and Media
3.Contractororganisation(s)
ADAS UK Ltd
North Wyke Research
54.Total Defra project costs
£49,868
(agreed fixed price)
5.Project:start date
01 September 2008
end date
30 April 2009
6.It is Defra’s intention to publish this form.
Please confirm your agreement to do so.YES FORMCHECKBOX NO FORMCHECKBOX
(a)When preparing SID 5s 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 SID 5 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
Total Cost
Percentage Reduction (%):
ID
Type
Method Title
(£ Million per Annum)
NO
3
-N
P
Z
CH
4
NH
3
-N
N
2
O
4
Soil
Establish cover crops in the autumn
47
5.2
1.0
1.9
0.0
0.0
0.0
5
Soil
Early harvesting and establishment of crops in the autumn
148
1.0
0.2
0.3
0.0
0.0
0.0
6
Soil
Cultivate land for crops in spring rather than autumn
112
1.5
0.6
1.2
0.0
0.0
0.0
7
Soil
Adopt reduced cultivation systems
-81
2.4
2.1
10.4
0.0
0.0
0.0
8
Soil
Cultivate compacted tillage soils
16
0.7
7.9
13.6
0.0
0.0
0.0
9
Soil
Cultivate and drill across the slope
34
0.2
0.4
2.3
0.0
0.0
0.0
10
Soil
Leave autumn seedbeds rough
42
0.1
0.1
0.2
0.0
0.0
0.0
11
Soil
Manage over-winter tramlines
29
0.3
0.6
1.1
0.0
0.0
0.0
13
Soil
Establish in-field grass buffer strips
22
0.3
0.6
3.3
0.0
0.1
0.8
14
Soil
Loosen compacted soil layers in grassland fields
43
0.0
11.7
8.6
0.0
0.0
0.0
15
Soil
Establish riparian buffer strips
40
0.3
1.5
6.5
0.0
0.4
0.7
16
Soil
Allow field drainage systems to deteriorate
154
11.4
1.2
1.5
0.0
0.0
0.0
19
Research
Make use of improved genetic resources in livestock
-173
0.5
0.7
0.0
1.5
5.1
0.3
20
Research
Use plants with improved nitrogen use efficiency
-138
5.7
0.0
0.0
0.0
1.9
3.8
21
Fertiliser
Fertiliser spreader calibration
9
1.1
0.2
0.0
0.0
0.0
0.8
22
Fertiliser
Use a fertiliser recommendation system
-25
2.1
2.9
0.0
0.0
0.7
1.6
23
Fertiliser
Integrate fertiliser and manure nutrient supply
-261
0.4
2.9
0.0
0.0
0.2
1.6
25
Fertiliser
Do not apply fertiliser to high-risk areas
19
0.4
0.3
0.0
0.0
0.2
0.7
26
Fertiliser
Avoid spreading fertiliser to fields at high-risk times
23
0.8
0.9
0.0
0.0
0.1
0.0
27
Fertiliser
Fertiliser placement
0
0.0
0.0
0.0
0.0
0.0
0.0
29
Fertiliser
Replace urea fertiliser with another form
-28
0.0
0.0
0.0
0.0
10.9
0.0
30
Fertiliser
Incorporate a urease inhibitor into urea fertilisers
0
0.0
0.0
0.0
0.0
8.9
0.0
31
Fertiliser
Use clover in place of grass
-253
0.6
0.0
0.0
0.0
2.0
6.6
32
Fertiliser
Do not apply phosphorus fertilisers to high index soils
-46
0.0
5.8
0.0
0.0
0.0
0.0
33
Livestock
Reduce dietary nitrogen and phosphorus intakes
62
0.2
0.2
0.0
0.7
1.2
0.2
34
Livestock
Adopt phase feeding of livestock
6
0.1
0.1
0.0
0.0
0.4
0.1
35
Livestock
Reduce the length of the grazing day and grazing season
99
0.0
0.5
1.6
1.5
-8.4
0.9
36
Livestock
Extend the grazing season for cattle
32
0.0
-0.5
-1.6
-1.5
8.4
-0.9
37
Livestock
Reduce field stocking rates when soils are wet
91
0.0
0.8
1.5
1.4
-7.5
0.9
38
Livestock
Move feeders at regular intervals
126
0.0
0.8
1.5
0.0
0.0
0.0
39
Livestock
Construct troughs with concrete base
23
0.0
0.9
1.6
0.0
0.0
0.0
41
Livestock
Improved feed characterisation
0
0.6
0.7
0.0
1.9
2.6
0.8
44
Manure Housing
Increase scraping frequency in dairy cow cubicle housing
74
0.0
0.0
0.0
0.0
0.7
0.0
45
Manure Housing
Additional targeted bedding for straw-bedded cattle housing
160
0.0
0.0
0.0
0.0
0.9
0.0
46
Manure Housing
Washing down of dairy cow collecting yards
4
0.0
0.0
0.0
0.0
0.5
0.0
47
Manure Housing
Outwintering of beef cattle on woodchip stand-off pads
80
0.0
0.0
0.0
0.0
1.9
0.0
48
Manure Housing
Frequent removal of slurry from beneath-slat storage in pig housing
6
0.0
0.0
0.0
0.0
0.1
0.0
49
Manure Housing
Improved slatted-floor design for pig buildings
6
0.0
0.0
0.0
0.0
0.1
0.0
50
Manure Housing
Install air-scrubbers to mechanically ventilated pig housing
15
0.0
0.0
0.0
0.0
0.3
0.0
51
Manure Housing
Convert layer hen housing from deep-pit to belt manure removal
74
0.0
0.0
0.0
0.0
0.3
0.0
52
Manure Housing
Frequent manure removal from layer hen housing
0
0.0
0.0
0.0
0.0
0.5
0.0
53
Manure Housing
In-house poultry manure drying
9
0.0
0.0
0.0
1.1
1.1
0.0
54
Manure Storage
Increase the capacity of farm manure (slurry) stores
34
0.1
0.8
0.0
0.0
-1.8
0.0
56
Manure Storage
Install covers to slurry stores
43
0.0
0.0
0.0
2.1
0.6
0.0
57
Manure Storage
Allow cattle slurry stores to develop a natural crust
1
0.0
0.0
0.0
1.1
0.0
0.0
58
Manure Storage
Anaerobic digestion
966
0.0
0.0
0.0
8.4
0.0
0.0
59
Manure Storage
Minimise the volume of dirty water produced
12
0.1
0.8
0.0
0.0
0.0
0.0
62
Manure Storage
Store solid manure heaps away from watercourses and drains
1
0.1
0.7
0.0
0.0
0.0
0.0
63
Manure Storage
Store solid manure heaps on concrete and collect effluent
14
0.2
1.8
0.0
0.0
0.0
0.0
64
Manure Storage
Cover manure stores with polythene sheeting
18
0.0
0.4
0.0
0.0
1.6
0.0
65
Manure Storage
Manure liquid and solid separation
50
0.0
0.2
0.0
0.0
1.6
0.0
66
Manure Storage
Manure additives (e.g. Alum)
3
0.0
0.1
0.0
0.0
17.6
0.0
69
Manure Spreading
Manure spreader calibration
18
0.0
0.3
0.0
0.0
0.0
0.2
70
Manure Spreading
Do not apply manure to high-risk areas
5
0.1
0.6
0.0
0.0
0.0
0.0
71
Manure Spreading
Do not spread slurry at high-risk times
4
0.6
4.0
0.0
0.0
0.0
5.4
72
Manure Spreading
Slurry band spreading application techniques
64
0.0
0.0
0.0
0.0
4.9
0.0
73
Manure Spreading
Use slurry injection techniques
81
0.1
3.8
0.0
0.0
7.1
0.0
74
Manure Spreading
Do not spread solid manure to fields at high-risk times
3
0.4
2.9
0.0
0.0
0.0
0.0
75
Manure Spreading
Incorporate manure into the soil
14
0.0
0.9
0.0
0.0
2.3
0.0
77
Manure Spreading
Incinerate poultry litter
0
4.8
0.1
0.0
0.0
3.1
1.4
85
Connectivity
Fence off rivers and streams from livestock
35
0.3
2.6
0.0
0.0
0.0
0.0
86
Connectivity
Construct bridges for livestock crossing rivers and streams
46
0.1
0.6
0.0
0.0
0.0
0.0
87
Connectivity
Re-site gateways away from high-risk areas
19
0.1
1.3
3.0
0.0
0.0
0.0
88
Connectivity
Farm track management
23
0.0
0.1
0.0
0.0
0.0
0.0
89
Connectivity
Establish new hedges
0
0.0
0.0
0.0
0.0
0.0
0.0
92
Connectivity
Establish and maintain artificial wetlands - steading runoff
52
0.2
3.7
0.0
0.0
0.0
0.0
93
Connectivity
Irrigate crops to achieve maximum yield
17
0.1
0.0
0.0
0.0
0.0
0.0
94
Connectivity
Establish tree shelter belts around livestock housing
1
0.0
0.0
0.0
0.0
0.4
0.0
95
Connectivity
Establish and maintain artificial wetlands - field runoff
6
1.2
3.5
3.8
0.0
0.0
0.0
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 Defra Nutrient Management Programme aims to reduce the nutrient over-supply in the ecosystem through maximising the efficiency of the nutrient cycle on farms and thereby reducing emissions of pollutants to air and water. An important aspect of this programme is the development of a quantitative understanding of the impacts that potential pollution mitigation methods may have on multiple pollutants. Work completed by ADAS under Defra project WQ0106 included the use of a model framework to assess the impact of farm pollution control options on nutrient and sediment loss in response to a target for phosphorus loadings (Anthony, 2006; Anthony and Collins, 2006). This work built upon the characterisation of potential mitigation methods in the ‘User Manual’ of diffuse pollution methods (Cuttle et al., 2006). The manual is now being updated to include an extended list of mitigation methods, and impacts on gaseous emissions (ammonia, nitrous oxide and methane) in addition to losses to water (Newell-Price et al., 2009). The policy targets for the reduction of diffuse pollutant losses to air and water vary by pollutant. Achieving these targets will require catchment specific and wide ranging reductions in pollutant loads from the agricultural sector. There was a need to develop and apply a modelling framework for calculating the national cost and effect of mitigation methods for controlling multiple diffuse pollutants from agriculture. This project therefore aimed to evaluate:
· The baseline agricultural diffuse pollutant loads to air and water (nitrate, sediment, phosphorus, nitrous oxide, ammonia and methane) under normal agricultural practice for present day (year 2004) and projected (year 2020) land areas and stock numbers under a ‘Business as Usual’ scenario (Shepherd et al., 2007);
· The maximum reduction that could be achieved against this baseline for each pollutant by the implementation of mitigation methods selected from a revised ‘User Manual’ (Newell-Price et al., 2009), and the associated agricultural sector cost; and
· The most cost effective combinations of mitigation methods that achieved a target reduction of at least 30% for each and every pollutant;
The analyses were to be carried out separately by farm type, and reported both nationally (England and Wales) and at river basin district scale. The financial value of any environmental benefits from the reduction in pollutant load was to be estimated and set against the agricultural sector cost. Defra further requested that the calculated cost and effectiveness of a set of methods for controlling green house gas emissions were compared with the output from the independent and concurrent project RMP4950 (Moran et al., 2008).
This project has implemented a computer-modelling framework to evaluate the cost and effect of methods for the control of diffuse agricultural emissions to air and water. The framework implemented a novel genetic algorithm model to search for pareto-optimal combinations of mitigation methods. This proved valuable in generating a large number of potential solutions to pollution control, which could be analysed to provide insight into the range of possibilities. The search for optimal method combinations using the genetic algorithm framework has shown potential, but questions remain over the number of model generations required to accurately identify the optimal solutions when applied to all pollutants simultaneously. Further work is required to refine the methodology and to measure how well we are able to approximate the pareto front for a large number of dimensions.
The mitigation methods investigated were taken from the revised ‘User Manual’ of methods (Newell-Price et al., 2009) and generally represented potential for improved practice within existing farm systems rather than adoption of novel systems or technology. A total of 69 methods were investigated, each characterised for their impact on nitrate, phosphorus, sediment, nitrous oxide, methane and ammonia emissions. The effects of the mitigation methods were estimated from literature and represented as a percentage reduction against a specific source type, area and delivery pathway on representative model farms. Livestock, cropping, fertiliser and manure management practices on these model farms were derived from national stratified survey data. Baseline pollutant losses from the farms were calculated at field scale using a range of existing policy models and scaled nationally using agricultural census data, to provide outputs for Water Framework Directive river basins and farm types.
Modelled present day (year 2004) pollutant losses for England and Wales were 295 kt NO3-N; 4.4 kt P; 1906 kt Z; 595 kt CH4; 64 kt N2O; and 207 kt NH3. Projected changes in livestock numbers under a ‘Business as Usual’ scenario and small changes in method implementation due to the Nitrate Vulnerable Zones legislation and the England Catchment Sensitive Farming Delivery Initiative were forecast to result in net pollutant reductions of 5 to 9% by 2020. The maximum potential pollutant reductions achieved by the additional implementation of all available mitigation methods in the year 2020 were 39% for NO3-N; 55% for P; 47% for Z; 21% for CH4; 25% for N2O; and 57% for NH3, relative to the 2004 losses. The calculated potential ammonia reduction is high relative to previous published estimates as a consequence of the inclusion of methods effective against the fertiliser contribution to ammonia emissions, and also the inclusion of aluminium sulphate (Alum) additions to poultry litter as a mitigation method. Although presently unproven in the United Kingdom, the use of aluminium sulphate was assigned a very high efficiency and assumed applicable to all layer and broiler litter. If excluded from the analysis, then the maximum potential reduction in NH3 relative to the present day is reduced to 46%. If the replacement of urea fertiliser or the use of urease inhibitors were also excluded then the potential reduction declined further to a maximum of only 35%. This was illustrative of the sensitivity of the model results to the characterisation of only a few mitigation methods.
The implementation of all the methods resulted in a net cost to the agricultural sector of £2010 M yr-1, and were estimated to result in a society benefit of up to £705 M yr-1 based on use value and treatment costs avoided for emissions to water, and the human health and climate change impacts of gaseous emissions. The cost effect and cost benefit of pollution control were generally better on arable farm systems relative to livestock farms. Significant potential savings were calculated from the implementation of mitigation methods that reduced the use of inorganic fertiliser nitrogen and phosphorus. These methods included the ‘Use of Improved Livestock Genetic Resource’, ‘Use of Plants with Improved Nitrogen Efficiency’, ‘Sowing Clover to Grass Swards’ and the ‘Integration of Fertiliser and Manure Nutrient Value’ in fertiliser planning. The estimated society benefits of pollutant reduction were also greater than the implementation costs of methods for ‘Cultivation or Loosening of Compacted Soils’. The implementation of all methods with negative costs or positive net benefits in the year 2020 made it possible to achieve a 24% reduction in nitrate and 34 to 42% reduction in phosphorus, sediment and ammonia emissions for a net saving of £774 M yr-1 relative to 2004. However, if it were also necessary to maximise the reduction of methane emissions, then this would require the implementation of the ‘Anaerobic Digestion’ method for a net additional cost of £961 M yr-1.
It is important to caution that the calculations of method effect assumed that mitigation methods were implemented across 100% of the applicable land area. A number of the mitigation methods would prove unpopular for agronomic reasons, or are subject to ongoing research to prove their value. Farm economics, farmer attitude and the level of government support will determine the actual level of uptake. Further analyses of the potential for pollution control require an enhancement of the modelling framework to include an assessment of the realistic farm uptake of methods. This assessment would critically depend on whether a supportive or regulatory approach is taken by government.
The benefit calculations considered only reductions in pollutant load to receiving waters, and not the translation to pollutant concentrations and hence ecological status with respect to standards set under the Water Framework Directive. Further work might therefore take the spatial calculated pollutant losses and integrate them with estimates of loads from other sectors (including sewage effluent discharges and urban runoff) and an appropriate river model to determine whether the reductions in load have a significant impact on quality given the threshold concentration values for good ecological status. More critically, a comparison of the cost effect values for gaseous emission mitigation methods with those developed independently under project RMP4950 has highlighted the uncertainty in the estimation of cost and effect of individual methods. Different approaches to calculating method costs (whether it be linear-programming; regression modelling of farm accounts; or partial budgeting; see, for example, Moran et al., 2008; Bateman et al., 2008; Newell-Price et al., 2009) and different delivery groups can result in markedly different estimates of mitigation costs. Combined with the uncertainties in method effects due to sparse monitoring data or choice of model, the relative merits of individual methods can vary significantly between analyses. It is recommended that further work be done to develop more robust cost and effect estimates for packages of methods that can be related to farm practices, such as a manure or soil management plan. By aggregation of related methods into packages, it should be possible to relate them more easily to the delivery mechanism or potential policy instruments, and to reduce the overall uncertainty by considering change at a whole system level.
Scenario
Valuation
Annual Benefit (£ Million) Accruing from Pollutant Reduction
Total Benefit
Method
NO
3
-N
P
Z
CH
4
N
2
O
NH
3
-N
(£ Million)
2004 Maximum
Method (a)
23.2
23.2
21.8
40.8
116.6
201.1
426.7
Method (b)
71.8
108.6
95.0
40.8
116.6
201.1
634.0
2020 Prior
Method (a)
4.1
3.9
2.8
30.9
28.9
26.2
96.9
Method (b)
12.7
18.3
12.3
30.9
28.9
26.2
129.4
2020 Maximum
Method (a)
25.7
24.1
22.4
67.7
130.4
217.4
487.7
Method (b)
79.6
112.9
97.5
67.7
130.4
217.4
705.5
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 scientific 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 Transfer).
Quantitative Assessment of Scenarios for Managing Trade-Off between the Economic Performance of Agriculture and the Environment and Between Different Environmental Media
Defra Project WQ0106 (Module 6)
Prepared for:
Dr Phillip Earl
Defra Economic Advisor
Agriculture and Natural Resource Economics
April 2009
Steven Anthony1, Doris Duethman1, Richard Gooday1
David Harris1, Paul Newell-Price1, David Chadwick2, Tom Misselbrook2
1ADAS UK Ltd
2North Wyke Research
River Basin
Cattle and
Sheep -
Lowland
Cattle and
Sheep -
Upland
Dairy
Horticulture
Mixed
Mixed
Combinable
and Pig
Outdoor Pigs
Roots and
Poultry
Winter
Combinable
and Pig
All Farm
Types
Anglian
9.43
0.00
6.45
1.11
2.99
2.30
16.72
2.72
1.87
4.84
Dee
5.20
3.12
5.14
0.48
2.64
1.26
1.78
1.60
1.08
2.48
Humber
7.91
5.83
5.60
0.75
3.28
2.16
10.28
2.47
1.82
4.45
North West
4.54
3.21
4.60
0.37
2.53
1.42
3.54
1.97
1.08
2.58
Northumbria
6.05
4.95
4.39
0.61
2.49
1.47
4.88
2.01
1.15
3.11
Severn
6.01
3.33
5.20
0.64
2.92
1.31
3.60
2.13
1.24
2.93
Solway Tweed
4.52
3.95
3.91
0.31
2.02
1.27
3.04
1.82
0.71
2.40
South East
5.20
0.00
5.68
0.50
3.50
1.61
6.30
1.80
1.00
2.84
South West
5.48
5.19
4.90
0.23
2.96
1.45
5.83
1.79
0.87
3.19
Thames
6.49
0.00
5.87
0.59
3.69
1.59
7.00
1.70
1.29
3.14
Western Wales
4.54
3.26
4.31
0.09
2.15
1.43
0.88
1.76
0.78
2.13
England and Wales
6.03
3.85
4.94
0.64
3.02
1.89
8.46
2.31
1.48
3.62
Executive Summary
The Defra Nutrient Management Programme aims to reduce the nutrient over-supply in the ecosystem through maximising the efficiency of the nutrient cycle on farms and thereby reducing emissions of pollutants to air and water. An important aspect of this programme is the development of a quantitative understanding of the impacts that potential pollution mitigation methods may have on multiple pollutants. Work completed by ADAS under Defra project WQ0106 included the use of a model framework to assess the impact of farm pollution control options on nutrient and sediment loss in response to a target for phosphorus loadings (Anthony, 2006; Anthony and Collins, 2006). This work built upon the characterisation of potential mitigation methods in the ‘User Manual’ of diffuse pollution methods (Cuttle et al., 2006). The manual is now being updated to include an extended list of mitigation methods, and impacts on gaseous emissions (ammonia, nitrous oxide and methane) in addition to losses to water (Newell-Price et al., 2009). The policy targets for the reduction of diffuse pollutant losses to air and water vary by pollutant. Achieving these targets will require catchment specific and wide ranging reductions in pollutant loads from the agricultural sector. There was a need to develop and apply a modelling framework for calculating the national cost and effect of mitigation methods for controlling multiple diffuse pollutants from agriculture. This project therefore aimed to evaluate:
· The baseline agricultural diffuse pollutant loads to air and water (nitrate, sediment, phosphorus, nitrous oxide, ammonia and methane) under normal agricultural practice for present day (year 2004) and projected (year 2020) land areas and stock numbers under a ‘Business as Usual’ scenario (Shepherd et al., 2007);
· The maximum reduction that could be achieved against this baseline for each pollutant by the implementation of mitigation methods selected from a revised ‘User Manual’ (Newell-Price et al., 2009), and the associated agricultural sector cost; and
· The most cost effective combinations of mitigation methods that achieved a target reduction of at least 30% for each and every pollutant;
The analyses were to be carried out separately by farm type, and reported both nationally (England and Wales) and at river basin district scale. The financial value of any environmental benefits from the reduction in pollutant load was to be estimated and set against the agricultural sector cost. Defra further requested that the calculated cost and effectiveness of a set of methods for controlling green house gas emissions were compared with the output from the independent and concurrent project RMP4950 (Moran et al., 2008).
This project has implemented a computer-modelling framework to evaluate the cost and effect of methods for the control of diffuse agricultural emissions to air and water. The framework implemented a novel genetic algorithm model to search for pareto-optimal combinations of mitigation methods. This proved valuable in generating a large number of potential solutions to pollution control, which could be analysed to provide insight into the range of possibilities. The search for optimal method combinations using the genetic algorithm framework has shown potential, but questions remain over the number of model generations required to accurately identify the optimal solutions when applied to all pollutants simultaneously. Further work is required to refine the methodology and to measure how well we are able to approximate the pareto front for a large number of dimensions.
The mitigation methods investigated were taken from the revised ‘User Manual’ of methods (Newell-Price et al., 2009) and generally represented potential for improved practice within existing farm systems rather than adoption of novel systems or technology. A total of 69 methods were investigated, each characterised for their impact on nitrate, phosphorus, sediment, nitrous oxide, methane and ammonia emissions. The effects of the mitigation methods were estimated from literature and represented as a percentage reduction against a specific source type, area and delivery pathway on representative model farms. Livestock, cropping, fertiliser and manure management practices on these model farms were derived from national stratified survey data. Baseline pollutant losses from the farms were calculated at field scale using a range of existing policy models and scaled nationally using agricultural census data, to provide outputs for Water Framework Directive river basins and farm types.
Modelled present day (year 2004) pollutant losses for England and Wales were 295 kt NO3-N; 4.4 kt P; 1906 kt Z; 595 kt CH4; 64 kt N2O; and 207 kt NH3. Projected changes in livestock numbers under a ‘Business as Usual’ scenario and small changes in method implementation due to the Nitrate Vulnerable Zones legislation and the England Catchment Sensitive Farming Delivery Initiative were forecast to result in net pollutant reductions of 5 to 9% by 2020. The maximum potential pollutant reductions achieved by the additional implementation of all available mitigation methods in the year 2020 were 39% for NO3-N; 55% for P; 47% for Z; 21% for CH4; 25% for N2O; and 57% for NH3, relative to the 2004 losses. The calculated potential ammonia reduction is high relative to previous published estimates as a consequence of the inclusion of methods effective against the fertiliser contribution to ammonia emissions, and also the inclusion of aluminium sulphate (Alum) additions to poultry litter as a mitigation method. Although presently unproven in the United Kingdom, the use of aluminium sulphate was assigned a very high efficiency and assumed applicable to all layer and broiler litter. If excluded from the analysis, then the maximum potential reduction in NH3 relative to the present day is reduced to 46%. If the replacement of urea fertiliser or the use of urease inhibitors were also excluded then the potential reduction declined further to a maximum of only 35%. This was illustrative of the sensitivity of the model results to the characterisation of only a few mitigation methods.
The implementation of all the methods resulted in a net cost to the agricultural sector of £2010 M yr-1, and were estimated to result in a society benefit of up to £705 M yr-1 based on use value and treatment costs avoided for emissions to water, and the human health and climate change impacts of gaseous emissions. The cost effect and cost benefit of pollution control were generally better on arable farm systems relative to livestock farms. Significant potential savings were calculated from the implementation of mitigation methods that reduced the use of inorganic fertiliser nitrogen and phosphorus. These methods included the ‘Use of Improved Livestock Genetic Resource’, ‘Use of Plants with Improved Nitrogen Efficiency’, ‘Sowing Clover to Grass Swards’ and the ‘Integration of Fertiliser and Manure Nutrient Value’ in fertiliser planning. The estimated society benefits of pollutant reduction were also greater than the implementation costs of methods for ‘Cultivation or Loosening of Compacted Soils’. The implementation of all methods with negative costs or positive net benefits in the year 2020 made it possible to achieve a 24% reduction in nitrate and 34 to 42% reduction in phosphorus, sediment and ammonia emissions for a net saving of £774 M yr-1 relative to 2004. However, if it were also necessary to maximise the reduction of methane emissions, then this would require the implementation of the ‘Anaerobic Digestion’ method for a net additional cost of £961 M yr-1.
It is important to caution that the calculations of method effect assumed that mitigation methods were implemented across 100% of the applicable land area. A number of the mitigation methods would prove unpopular for agronomic reasons, or are subject to ongoing research to prove their value. Farm economics, farmer attitude and the level of government support will determine the actual level of uptake. Further analyses of the potential for pollution control require an enhancement of the modelling framework to include an assessment of the realistic farm uptake of methods. This assessment would critically depend on whether a supportive or regulatory approach is taken by government.
The benefit calculations considered only reductions in pollutant load to receiving waters, and not the translation to pollutant concentrations and hence ecological status with respect to standards set under the Water Framework Directive. Further work might therefore take the spatial calculated pollutant losses and integrate them with estimates of loads from other sectors (including sewage effluent discharges and urban runoff) and an appropriate river model to determine whether the reductions in load have a significant impact on quality given the threshold concentration values for good ecological status. More critically, a comparison of the cost effect values for gaseous emission mitigation methods with those developed independently under project RMP4950 has highlighted the uncertainty in the estimation of cost and effect of individual methods. Different approaches to calculating method costs (whether it be linear-programming; regression modelling of farm accounts; or partial budgeting; see, for example, Moran et al., 2008; Bateman et al., 2008; Newell-Price et al., 2009) and different delivery groups can result in markedly different estimates of mitigation costs. Combined with the uncertainties in method effects due to sparse monitoring data or choice of model, the relative merits of individual methods can vary significantly between analyses. It is recommended that further work be done to develop more robust cost and effect estimates for packages of methods that can be related to farm practices, such as a manure or soil management plan. By aggregation of related methods into packages, it should be possible to relate them more easily to the delivery mechanism or potential policy instruments, and to reduce the overall uncertainty by considering change at a whole system level.
1 Introduction
This report summarises technical work for the Defra Water Quality Division to provide a model-based assessment of the potential to reduce agricultural emissions of pollutants to air and water. The work links to and builds upon the results of Defra projects WQ0106: Cost and Effectiveness of Policy Instruments; WT0706CSF: Benefits and Pollution Swapping; SFF0601: Business as Usual III; and WT0743CSF: Evaluating the Extent of Agricultural Phosphorus Losses across Wales. In particular, this project develops the farm assessments of pollutant loss and mitigation method effect developed in project WQ0106 Module 5 (revised ‘User Manual’ of mitigation methods; Newell-Price et al., 2009) for application at national scale using the methodology developed under project WT0743CSF (Anthony et al., 2008).
1.1 Background
The Defra Nutrient Management Programme aims to reduce the nutrient over-supply in the ecosystem through maximising the efficiency of the nutrient cycle on farms and thereby reducing emissions of pollutants to air and water. The programme will achieve this by improving the coordination of the many areas of diffuse pollutant control work within Defra. An important aspect of this coordination is the development of a quantitative understanding of the impacts that potential pollution mitigation methods may have on multiple pollutants.
Work completed by ADAS under Defra project WQ0106 included the use of a model framework to assess the impact of farm pollution control options on nutrient and sediment loss in response to a target for phosphorus loadings (Anthony, 2006; Anthony and Collins, 2006). The work demonstrated that methods for the control of phosphorus losses would also benefit sediment and nitrate, and that the overall cost of pollution control could be minimised by careful selection of a suite of methods. This work built upon the characterisation of potential mitigation methods in the ‘User Manual’ of diffuse pollution methods (Cuttle et al., 2006). The manual is now being updated to include an extended list of mitigation methods, and impacts on gaseous emissions (ammonia, nitrous oxide and methane) in addition to losses to water (Newell-Price et al., 2009).
The policy targets for the reduction of diffuse pollutant losses to air and water vary by pollutant. The United Kingdom is legally bound to reduce green house gas emissions by 80% by 2050, and to reduce CO2 emissions by 26% by 2020 (Climate Change Act, 2008). The United Kingdom is also required to reduce ammonia emissions to 297 kt yr-1 under the Gothenburg Protocol and National Emission Ceilings Directive by 2010. The Nitrates Directive (81/676/EEC) sets a standard of 50 mg l-1 NO3 for nitrate concentrations in surface and ground waters. The UK implementation of the Water Framework Directive (2000/60/EC) sets phosphorus concentration standards of 50 to 120 ug l-1 for good ecological status, and the Freshwater Fish Directive (78/659/EC) sets a guideline target of 25 mg l-1 suspended solids for salmonid and cyprinid waters. Achieving these quality standards will require catchment specific and wide ranging reductions in pollutant loads from the agricultural sector. The potential synergy between mitigation methods selected to target water and air pollutants has not previously been quantified at national scale. There was therefore a need to develop a new modelling framework fit for the purpose of assessing the cost and effectiveness of mitigation methods against multiple pollutants and multiple targets.
1.2 Objectives
The objectives of this project were to develop and apply a modelling framework for calculating the national cost and effect of mitigation methods for controlling multiple diffuse pollutants from agriculture. The project aimed to evaluate:
· The baseline agricultural diffuse pollutant loads to air and water (nitrate, sediment, phosphorus, nitrous oxide, ammonia and methane) under normal agricultural practice for present day (year 2004) and projected (year 2020) land areas and stock numbers under a ‘Business as Usual’ scenario (Shepherd et al., 2007);
· The maximum reduction that could be achieved against this baseline for each pollutant by the implementation of mitigation methods selected from a revised ‘User Manual’ (Newell-Price et al., 2009), and the associated agricultural sector cost;
· The most cost effective combinations of mitigation methods that achieved a target reduction of at least 30% for each and every pollutant;
The analyses were to be carried out separately by farm type, and reported both nationally (England and Wales) and at river basin district scale. The financial value of any environmental benefits from the reduction in pollutant load was to be estimated and set against the agricultural sector cost.
Defra further requested that the calculated cost and effectiveness of a set of methods for controlling green house gas emissions were compared with the output from the independent and concurrent project RMP4950 (Moran et al., 2008).
Sections 2 and 3 of this report present an overview of the approach and the environment and agricultural data used as the baseline for these analyses. Section 4 describes the pollutant loss models, and Sections 5 to 6 describe the candidate mitigation methods and the algorithm for identifying the most cost effective combinations of methods. Sections 7 and 8 presents the results of the baseline pollutant loss calculations and the potential cost and effect of mitigation method implementation. Section 9 presents a comparison of results with project RMP4950. And Section 10 concludes with a brief discussion of the results.
The primary physical output from this study is a database of potential solutions to the control of agricultural diffuse pollution for England and Wales. Each solution is a combination of mitigation methods drawn from the revised ‘User Manual’ of mitigation methods (Newell-Price et al., 2009). The solutions are presented with estimates of the national cost of implementation and the percentage reduction achieved for each pollutant relative to an estimate of present day (year 2004) losses, alongside an estimate of the society benefit arising from the total pollution reduction. The solutions have been created using an algorithm that searches for pareto-optimal solutions. This database is provided separately to Defra. The reported maximum achievable pollutant reductions and most cost effective solutions have been selected from this database.
2 General Methodology
Anthony (2006) and Anthony et al. (2008) demonstrated a generic methodology for calculating the cost and effect of mitigation methods for control of diffuse agricultural pollution. The methodology involved the derivation of export coefficients from the output of mechanistic models applied to model farm systems that are representative of typical practise. The export coefficients express the pollutant loss as a linear function of the potential pollutant input to the farm system in the form of fertiliser and excreta. In a deviation from typical export coefficient models, losses are also expressed as a function of the land area where it is necessary to represent pollutant sources that are intrinsic or respond slowly to reducing inputs. The export coefficients are derived by applying the mechanistic models to a large number of scenarios representative of the national range of environment conditions. Area-weighted average coefficients are calculated for types of soil or climate so that the export coefficients represent a statistical summary of potential losses across a large area. When used with catchment and national scale agricultural census and practice data, the export coefficients provide an efficient and easily scalable estimate of baseline pollutant losses. The calculated losses are equivalent to the output from running the mechanistic models for each combination of soil and climate within an area, but are calculated considerably more quickly and can therefore be used in an iterative optimisation algorithm.
The mechanistic models supporting the export coefficients are used to explicitly disaggregate the total pollutant loss between each of the source types, areas and pathways on a farm, based on the detailed management assumptions for each model farm system. The effect of a potential mitigation method is then expressed as a percentage reduction in the pollutant loss from a specific source type, area or pathway. Hence it is possible to represent the targeted impact of mitigation methods. The effect values are taken from the literature, along with estimates of implementation cost. The net cost and effect of a combination of mitigation methods depends on the extent to which they target the same source types, areas and pathways.
A library of mitigation methods is integrated with the export coefficient model to calculate the cost and effect of policy instruments on diffuse pollutant losses. A policy instrument is represented as the implementation of a collection of individual mitigation methods. The extent of implementation can be based upon a theoretical scenario of maximum potential uptake, or estimates of likely landowner response to supportive and legislative instruments. More interestingly, computer tools have been developed that use iterative algorithms to calculate the most cost-effective set of mitigation methods to achieve a target pollutant reduction. The optimal policy instrument could in principle be found by simulating all possible combinations of mitigation methods, but this is a numerically intensive task that scales exponentially with the number of mitigation methods. Existing frameworks have therefore minimised the computing resource by calculating a cost curve for either a single pollutant (Anthony et al., 2008) or a weighted sum of pollutants (Anthony, 2006). However, this single-objective approach requires that a relative value or weight can be assigned to each pollutant at the outset of the computation. It is then possible that the optimal policy instrument will be over-looked as we pre-emptively constrain the optimisation process.
This project has therefore adopted a proper multi-objective methodology to identify optimal policy instruments. It is derived from the Non-Dominated Sorting Genetic Algorithm (NSGA-II; Deb et al., 2001) and uses principles of evolution to simultaneously optimise on all objectives and minimise the computing resource. The methodology identifies the pareto-optimal front of non-dominated solutions. A solution or combination of mitigation methods dominates another solution if it is superior or equal in all objectives but at least superior in one objective. The complete set of non-dominated solutions represents the best available solutions for achieving each objective. After calculation, the set can be sieved to identify the solutions that satisfy a range of constraints. It is therefore possible to delay the weighting of objectives (such as the relative worth of reducing phosphorus versus nitrate pollution) until all of the possible outcomes have been reviewed. This approach is also being investigated under Defra project WQ0106 Module 3 in support of farm scale mitigation method optimisation.
2.1 Spatial Framework
The modelling framework used in this study was designed to reflect the variability of environment risk factors across England and Wales. The basic spatial unit of calculation was the 10 by 10 km2 grid used by the ‘Business as Usual III’ project (Shepherd et al., 2007). Although pollutant losses have been calculated using environment data at a sub-grid scale, many of the assumptions about farm practice or, for example, soil properties are based on interpolated data from national surveys. The mapping of the Defra agricultural census from incompatible and large spatial units introduces additional spatial uncertainty. Model calculations for an individual grid cell are therefore subject to significant uncertainty. Outputs for grid cells were therefore aggregated to the Water Framework Directive River Basin Districts for reporting (Figure 2.1).
"
!
1
"
!
10
"!
8
"
!
9
"
!
11
"!
6
"
!
2
"
!
3
"
!
4
"
!
5
"!
7
1.Anglian
2.Dee
3.Humber
4.North West
5.Northumbria
6.Severn
7.Solway Tweed
8.South East
9.South West
10.Thames
11.Western Wales
Figure 2.1 Spatial extents of the River Basin Districts defined by the Water Framework Directive for England and Wales, superimposed on the 10 by 10km2 grid used for summarising environment risk factors and model calculations.
2.2 Environment Data
The mechanistic models selected for calculating pollutant emissions to water were applied to agricultural data at sub-grid scale across England and Wales, using statistical summaries of soil data and climate data derived from national inventories at a spatial scale of 1 km2. The results of these calculations were then area-weighted to create regional average loss values for generic soil types and climate zones (see Section 4).
2.2.1 Climate Data
Climate data for the standard period 1961 to 1990 were obtained from the Climate Research Unit at the University of East Anglia (Barrow et al., 1993). Six climate zones were defined based on the range of annual average rainfall, so that calculated pollutant losses for a River Basin District were responsive to the regional leaching and runoff pressure (Figure 2.2).
Annual
Rainfall (mm)
< 600
601 - 700
701 -900
901 - 1200
1201 - 1500
> 1500
Figure 2.2 Climate zones defined by ranges of average annual rainfall (mm)
for the standard period 1961-90 across England and Wales.
2.2.2 Soils Data
The National Soils Resources Institute (NSRI) at Cranfield University provided soils data for input to the mechanistic loss models under a Defra contractor licence. The input data required for the pollutant models were the particle size distribution (percentage sand, silt and clay), organic matter content and bulk density of the dominant soil series at every 1 km2 across England and Wales. The soils information also included the Hydrology of Soil Types (HOST) class (Boorman et al., 1995) from which the models determined the relative importance of surface and subsurface flow paths. This was critical for assessing the potential impact of mitigation methods that were effective against only one pathway.
To simplify the development of the pollutant export coefficients, two representative soil groups were defined based on the likelihood that a soil would normally have artificial under-drainage when supporting arable cultivation. This separated the individual soil series into a group that was generally regarded as free draining, and a group that was slowly permeable or impermeable (Figure 2.3). The need for under-drainage was estimated based on the soil HOST class using rules from the PSYCHIC model (Davison et al., 2008). The mechanistic pollutant loss models were run using the attributes of each and every soil series within a soil group, and the results area weighted to generate a single export coefficient for each drainage group.
Percentage
Drainage Soils
0 - 20
21 - 40
41 - 60
61 - 80
81 - 100
Figure 2.3 Percentage of agricultural land located on soils that would normally be expected to have artificial under-drainage when supporting arable cultivation.
3 Agricultural Land Use and Practice
Model farm descriptions and national agricultural census data were used as input to field scale mechanistic pollutant models to calculate representative pollutant loadings in the absence of mitigation methods for each of the 10 by 10 km2 grid cells. Pollutant loadings in the absence of any mitigation method implementation are referred to as baseline scenario losses. The baseline scenario results over-estimate present day actual pollutant load as many mitigation methods are already part implemented by farmers in response to existing policy instruments. Calculated pollutant loads that took account of present day method implementation are referred to as prior implementation scenario losses. The assumptions about prior implementation of mitigation methods are a key aspect of the definition of agricultural practise.
3.1 Model Farm Systems
The calculation of baseline and scenario pollutant losses from agricultural land using field scale mechanistic models required an explicit definition of standard agricultural practise for the present day. This was based on the representative model farm systems developed to support the revised ‘User Manual’ of mitigation methods (Newell-Price et al., 2009). The detailed assumptions about farm practise on the model farm systems are given in Newell-Price et al. (2009). We present only a summary of relevant data.
3.1.1 Crop Areas and Stock Numbers
The representative model farm systems were based on the nine ‘Robust Farm Types’ (RFT) used by the Farm Business Survey and defined by the dominant source of revenue (MAFF, 1993). The farm systems excluded the ‘Other’ RFT that defines holdings that either do not fit in well with mainstream agriculture or are of limited economic importance (Table 3.1).
Table 3.1 Representative model farm systems defined by the revised ‘User Manual’ of mitigation methods project, and mapping to the Defra ‘Robust Farm Types’ (Newell-Price et al., 2009).
‘User Manual’ Farm Type
‘Robust Farm Type’
Dairy
Specialist Dairy
Less Favoured Area - Grazing Livestock
Less Favoured Area - Grazing Livestock
Lowland - Grazing Livestock
Lowland - Grazing Livestock
Mixed
Mixed
Combinable Crops*
Specialist Cereal
Roots Combinable
General
Indoor Pig
Specialist Pig
Outdoor Pig
Specialist Pig
Poultry
Specialist Poultry
Horticulture
Horticulture
*Split into a ‘Mixed’ and a ‘Winter’ combinable crop farm;
The model farm sizes (total arable crop and grass area) were based on the average farm areas given in the Farm Business Survey for 2006, for England only. The farms surveyed by the Farm Business Survey are generally larger than the average census farm as the survey excludes minor holdings. The proportions of the land area occupied by each crop type and the stocking densities of each livestock type were derived from the Defra June Agricultural Census for 2004 for each farm type. Multiplying these ratios by the total farm areas derived the actual crop areas and stock counts on the model farms. The crop areas and stock numbers were then adjusted pro rata so that the England totals across all farm types agreed with the published census data. This accounted for the relatively small land area and livestock count found on the ‘Other’ RFT that had been discarded.
In order that the model farms had physically realistic crop rotations and livestock numbers a number of expert adjustments were made to the average farm statistics. For example, small numbers of pigs and poultry were removed from the specialist ‘Dairy’ farm and the total numbers of cattle were adjusted to achieve a typical economic stocking rate. These adjustments were necessary to convert a statistical farm definition, averaged across all surveyed farms of a type, into a simpler and more easily recognised farm definition suitable for use with the public version of the ‘User Manual’ of mitigation methods (Newell-Price et al., 2009).
This project introduced ‘Set-Aside’ to the model farm definitions to represent the 2004 baseline. This project also introduced an ‘Outdoor Pig’ farm category, and a sub-classification of the ‘Combinable Crops’ farm (‘Mixed’ and ‘Winter’) that are not explicitly represented in the Farm Business Survey (see Section 3.2).
Tables 3.2 and 3.3 summarise the total land areas and livestock numbers on the model farm systems. The farm systems were compliant with the current limits on nitrogen loading within Nitrate Vulnerable Zones.
Table 3.2 Summary of crop areas (ha) on each of the representative model farm systems.
Land Use
Dairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)
Mixed
Combinable
(Mixed)
Combinable
(Winter)
Root Crops
Indoor Pigs
Outdoor
Pigs
Poultry
Permanent Pasture
71
62
75
74.4
0
0
15
0
0
0
Rotation Pasture
24
5
16
22.2
0
0
0
0
18
0
Rough Grazing
6
79
4
4.7
5
5
2
0
3
0
Set-Aside
1
0
0
6
25
25
22
0
0
0
Winter Wheat
1
0
0
9
77
77
43
0
24
0
Winter Barley
0
0
4
10
21
21
9
0
6
0
Spring Barley
3
0
1
7.5
22.6
0
8
0
6
0
Maize
6
0
1
4.5
0
0
0
0
0
0
Sugar Beet
0
0
0
0
0
0
24.8
0
0
0
Oilseed Rape
0
0
0
8
30.7
30.7
0
0
0
0
Potatoes
0
0
0
0
0
0
18
0
0
0
Fodder Crops
1.5
0
0
1.8
0
0
0
0
0
0
Vegetables
0
0
0
0
0
0
9.8
0
0
0
Horticultural
0
0
0
0
0
0
0
0
0
0
Other Crops
0
0
0
6.3
14.5
0
28
0
0
0
Table 3.3 Summary of the livestock numbers on each of the representative model farm systems, by stock type and age.
Animal TypeDairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)Mixed
Combinable
(Mixed)
Combinable
(Winter)Root CropsIndoor Pigs
Outdoor
PigsPoultry
Dairy Cows and Heifers 1100031000000
Dairy Heifers in Calf, > 2 yrs 14000000000
Dairy Heifers in Calf, < 2yrs 14000000000
Beef Cows and Heifers 0222721000000
Beef Heifers in Calf, > 2 yrs 0323000000
Beed Heifers in Calf, < 2 yrs 0112000000
Bulls 1111000000
Other Cattle, > 2 yrs 011145000000
Other Cattle, 1 to 2 yrs 31143753000000
Other Cattle, < 1 yr 45203940000000
Sheep 50358184190000000
Lambs 54339170203000000
Sows in Pig and Other Sows 000180001592940
Gilts in Pig and Barren Sows 000200071620
Gilts Not Yet in Pig 0009000133780
Boars 0002000660
Other Pigs, > 110 kg 00040003200
Other Pigs, 80 to 110 kg 0006500024700
Other Pigs, 50 to 80 kg 0009200062100
Other Pigs, 20 to 50 kg 00010200098300
Other Pigs, < 20 kg 0001060001,27200
Layers 0002520000014,703
Pullet 00060000004,191
Broilers 0009280000055,772
Turkeys 000642000001,379
Breeding Birds 000358000002,602
Other Poultry 000365000002,704
3.1.2 Fertiliser Practice
The quantity of nitrate and phosphate fertiliser used on each land use type on the model farms was taken from the overall use figure reported by the British Survey of Fertiliser Practice (BFSP; 2004). The application rates were adjusted to account for livestock manures if present. The type and timing of fertiliser applications were taken from a detailed analysis of BFSP returns for 2003 conducted in support of Defra project NT2605 (Chadwick et al., 2005).
3.1.3 Livestock Calendar
The proportion of time animals spent in housing, gathering yards, milking parlour or at grazing were estimated by month for each livestock type based on data in Hellesten et al. (2007), Webb et al. (2001) and the Defra Farm Practice Survey (2001). The livestock activity data for each farm type were broadly consistent with the default assumptions in the National Ammonia Emissions inventory and the NARSES modelling system (Webb and Misselbrook, 2004). When grazing, animals were located on either the improved grassland or rough grazing. The grazing livestock calendar took account of the grass-cutting regime.
3.1.4 Manure Management
The quantity and nutrient content of excreta produced by each animal type was calculated according to Cottrill and Smith (2009; see also Table 3.10 below). The proportion of excreta that was managed as slurry or farmyard manure was defined based on animal type and national survey statistics on animal activity. Estimates were made of the volume of dirty water generated on the hard-standings, and the dilution of slurry in open stores. The timing and location of manure spreading to land was based on Smith et al. (2000; 2001a; 2001b). The model farm descriptions provided an explicit calendar of the location and amount of each type of animal manure spread to each crop type. The method of manure spreading and delay to incorporation (if applicable) were also provided in the farm descriptions.
3.2 Agricultural Census Data
Defra agricultural census data were used to scale the results of the model farm pollutant loss calculations by multiplication against the derived export coefficients. The total areas of agricultural land and stock numbers in each 10 by 10 km2 grid cell for 2004 and 2020 were obtained from the database prepared by the ‘Business as Usual III’ project (Shepherd et al., 2007).
3.2.1 Farm Type Disaggregation
The ‘Business as Usual III’ project (Shepherd et al., 2007) previously mapped the complete agricultural census for England and Wales at a 10 by 10 km2 spatial resolution. The procedure involved the explicit separation of the land areas and livestock numbers between the nine ‘Robust Farm Types’ used by the Farm Business Survey. The total land areas and livestock numbers assigned to each ‘Robust Farm Type’ agreed with the national June Agricultural Census statistics for 2004, wherein every farm holding had been assigned to a type. The census data were summarised on the same spatial grid and to the same crop and livestock categories used by the National Ammonia Reduction Strategy Evaluation System (NARSES) mass flow model (Webb and Misselbrook, 2004).
In order to calculate the cost and effect of potential pollution mitigation methods at national scale, it was necessary to create new versions of the ‘Business as Usual III’ national agricultural census datasets, in which the total land areas and livestock numbers were distributed between the new farm system definitions used in the revised ‘User Manual’ of mitigation methods (Newell-Price et al., 2009).
This was done by statistically re-distributing the total land areas and livestock numbers within each 10 by 10 km2 grid cell between the new farm systems. This was done in proportion to the relative number of farms of each type within a cell, and in proportion to the relative values of the agricultural census categories between each farm type. This procedure can be mathematically expressed as:
å
=
=
×
×
×
=
n
j
j
j
j
i
j
j
i
i
j
i
N
B
N
B
A
C
1
,
,
,
(EQ 3.1)
where Ai is the total value of census category i within the grid cell; Bi,j is the reference average value of census category i on farm type j (Tables 3.2 and 3.3); and Nj is the number of farms of farm type j within the grid cell.
The number of farms Nj of each type within a grid cell were estimated from the count of ‘Robust Farm Types’ derived from ‘Super Output Areas’ and ‘Small Area Statistics’ published by Defra and the Welsh Assembly Government. These farm type counts were necessarily corrected for the ratio of the ‘User Manual All’ average farm sizes and national average farm sizes derived from agricultural census statistics (Table 3.4).
The pig farms within each grid cell were re-distributed between the ‘Indoor Pig’ and ‘Outdoor Pig’ types in proportion to the ratio of the numbers of breeding sows and all pigs. This re-distribution was calibrated so that 36% of all breeding sows and 22% of all gilts were located on ‘Outdoor Pig’ farms in accordance with evidence from the national Defra Farm Practice Survey (2006). Overall, only 2% of all pigs were located on ‘Outdoor Pig’ farms. The effect of this calculation was that a high proportion of pig farms were of the ‘Outdoor Pig’ type in areas of light soils, as expected from industry practice and the requirement for good soil drainage (Figure 3.1).
Table 3.4 Average total agricultural land area (including rough grazing) by farm system type, defined by the revised ‘User Manual’ of mitigation methods and derived from the complete June Agricultural Census (Defra, 2004).
Land Area (ha)
Farm Type
User Manual All
Agricultural Census (2004)
England
Wales
Dairy
114
84
78
Less Favoured Area - Grazing Livestock
146
132
90
Lowland - Grazing Livestock
101
33
41
Mixed
154
97
61
Combinable Crops
177
123
82
Roots Combinable
180
140
75
Indoor Pig
0*
11
7
Outdoor Pig
57
11
7
Poultry
0*
6
7
Horticulture
18
18
4
*Average values of 24 ha for Poultry and 32 ha for Indoor Pig were used,
derived from the Farm Business Survey for England (Defra, 2004).
Percentage
Outdoor Pigs
0 - 10
11 - 20
21 - 30
31 - 60
61 - 100
Figure 3.1 Estimated percentage of pigs located on the ‘Outdoor Pig’ farm type.
By following these calculations, a new spatial dataset was created that described the total land areas and livestock numbers for each farm system type on a 10 by 10 km2 grid. The national total agricultural land areas and livestock counts were equal to the original values published by Shepherd et al. (2007) for all categories.
The calculations were also repeated using the ‘Business as Usual III’ projected agricultural census dataset for 2020, to obtain a revised forecast using the new farm types (see Section 3.3). The total land areas and livestock numbers mapped to each farm type in 2004 and 2020 are summarised by Tables 3.5 and 3.6.
As a consequence of the mapping methodology, the national and local ratios of the land areas and stock numbers do not necessarily agree with the values for the ‘User Manual’ reference farm system definitions. However, there is a satisfactorily close match that cannot be improved upon. It is also argued that the new datasets reflect the local variability of practice within a single farm system type. The ‘User Manual’ farm system types were used to provide data on typical practices, such as the timing and quantities of fertiliser applied, whilst the new census datasets were used for the calculation of national pollutant losses.
The reference combinable cropping farm was divided into two sub-types, with and without spring crops. The land area was preferentially allocated to the ‘Winter Cereal Only’ sub-type in areas where the NSRI Soil Workability Class was in the range 3 to 6, representing soils that are difficult to cultivate in spring. The latter was calculated from data on field capacity days, soil texture and soil wetness class (Defra project CET0501). Overall, 48% of the land area on the combinable crop farm type in England was allocated to the winter sub-type.
At national scale, the manure generated by the specialist ‘Indoor Pig’ farm type was assumed to be distributed to the land on the combinable farm types within each grid cell. The manure generated by the specialist ‘Poultry’ farm was distributed to the land on the ‘Root’ farm type. Where a farm of the correct type was not co-located with the housed animal farms, the manures produced were re-distributed across all farms of the correct target type, regardless of distance travelled. This affected less than 1% of pig and poultry manure production, so the general pattern of manure production should not be affected.
Table 3.5 Summary of total agricultural land area and livestock numbers for 2004 by farm type and River Basin District.
Land Area (ha) Animal Numbers (hd)
BasinArableGrassRoughPigPoultrySheepDairyBeefOther
Anglian1,586,801293,87439,1451,402,57542,813,022916,03560,12679,709243,153
Dee14,224112,70339,35616,0991,558,886879,37565,80820,24679,709
Humber916,738623,793266,0241,599,25625,051,2862,884,241313,231162,285641,572
North West97,318452,619255,974135,7368,173,1952,427,605298,57477,288357,929
Northumbria159,505217,553204,07492,1912,836,2251,828,27129,46180,816178,874
Severn474,129939,230136,530281,21427,213,8336,539,755316,070207,405674,664
Solway Tweed45,849172,16997,60228,3032,262,034978,54377,63245,099153,597
South East279,450201,78323,72499,4716,303,985728,02564,79638,368119,615
South West363,890854,564119,891353,57314,584,0952,772,804464,382184,290779,898
Thames499,276289,34032,382243,0439,563,855788,48279,86358,689187,022
Western Wales40,327651,559230,19710,5654,620,3684,852,345235,066138,162459,120
Total4,477,5074,809,1871,444,9014,262,027144,980,78425,595,4832,005,0091,092,3573,875,154
Table 3.6 Summary of total agricultural land area and livestock numbers for 2020 by farm type and River Basin District.
Land Area (ha)
Animal Numbers (hd)
Basin
Arable
Grass
Rough
Pig
Poultry
Sheep
Dairy
Beef
Other
Anglian
1,521,231
315,624
39,309
1,366,293
45,821,983
872,242
52,911
71,823
217,628
Dee
13,965
112,737
39,357
15,155
1,653,633
822,780
52,994
18,572
72,594
Humber
881,317
636,014
264,064
1,552,011
26,811,718
2,775,632
275,633
145,866
574,584
North West
92,890
454,007
253,894
130,984
8,747,544
2,352,211
262,645
69,463
321,316
Northumbria
152,735
220,713
202,354
87,325
3,035,347
1,764,699
25,922
72,639
160,045
Severn
454,125
945,951
136,599
267,791
29,069,870
6,158,411
271,900
188,184
608,092
Solway Tweed
44,122
172,809
96,797
26,643
2,421,190
944,804
68,313
40,505
137,709
South East
266,351
207,123
23,783
95,383
6,746,357
692,434
57,021
34,466
107,013
South West
347,447
860,385
119,344
337,868
15,608,532
2,662,328
408,650
165,424
698,659
Thames
475,729
300,014
32,466
232,368
10,234,987
750,172
70,280
52,777
167,322
Western Wales
39,250
651,762
230,192
9,825
4,876,926
4,546,080
177,199
127,015
420,000
Total
4,289,161
4,877,139
1,438,159
4,121,646
155,028,088
24,341,793
1,723,467
986,733
3,484,962
3.3 Business as Usual Projection
The projections of land use and stock numbers for 2020 were taken from the ‘Business as Usual III’ project (Shepherd et al., 2007). These projections were made by farm type based on a review of industry trends and economic modelling of structural change. The projections assumed that set-aside would be phased out by 2020. As a consequence, there was a large increase in the area of winter cereals, and of maize on farm types with grazing animals.
Across the whole of England and Wales, Dairy and beef cattle were forecast to decline by 14% and 10% respectively. Sheep numbers declined by 5%. Pig numbers were forecast to decline by 3% and poultry numbers to increase by 7%. Set-aside declined to zero, whilst the area of winter wheat increased by 13% and of oilseed rape by 14%. Tables 3.7 and 3.8 summarise the estimated total crop areas and livestock numbers in 2004 and 2020 by farm type for England and Wales.
Table 3.7.a Summary of national crop areas (ha) on each of the representative model farm systems, for the present day (year 2004) scenario.
Land Use
Dairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)
Mixed
Combinable
(Mixed)
Combinable
(Winter)
Root Crops
Indoor Pigs
Outdoor
Pigs
Poultry
Permanent Pasture
880,963
1,296,602
1,121,320
601,523
0
0
96,612
0
0
0
Rotation Pasture
280,031
84,588
228,557
182,807
0
0
0
0
5,922
0
Rough Grazing
53,380
1,214,109
52,038
32,768
39,124
35,793
16,548
0
1,141
0
Set-Aside
6,257
0
0
31,174
152,983
186,346
103,804
0
242
0
Winter Wheat
8,803
0
0
61,955
656,396
805,160
341,001
0
6,518
0
Winter Barley
0
0
44,353
66,066
140,825
151,785
60,890
0
1,682
0
Spring Barley
30,121
0
11,568
49,199
152,367
0
61,474
0
1,884
0
Maize
63,287
0
14,883
35,531
0
0
0
0
0
0
Sugar Beet
0
0
0
0
0
0
153,716
0
0
0
Oilseed Rape
0
0
0
36,967
196,360
225,143
0
0
0
0
Potatoes
0
0
0
0
0
0
113,408
0
0
0
Fodder Crops
17,546
0
0
19,168
0
0
0
0
0
0
Vegetables
0
0
0
0
0
0
74,237
0
0
0
Horticultural
0
0
0
0
0
0
0
0
0
0
Other Crops
0
0
0
38,956
137,014
0
136,223
0
0
0
Table 3.7.b Summary of national livestock numbers on each of the representative model farm systems, by stock type and age, for the present day (year 2004) scenario.
Animal Type
Dairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)
Mixed
Combinable
(Mixed)
Combinable
(Winter)
Root Crops
Indoor Pigs
Outdoor
Pigs
Poultry
Dairy Cows and Heifers
1,419,709
0
0
231,833
0
0
0
0
0
0
Dairy Heifers in Calf, > 2 yrs
180,126
0
0
0
0
0
0
0
0
0
Dairy Heifers in Calf, < 2yrs
173,341
0
0
0
0
0
0
0
0
0
Beef Cows and Heifers
0
399,301
374,885
172,953
0
0
0
0
0
0
Beef Heifers in Calf, > 2 yrs
0
46,298
28,412
23,387
0
0
0
0
0
0
Beed Heifers in Calf, < 2 yrs
0
15,976
14,856
16,290
0
0
0
0
0
0
Bulls
14,980
18,709
18,292
10,635
0
0
0
0
0
0
Other Cattle, > 2 yrs
0
116,869
231,168
45,380
0
0
0
0
0
0
Other Cattle, 1 to 2 yrs
409,882
271,680
531,585
418,002
0
0
0
0
0
0
Other Cattle, < 1 yr
555,864
359,539
548,174
324,395
0
0
0
0
0
0
Sheep
562,100
8,396,781
2,733,764
1,570,001
0
0
0
0
0
0
Lambs
578,268
7,708,289
2,414,723
1,631,557
0
0
0
0
0
0
Sows in Pig and Other Sows
0
0
0
101,207
0
0
0
83,129
105,080
0
Gilts in Pig and Barren Sows
0
0
0
11,991
0
0
0
32,736
17,620
0
Gilts Not Yet in Pig
0
0
0
48,681
0
0
0
64,909
25,520
0
Boars
0
0
0
12,160
0
0
0
3,789
2,539
0
Other Pigs, > 110 kg
0
0
0
20,149
0
0
0
19,192
0
0
Other Pigs, 80 to 110 kg
0
0
0
358,837
0
0
0
189,500
0
0
Other Pigs, 50 to 80 kg
0
0
0
477,556
0
0
0
402,304
0
0
Other Pigs, 20 to 50 kg
0
0
0
514,152
0
0
0
565,559
0
0
Other Pigs, < 20 kg
0
0
0
541,115
0
0
0
664,303
0
0
Layers
0
0
0
2,046,566
0
0
0
0
0
22,430,640
Pullet
0
0
0
398,467
0
0
0
0
0
6,397,218
Broilers
0
0
0
6,574,407
0
0
0
0
0
87,182,441
Turkeys
0
0
0
4,343,663
0
0
0
0
0
2,535,262
Breeding Birds
0
0
0
2,329,809
0
0
0
0
0
4,196,993
Other Poultry
0
0
0
2,393,628
0
0
0
0
0
4,151,691
Table 3.8.a Summary of national crop areas (ha) on each of the representative model farm systems, for the projected (year 2020) scenario.
Land Use
Dairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)
Mixed
Combinable
(Mixed)
Combinable
(Winter)
Root Crops
Indoor Pigs
Outdoor
Pigs
Poultry
Permanent Pasture
889,610
1,296,602
1,121,320
601,523
0
0
96,612
0
0
0
Rotation Pasture
281,773
84,588
228,557
182,807
0
0
0
0
5,922
0
Rough Grazing
53,324
1,214,109
52,038
32,768
39,124
35,793
16,548
0
1,141
0
Set-Aside
0
0
0
31,174
152,983
186,346
103,804
0
242
0
Winter Wheat
9,780
0
0
61,955
656,396
805,160
341,001
0
6,518
0
Winter Barley
0
0
44,353
66,066
140,825
151,785
60,890
0
1,682
0
Spring Barley
29,526
0
11,568
49,199
152,367
0
61,474
0
1,884
0
Maize
71,275
0
14,883
35,531
0
0
0
0
0
0
Sugar Beet
0
0
0
0
0
0
153,716
0
0
0
Oilseed Rape
0
0
0
36,967
196,360
225,143
0
0
0
0
Potatoes
0
0
0
0
0
0
113,408
0
0
0
Fodder Crops
17,583
0
0
19,168
0
0
0
0
0
0
Vegetables
0
0
0
0
0
0
74,237
0
0
0
Horticultural
0
0
0
0
0
0
0
0
0
0
Other Crops
0
0
0
38,956
137,014
0
136,223
0
0
0
Table 3.8.b Summary of national livestock numbers on each of the representative model farm systems, by stock type and age, for the projected (year 2020) scenario.
Animal TypeDairy
Cattle and Sheep (Less
Favoured Area)
Cattle and Sheep
(Lowland)Mixed
Combinable
(Mixed)
Combinable
(Winter)Root CropsIndoor Pigs
Outdoor
PigsPoultry
Dairy Cows and Heifers 1,216,56400231,833000000
Dairy Heifers in Calf, > 2 yrs 154,694000000000
Dairy Heifers in Calf, < 2yrs 149,513000000000
Beef Cows and Heifers 0399,301374,885172,953000000
Beef Heifers in Calf, > 2 yrs 046,29828,41223,387000000
Beed Heifers in Calf, < 2 yrs 015,97614,85616,290000000
Bulls 13,47518,70918,29210,635000000
Other Cattle, > 2 yrs 0116,869231,16845,380000000
Other Cattle, 1 to 2 yrs 368,871271,680531,585418,002000000
Other Cattle, < 1 yr 500,002359,539548,174324,395000000
Sheep 545,9238,396,7812,733,7641,570,001000000
Lambs 542,5407,708,2892,414,7231,631,557000000
Sows in Pig and Other Sows 000101,20700083,129105,0800
Gilts in Pig and Barren Sows 00011,99100032,73617,6200
Gilts Not Yet in Pig 00048,68100064,90925,5200
Boars 00012,1600003,7892,5390
Other Pigs, > 110 kg 00020,14900019,19200
Other Pigs, 80 to 110 kg 000358,837000189,50000
Other Pigs, 50 to 80 kg 000477,556000402,30400
Other Pigs, 20 to 50 kg 000514,152000565,55900
Other Pigs, < 20 kg 000541,115000664,30300
Layers 0002,046,5660000022,430,640
Pullet 000398,467000006,397,218
Broilers 0006,574,4070000087,182,441
Turkeys 0004,343,663000002,535,262
Breeding Birds 0002,329,809000004,196,993
Other Poultry 0002,393,628000004,151,691
3.3.1 Fertiliser Use
Historical trends in national average fertiliser use and crops were reviewed by ADAS under Defra project SFF0601 (Temple, pers. comm.; Shepherd et al., 2007). Factors taken account in assessing future usage included a) changing fertiliser and yield prices; b) improved accounting for the phosphorus content of managed manure; and c) then balance of phosphorus inputs and off-take. For the majority of crops there was no forecast change in application rates for 2020. The exceptions were spring barley; potatoes; sugar beet; and temporary grassland (Table 3.9).
Table 3.9 Forecast change in the total fertiliser application rate (kg ha-1 N and P2O5) made to selected crops for 2020.
Crop Type
Change in Rate
P2O5
N
Spring Barley
+3
0
Sugar Beet
+4
0
Temporary Grass
-7
-28
Potatoes
-25
-17
3.3.2 Excreta Production
The present day quantity and nutrient content of excreta produced by livestock for the model farm systems were defined by Newell-Price et al. (2009), based on the work of Cotrill and Smith (2009). The data were also based on the latest estimates prepared for the 2008 Nitrate Vulnerable Zones guidance documents. However, these data also assumed that a proportion of the pig and poultry diets contained phytase, and thus phosphorus excretion for the 2004 baseline took into account partial implementation of dietary methods.
Therefore, for calculation of the change in national excreta pollutant load between 2004 and 2020, the excreta values reported by Shepherd et al. (2007) were used and were adjusted to represent a baseline of no phytase use (Table 3.10).
Table 3.10 Undiluted excreta volume and total nutrient content of animal excreta, excluding the effect of any phytase supplements, for 2004 and 2020.
Livestock Type
Volume (m
3
yr
-1
)
2004202020042020
1 Dairy Cows and Heifers 19.35117126.744.147.8
2 Dairy Heifers in Calf, 2 Years and Over 14.6676727.727.7
3 Dairy Heifers in Calf, Less than 2 Years 14.6323213.213.2
4 Beef Cows and Heifers 16.43797926.826.8
5 Beef Heifers in Calf, 2 Years and Over 14.6565618.218.2
6 Beef Heifers in Calf, Less than 2 Years 14.6565618.218.2
Bulls 9.4954.554.521.921.9
9 Other Cattle, 2 Years and Over 11.68565618.218.2
10 Other Cattle, 1 to 2 Years 9.49565618.218.2
11 Other Cattle, Less than 1 Year 6.21383818.418.4
12 Sheep 1.8310103.43.4
13 Lambs Less than 1 Year 0.730.60.60.40.4
14 Sows in Pig and Other Sows 3.9820.118.116.114.5
15 Gilts in Pig and Barren Sows 2.0415.515.51010
16 Gilts Not Yet in Pig 2.0415.515.51010
17 Boars 3.182521.811.810.3
18 Other Pigs, 110kg and Over 1.8617.615.48.67.5
19 Other Pigs, 80 to 110kg 1.8617.615.48.67.5
20 Other Pigs, 50 to 80kg 1.3515.213.37.36.4
21 Other Pigs, 20 to 50kg 1.3510.28.94.84.2
22 Other Pigs, Under 20kg 0.