SID 5 Research Project Final Report -...

SID 5 (Rev. 3/06) of 25

General enquiries on this form should be made to: Defra, Science Directorate, Management Support and Finance Team, Telephone No. 020 7238 1612 E-mail: [email protected]

SID 5 Research 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 IS0214

2. Project title

New integrated dairy production systems: specification, practical feasibility and ways of implementation

3. Contractor

organisation(s) Institute of Grassland and Environmental research Plas Gogerddan Aberystwyth Ceredigion SY23 3EB

4. Total Defra project costs £ £446,603

(agreed fixed price)

5. Project: start date ................ 01 April 2004 end date ................. 31 March 2007

SID 5 (Rev. 3/06) of 25

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so....................................................................................YES NO

(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

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.New approaches and modifications of existing approaches are required to make UK dairy farming more environmentally and economically sustainable. This project was undertaken in order to identify new system specifications that are effective in ensuring greatly reduced environmental impact while maintaining farm viability. The implementation of management changes required to evolve the new set of sustainable systems should be technically feasible and practicable over a reasonably short time period. First the indicators of the economic and environmental sustainability of UK dairy systems were identified and their modes of application to improve system management researched. The performance of current systems was quantitatively evaluated against these criteria:- almost all were environmentally unsustainable in some respects. Next, the relative effects of different inputs, managements and system component changes on sustainability were evaluated. As previously determined, the sustainability of dairy systems was found to be most sensitive to the level of nitrogen (N) input and the efficiency with which it is utilised by the plant and the animal. It was found that whether an improvement to such efficiency would be taken advantage of by the farmer to reduce N inputs, or to boost production at the same level of N use is of great importance. Only the former choice would result in improved sustainability. Existing models and studies were used to develop an integrated mathematical simulation modelling framework – SIMSDAIRY This framework is capable of describing all of the complexity of the interacting nutrient transformations within the soil, plant and animal components of the system to calculate levels of losses to water [nitrate (NO3

-) and phosphorus (P)] and the air [greenhouse gases (GHG) and ammonia (NH3)], in response to nutrient inputs, farm managements (stock, swards and manure) and specific site conditions (climate and soil). At the same time it calculates the levels of production and farm profitability within its economic module so that pollution and farm viability can be explored within the same system. In doing this SIMSDAIRY has the capability to search for and specify optimal conditions for satisfying both sets of criteria. Additionally, SIMSDAIRY has the unique capability to integrate the effects of managements on some of the more socio-economic strands of sustainability that are not yet amenable to most modelling systems and to quantify how changes to these criteria might affect profitability. These include animal welfare, food quality for human health and product saleability, landscape aesthetics and biodiversity. A major modelling exercise was carried out of over 10,000 runs in order to identify and specify what combination of management techniques and measures would be necessary to render UK dairy systems fully sustainable. It was discovered that the order in which these individual changes are introduced determines whether or not and how quickly current systems can become sustainable. The most effective ordering (or trajectory) towards sustainability may not be the most practical to introduce by the farmer,

SID 5 (Rev. 3/06) of 25

however. SIMSDAIRY was thus used to demonstrate how the 3 most common UK dairy system typologies (‘conventional-180 day grazing’, ‘extended grazing’ and ‘fully housed’) could become sustainable by stepwise introduction of new managements and inputs along specified trajectories. There was great variation in the ease (and cost) with which different criteria of sustainability could be achieved, depending on specific location (soil and climate), degree of intensification and farm size. The important result of this exercise is that full environmental sustainability was found to be technically feasible, but this would result in economic penalties ranging from minor to severe. Dairy Industry stakeholders were invited to a workshop and asked to comment on the scope, performance and outputs of the new modelling framework. The general consensus was that SIMSDAIRY adequately and realistically described the 3 typologies, although some specific input values were disputed, and that most of the improvements could be implemented over a 5-10 year period, given the appropriate financial incentives through Policy. The stakeholders considered that it was important for Defra to consider the large effect of site conditions on ease of achieving sustainability, when developing Policy and that SIMSDAIRY should be up-scaled to enable catchment and/or regional simulations. It was generally thought that a simplified version of the model, or a suite of common pre-run scenarios, should be developed for use by farmers and their advisors, advocating that the full system be used for Policy development studies. Nevertheless, it was thought that the existing framework should be validated and used on real farms in the near future.

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).

SID 5 (Rev. 3/06) of 25

RATIONALE The UK dairy sector has faced severe, multiple challenges over the last couple of decades (Curry Commission, 2002; Defra, 2002). The policy and economic drivers, which for some 60 years until 1994 shaped UK dairying through the operation of the Milk Marketing Boards have been replaced, inter alia, by a combination of market and policy drivers of change and development in dairying. For the market, the signals transmitted through the food chain encompass a range of ‘quality’ requirements (milk composition, milk hygiene, dairy and herd management issues, infra-structural requirements, seasonality, production system, etc.) and comprise both contractual (via a variety of usually bespoke QA schemes) and financial (via the determinants of milk price). Meanwhile, policy drivers of the dairy sector comprise an increasing range of environmental and animal welfare goals, which may or may not be broadly consistent with some of the ‘quality’ criteria set by the market. Dairying systems in particular are faced with the need for a comprehensive ‘re-engineering’ in order to meet market, animal welfare and environmental targets in the context of government policy for an internationally competitive industry. All this has to be achieved within an essentially market-facing business environment while maintaining, or even enhancing, broader socio-economic sustainability. Thus, integral to the new policy context is a requirement to both sustain and enhance the rural environment whilst promoting and developing the rural economy. More sustainable systems, based on integrated farming principles that can be implemented in a cost-effective manner, are seen as the route towards reconciling these economic and environmental pressures and this is the context of the research reported here. It should be noted, however, that it is not self-evident that the combination of market-led production, increasing economic efficiency and an appropriate regulatory framework are actually reconcilable and can deliver this ideal (Winter, 2002). Meanwhile persistent price pressures until very recently have driven many milk producers to the margins of economic viability (Colman et al., 2004; Colman and Zhuang, 2005; NFU/RABDF, 2007) and fuelled concerns about the possible abuse of market power within the UK dairy chain (MDC, 2003; Competition Commission, 2007). In recent months however there has been an increase in food commodity prices and costs around the world, as plant tissue production for food has been supplanted by production for energy acquisition. At the same time, demand for ruminant products by the developing world is increasing, due to changes in preference and population growth. These changes have had a range of impacts on the UK dairy industry, including increased costs of grains for animal feeding, but increased farm-gate milk prices. In practice, however, the identification and implementation of integrated dairy production systems require research to explore a wide range of alternatives, and to show how far the negative environmental impact of production can be reduced or eliminated without having a negative effect on the economic viability of the farm. While such research will make high demand of relevant farm data, the use of process based simulation models will be the only way possible to fully explore the economic, environmental and social performance of dairy systems that it would be feasible to develop along evolutionary trajectories from our predominant systems currently operating. This is due to the extreme complexity of the biological, physical and chemical controls and their interactions on the ways dairy systems operate as well as the impacts of weather and the farmers’ management of nutrient inputs, sward growth and stock feeding. Therefore this modelling-led desk-study project was undertaken to specify new sustainable dairy systems for operating under UK conditions and to investigate the feasibility and likely impacts of their implementation. First, the objective criteria by which sustainability is defined had to be obtained through literature review and utilising the results of previous relevant projects. Next, a novel modelling framework was constructed to enable current systems to be specified and the new systems to be identified. Thirdly, the new systems were evaluated by stakeholders and ease of implementation judged. Lastly, the work was disseminated through a range of knowledge transfer methods. DEFINITION OF CRITERIA OF SUSTAINABILITY The criteria used for defining the sustainability of UK dairy systems were identified through literature search, reference to previous relevant projects and by access to unpublished data and model outputs archived with the project participants. The sources of information and the tasks of the various project partners are detailed in the SID3 documentation. Economic criteria were reviewed by University of Exeter and IGER and environmental criteria were reviewed by IGER, ADAS and Wageningen University. There have been many attempts to define sets of indicators of sustainability of agricultural systems, each differing in emphasis, degree of quantification and functional coherence. Perhaps the most basic is that set out in the DEFRA Sustainable Food and Farming Strategy, which identified Headline Indicators for each of the 3 Pillars of Sustainability – economic, environmental and social sustainability and a set of Core Indicators under each Headline. Progress towards this long-term vision, which is underpinned by a strategy agreed with the Sustainable Development Commission, is being monitored through the use of these indicators. There are 11 Headline indicators, which focus on broad measures of progress; and some 50 Core indicators, some of which extend the scope of the Headline indicators while others provide additional detail in assessing the speed and nature of change. Table 1 shows those indicators relevant to defining more integrated dairy systems.

SID 5 (Rev. 3/06) of 25

Table 1. A simple classification of some Indicators of sustainability

Table 1 however embodies little information about how the different indicators relate to each other in the context of a land-use system. Neither has there been any attempt to quantify the indicators in terms of what are currently considered as threshold values. Both of these features are essential prerequisites for the present study: - each indicator must be a variable or parameter that can be directly or indirectly considered by the model used to specify the new dairy systems and the model itself should be able to simulate processes that identify and quantify the important interactions between indicators. A set of criteria (indicators) functionally related to the major system controls were defined by Van Calker et al. (2005). The basic design of the scheme is given in Fig. 1. Here again there was no attempt made to quantify the indicators of sustainability. Nor is there any information on how indicators in different limbs of the scheme might interact. Both of these criteria are important if the chosen indicator(s) might be used to compare between farms or eventually form the basis of legislation. Fig. 2 shows a hierarchy of indicators identified by Schroder et al. (2004) capable of being used as the basis to assess (and control by legislation) the impact of nutrient loss from agriculture on water quality. This represents an advance on the previous schemes in that it represents a logical flow of nutrients around a livestock system and thus indicates functional relationships between different indicators. It also leads one to consider the relative values of the most effective indicators, close to the point of legislation (towards to top) and the most attributable and responsive indicators, close to nutrient management (towards the bottom) for a range of purposes. It also makes the distinction between ‘single measurement’ and integrated indicators and suggests that integrated indicators, while enabling more aspects of the system to be accounted for, are generally more uncertain in whether or not they are a means to achieve a desired goal.

Strategic outcome

Headline indicators Core indicators

Farming sector focused on market

Greater value added per head

Farm incomes

Value-added activities Economic sustainability

Commodity yields

Farm assurance schemes Organic action plan Skills and training Financial risks Reduced

environmental cost of food chain

River water quality

GHG Air quality Pesticide use Pollution incidents Waste Good farm practice Environmental sustainability

Energy use

Food miles Better use of natural

resources Soil strategy

Use of water for irrigation Improved landscape and

biodiversity Species and biodiversity

Habitats Landscape value Access to countryside Higher animal

welfare Animal welfare indicator Animal health and welfare

strategy indicators Social sustainability

Reduce the gap in productivity

Rural economy

Rural business Labour

SID 5 (Rev. 3/06) of 25

Figure 1. Classification of factors (attributes, aspects and controls) affecting the overall sustainability of dairy farms (after Van Calker et al., 2006) Composite (integrated) and single measurement indicators comprising nutrient levels in different parts of the livestock system have been proposed (Jarvis, 1999; Goodlass et al., 2003) and critically examined (Scholefield et al., 2007) for their use in defining sustainability. These kinds of indicators (see list in Table 2) appear extremely attractive in that they are capable of integrating over at least two of the pillars of sustainability (e. g. N loss to the environment/1000l of milk produced), are simple to understand and can be used to compare between farms and management regimes. Others, like the farm-gate nutrient balance (N and/or P) are easier to measure or calculate accurately and could be used to compare both within and between farms as the basis for legislation, as with the Dutch MINAS system (see Ondersteijn et al., 2002). Schroder et al. (2003) reviewed the application of nutrient balance approaches for improving the efficiency of nutrient use and concluded that balances do not in themselves reveal the nature or magnitude of nutrient losses to the environment, nor do they provide the information required to reduce those losses. Table 2. Some indicators of economic and environmental performance of dairy production systems which may be used to assess sustainability (from information in Jarvis, 1999; Defra project NT1854) Economic Concentrate use kg/litre Milk £/cow Milk £/hectare Milk litres /cow Milk litres/ hectare Dairy LU /hectare Margin over Concentrates £/cow Margin over Concentrates £/herd Gross margin £/cow Gross margin £/herd Calving pattern (code) Labour input - total hours Total N in milk (also P) N in milk per cow N in milk per ha N Yield / ha Total N yield Total N harvested

Environmental Grass fertiliser N/ha (also P fert) Total feed N per cow Total dairy enterprise N (kg) (inputs) NO3 leach grass / ha Denitrification / ha NH3 Volatilisation / ha N2O from denitrification / ha N2O from nitrification / ha NO / ha N losses grass / ha Total grass N losses Total farm N losses Total farm env. losses Farm gate P surplus Farm gate N surplus Losses N grazing / ha Losses N cutting / ha Environmental losses N grazing / ha Environmental losses N cutting / ha Total environmental losses Total NH3 losses from stable

Measurement and/or calculation of sets of indicators for similar farm types in a particular area or region tend to show wide ranges of values (e. g. Jarvis, 1999; Scholefield et al., 2007) and this has lead to the idea that the worst farm can be changed to be like the best farm through improved management. This assumes that these differences in indicator values are not merely due to immutable site factors (soil, drainage, climate) and/or special

SID 5 (Rev. 3/06) of 25

socio-economic circumstances. Another complication with interpreting the use of indicators to assess sustainability is whether a particular environmental or economic (or integrated) indicator is judged either on a per area or on a per unit of output or input basis. For example, highly intensive dairy farms tend to be more polluting on a per ha basis than moderately intensive farms, but less polluting on a per unit of milk output basis. Resolution of this complication takes on even more importance when considering the units by which any aspect of sustainability is defined and legislated for. For example, should we apply water quality legislation to limit NO3

- in water on a field, farm, hydrological catchment or administrative regional basis? Each one will have different consequences for how a relevant indicator value is interpreted and with the ease of compliance with the legislation that a farmer can achieve. With the present study we selected a range of ‘single measurement’ indicators of sustainability given in Table 3. These were chosen to be at or close to the ‘points of assessment’ in the system, or towards the top of the scheme shown in Fig. 2 and therefore more accurate and effective, than those far removed from these points. We then relied upon the model or computer simulation of the nutrient flows within and out of the system to relate and reconcile those indicators of the different pillars of sustainability and to indicate how systems could be modified to improve them. These are, of course, the overriding advantages of using a modelling approach to defining the sustainability of livestock systems, relative to an indicators-only approach.

Figure 2. Hierarchy of indicators to assess impact of nutrient loss from agriculture on water quality (taken from Schroder et al., 2004)

SID 5 (Rev. 3/06) of 25

Each criterion was defined according to ranges of values considered to be compatible with the notion of sustainability. In some cases such values may be fixed as those required for strict compliance with legislation, such as for example 11.3 mg N l-1 concentration of N in drainage water, as the limit on NO3

- leaching. In most cases however, this is not possible [for example for P, nitrous oxide (N2O), methane (CH4) and NH3 emissions] and so we have assumed values which correspond to those at the low ends of ranges measured and/or modelled for economically viable dairy systems. Similarly, for economic sustainability, we have assumed a profit margin for a given milk price, which is considered to be a limit below which the average farm is considered non-viable. For the other attributes of sustainability such as biodiversity, landscape aesthetics, animal health and welfare, we have constructed semi-quantitative score matrices, using approaches described in the recent literature where relevant, which enable these attributes to interact with the quantitative model structures, so that they can act both as driving and responsive variables. Table 3. Sustainability targets Sustainability Attributes Environmental Climate change Eutrophication Acidification

thresholds manure limitations

GHG NO3

-

(mg/L) P (µg/L) if NVZ NH3 NOx -12.50% 11.3 100 -15% -40% Economic 4 pence/L profit Biodiversity 4 score Milk quality 4 score Soil quality 4 score Landscape 4 score Animal welfare 4 score

Table 3 shows the set of desirable sustainable targets for the predicted farms. Targets associated with impacts on water quality were set to comply with Nitrates Directive (<11.3 mg l-1 in the leachate) and the P threshold for eutrophication (<100µg l-1). A greenhouse gas (GHG - N2O and CH4) reduction (12.5 %) from the baseline farm was proposed in order to comply with Kyoto protocol targets both individually and as a whole. Gothenburg protocol set emission ceilings for 2010 for NH3 and nitric oxide + nitrogen dioxide (NOx) (Gothenburg protocol: UNECE, 1999) in Europe. According to this protocol, a reduction from the baseline farm was proposed at 17% and 41% for NH3 and NOx, respectively. An adequate net farm income for standard living and acceptable standards (score of 4/6) of quality of milk [milk with enhanced profile of CLA (conjugated linoleic acids) and omega-3 fatty acids], animal welfare (e.g. welfare driven by an adequate moving space per cow in the stable and in the fields, less than average incidence of illnesses related to stress, bad management conditions), level of biodiversity and landscape aesthetics (e.g. this would generally imply capital investment in hedging or wild-life strips and enhancing the sense of traditional methods or natural use of the land), and soil quality (a % of total land not under pressures of treading, compaction) were also set as sustainable targets. The acceptable net farm income was assumed to be net farm income of 4 pence/L of milk. This value was arbitrarily chosen for the purposes of this modelling exercise as no data were available to support any specific value. Van Calker et al. (2005) suggested other attributes for economic sustainability such as liquidity, profitability and solvability. As all of these attributes are interrelated, profitability, through using the total net farm income, was chosen as the indicator. Ideally we should have had some information about the distribution of net farm incomes of our UK dairy farmers and this should have been related to the adequate net farm income for standard of living. Compliance with NVZ170 rules was simplified as the limitation at farm level of organic manure (including grazing) applications of 170 kg N ha-1 yr-1 (assuming that a 600 dairy cow will have an annual excretion of 106 kg N ha-1 yr-

1: Guidelines for farmers in NVZs: MAFF, 2001) and establishing closed periods for manure application (September-November for grasslands and August-November for maize). ASSESSING THE PERFORMANCE OF CURRENT SYSTEMS The range of economic performance of current dairy systems was determined, through analysis of archived data for over 300 farms. A regression equation based model was derived (Economic Dairy Management model – EDM) and used to calculate various economic indicators in £/cow as:-

(i) Total gross output (range 572-1597, mean 1085) (ii) Total variable costs (range 202-765, mean 483) (iii) Gross margin (range 190-1013, mean 602) (iv) Total fixed costs (range 223-859, mean 541) (v) Overheads (range 46-182, mean 114) (vi) Net margin after overheads (range -651-544, mean -54)

SID 5 (Rev. 3/06) of 25

[See Appendix 2 and 6 (M Turner) for more details] This information was then used as the basis for interaction of the EDM model and the rest of the SIMSDAIRY framework, which describes and quantifies the environmental performance of dairy farming. Recently, a high proportion of farms are becoming non-viable due to poor and declining milk price as indicated by the above analysis. Thus dairy farming is ‘evolving’ due to this pressure from the ‘average’ 100 cow/180day grazing herd either to extended grazing (with much reduced indoor feeding), or to larger herd size of higher yielding cows, with reduced grazing and greater indoor feeding. Environmental performance of current systems was evaluated through analysis of data and modelled outputs derived as part of the DEFRA funded ‘Indicators’ (NT1854), ‘Extended grazing’ (NT2509) and ‘Blueprints’ (IS0101) projects. Thus, N losses as NO3

-, NH3 and N2O were either measured or modelled for over 60 dairy farms operating a range of management practices and levels of N input. Then, using the methodologies developed under project NT2506 (NGAUGE simulation of NVZ compliance) the environmental performance of current systems was compared with the sustainability criteria values as derived above. Most current systems did not meet these criteria, although there were clear interaction between forms of pollution and managements (pollution swapping). The total losses of N from the farms ranged generally from 5 to 25 tonnes, with two larger farms losing 57 and 145 tonnes N. On a per hectare basis this was divided into N leaching (range 10-142 kg N), denitrification (range 6 – 170 kg N) and volatilisation of NH3 (range 30-60 kg N). There have been few recent experimental assessments of P losses from UK dairy farms, but it is generally accepted that these range between 1 and 2 kg/ha with P concentrations in drainage water frequently exceeding 200 mg/l and occasionally 2000mg/l (Haygarth et al., 1998). Losses of GHGs from dairy farms are even more difficult to evaluate experimentally and modelling-based studies are being used in conjunction with field assessments in the major Defra project AC0101 in order to specify these as a basis for updating UK GHG Inventory figures. The modelling framework SIMSDAIRY, developed and used for the present project is at present the only means of assessing these environmental effects of dairy farming, taking into account all of the complexity and variability within and between farming systems. Thus the baseline scenarios specified by this framework are those used to evaluate these losses for a range of UK systems and to evaluate whether management options used singly or in combinations can achieve sustainability. This information is given in later sections of this report. FACTORS MOST INFLUENTIAL ON SUSTAINABILITY A sensitivity analyses were performed with the NGAUGE model to assess the relative impacts on sustainability which would be caused by modification to the efficiency of N flows through the various system components. Most improvement would be caused by increasing the efficiency of plant uptake from the soil. Improving the efficiency of rumen capture would only be effective if farmers used the benefit in efficiency to reduce N input rather than increase output. The latter would result in negligible improvement to sustainability. This indicates that how a potential N-loss mitigation strategy is implemented on the farm has large consequences for its effectiveness. The ‘gearing’ of nutrient supply from the soil through mineralization to the requirements by the sward over the growing season is also found to have a substantial effect on N flows and losses for a given level of input. With high levels of mineralization the system is apparently more N efficient, with a greater proportion of input going to product. However, the ratio N in product: N lost remains the same. A more comprehensive sensitivity analysis, involving many more factors, was performed using the modelling framework SIMSDAIRY as described in the next section. STRUCTURE AND USE OF A NEW MODELLING FRAMEWORK TO IDENTIFY SUSTAINABLE DARY SYSTEMS A new modelling framework was constructed from a further development of existing models and some new modules and sub-models named Sustainable and Integrated Management Systems for Dairy Production (SIMSDAIRY). This integrates models for N (NGAUGE: Brown et al., 2005; NARSES: Webb and Misselbrook, 2004), P (PSYCHIC: Davison et al., 2008; Stromqvist et al., 2008) and farm economics (EDM: Butler and Turner, 2007), equations to simulate NH3 losses from manure application (Chambers et al., 1999), predict CH4 losses (Chadwick and Pain, 1997; Giger-Reverdin et al., 2003) and cows’ nutrient requirements [Feed into Milk (FiM) (Thomas, 2004)], with ‘score matrices’ for measuring sustainability attributes of biodiversity, landscape, product quality, soil quality and animal welfare. NGAUGE (Brown et al., 2005) is an empirically-based model which simulates monthly N flows within and between the main components of a grazed grass field according to user inputs describing site conditions and field

SID 5 (Rev. 3/06) of 25

management characteristics. Outputs of NGAUGE include a field and N fertiliser recommendation, comprising amounts of N in both production [N and dry matter (DM)] and loss components of the system [NO3

- leaching, NH3, dinitrogen (N2), N2O and NOx]. This optimisation enables user-specified targets of herbage, N loss or fertiliser use to be achieved while maximising efficiency of N use. NARSES (Webb and Misselbrook, 2004; Webb et al., 2006) is a model which estimates the magnitude, spatial distribution and time course of agricultural NH3 emissions, together with the potential applicability of abatement measures and their associated costs. Phosphorus and Sediment Yield Characterisation In Catchments (PSYCHIC) is a model which predicts the risk of diffuse pollution from a source area by estimating source, mobilisation and delivery of P and sediment: P inputs in manure and fertilisers and soil residual P, the mobilisation of P and sediment through dissolution and soil detachment and the delivery of dissolved and particulate P, and associated sediment, to watercourses in surface and subsurface runoff (Davison et al., 2008). PSYCHIC takes account of management practices as well as landscape factors and climate to predict the spatial and temporal distribution of flow, sediment and P, on a monthly time step. EDM is a newly developed empirically-based model (Butler and Turner, 2007) using data that was originally collected by Universities and Colleges under the aegis of Defra’s Commissioned Work Programme and was collated by the University of Manchester in 2003/04 (Colman et al., 2004); and recalibrated to reflect 2006 prices. This data is used to model production and cost regression equations to represent the relationships between milk production, prices, input requirements and cost functions thus providing the necessary parameters to replicate the underlying economic structures of English and Welsh dairy farms. From this model, net margins are calculated by deducting total fixed and total overhead costs from gross margin to provide a measure of economic farm performance.

SIMSDAIRY main components can be classified in 2 different hierarchical levels: modules and sub-models. Modules represent the highest level of control and comprise the code in which the main general functions of SIMSDAIRY are represented (MODGENERATOR, MODSIMULATOR and MODEVALUATOR). The sub-models carry out specific tasks within SIMSDAIRY (e.g. predict N flows within a grazed grassland field) and are either modifications of existing models or new developments (SIMSMANAGEMENT, SIMSNGAUGE, SIMSPSYCHIC, SIMSECONOMICS, SIMSSCORE and SIMSRANK).

The whole framework operates automatically and, except for SIMSPSYCHIC submodel which is an external link through a Visual Basic (VB) Dynamic Linking Library (DLL) file, has been coded into a program compiled with Borland Delphi 5. Modules MODGENERATOR generates management options to be optimised. Management can be optimised as a single option or in combination through a matrix (Man i,j). The indices i and j represents different ‘i’ management options and ‘j’ values, respectively. For any given farm, existing management factors (e.g. manure application on different land use areas) and/or genetic traits (e.g. plants with enhanced ability to absorb N) can be optimised. The whole list of optimisable factors together with the main input environmental/management (E/M) variables data can be viewed in the input screen interface of SIMSDAIRY (Fig 3).

Figure 3. Input-screen of SIMSDAIRY with the main input data E/M variables used for the simulation of nutrient flows, transformations and losses and the whole list of optimisable farm factors (in white) (all input data in Appendix 5).

SID 5 (Rev. 3/06) of 25

MODSIMULATOR main tasks are to simulate within the farm: (i) the nutrient and energy flows, the impact of these flows on sustainability issues (e.g. milk quality) and the effect on farm profitability. MODEVALUATOR ranks farms management values by best match to a user-weighted multiple criteria result [e.g. Min CO2 eq-GWP/L milk or max environmental index (Van Calker et al., 2006)] giving a level of farm milk production as a target to achieve.

Submodels SIMSMANAGEMENT is a new development which is used to enter the main data inputs, initialise E/M variables, calculate nutrient/energy flows at the herd level and link the main interactions among submodels. SIMSNGAUGE is a modification of the NGAUGE model (Brown et al., 2005) by which, DM and N plant yield and N losses (N2, N2O, NOx, NH3 and NO3

- leaching) per unit of hectare are predicted for growing grass (grazed and/or conserved) and

arable crops (maize). Modifications have been carried out to incorporate the simulation of flows of P (P losses are not calculated at this stage). SIMSPSYCHIC is a Psychic (Davison et al., 2008) metamodel which has been developed for SIMSDAIRY and includes simple functions to represent the annual total loss of elemental P (dissolved and particulate P from soil and losses from fertiliser and manure) to watercourses [see Appendix 6 (P. Davison) for more details]. . SIMSSCORE is a new submodel which simulates the effect of both nutrient management variables (e.g. effect of unsaturation of fatty acids in the diet on milk yield) and non-nutrient management variables (e.g. available surface per cow during housing) on the sustainability of the farm in terms of biodiversity, landscape, milk quality, soil quality and animal welfare. The scores assigned reflect poor (0) to very satisfactory (6) sustainability.

Figure. 4. General flowchart diagram of SIMSDAIRY.

SID 5 (Rev. 3/06) of 25

The scores themselves and the way each matrix interacts with the conventional modelling framework have been determined through expert judgement and are not based on any common underlying principles, nor on any recent validating data. It is hoped that as the value of this novel approach is realised, such principles and data can be developed and obtained. The milk quality matrix is based on the magnitude of 4 interacting factors:- content of polyunsaturated fatty acids (PUFAs) in the diet; content of CLA in the diet; degree of off-flavours in the milk; shelf life of milk; and ‘spreadability’ of products. SIMSSCORE calculates a score index of sustainability for biodiversity based on 5 management factors: (i) grazing intensity, (ii) fertiliser rate, (iii) cutting management, (iv) reseeding management and (v) the inclusion of patches for biodiversity (margins, hedges and buffer-strips). A conceptual relationship between grazing pressure and species diversity (Grime, 1979) has been modified and incorporated into an equation in SIMSSCORE. According to this equation, vegetation cannot maintain its integrity if grazing pressure is either too high or to low [see Appendix 1 and 6 (A del Prado, R. Drewhurst & A. Hopkins) for specification of relationship]. In order to score animal welfare of the farm SIMSDAIRY incorporates the prototype of clinical welfare score for dairy cattle proposed by Noordhuizen and Metz (2005), in which information about: (i) general husbandry, (ii) pasturing, (iii) housing, (iv) milk harvesting and (v) dairy cows per se needed to be input. The resulting score is weighted with a newly developed factor which takes into account: housing period, livestock density, genetic merit of the cows, proportion of silage feed of the total diet and PUFA concentration in the diet. For assessing the effects of soil quality and quality changes on sustainability 3 matrices have been defined to describe poaching susceptibility, risk of compaction and risk of erosion. All are based on combinations of model inputs for soil conditions (texture, structural stability assessment, water content) and stock management intensity. Subsequently, SIMSECONOMICS calculates the net farm margin by subtracting the total fixed costs and overheads from the gross margin. SIMSECONOMICS is a newly developed model (Butler and Turner, 2007) which calculates the net farm margin by subtracting the total fixed costs and overheads from the gross margin. Variables costs are calculated by the model as a function of management variables (e.g. £ per unit of applied manure volume). Some management strategies (e.g. those resulting in enhancing landscape), due to large cost variability, are user-proposed inputs and hence do not intend to reflect an accurate value. The farm sustainability is then evaluated by the submodel SIMSRANK, which ranks farms under different management strategies according to the user-defined criteria (e.g. minimising overall environmental impacts over unit of milk). Sensitivity analysis The influence of a number of selected E/M variables (Table 3 in Appendix 2) on state variables values (Table 4 in Appendix 2) was tested through sensitivity analysis. Two types of E/M variable were identified:- those that could be numerically accounted for in the model (e. g. fertiliser rates, housing days, starch in diet, dead plant mineralised) and those that had to be assessed via relative high-low classifications (e. g. increased dietary PUFA, poor silage making conditions, area of loam soil). The sensitivity of a number of state variables was calculated as the change in state value relative to the change in E/M variable value: Sensitivity = [(S1-S2) / Sb] / [ (P1-P2) / Pb] x 100 % In which S1 and S2 are the state variable values for the minimum (P1) and maximum (P2) E/M variable values, and Sb and Pb the state value for the basal E/M variable value. To a certain extent, most of the state variables showed some sensitivity to changes in most of the accountable and unaccountable E/M variables. However, as expected, a large degree of variability could be found on the levels of sensitivity. The results of this analysis are detailed in Appendix 2. Briefly, most of the selected state variables, except for landscape aesthetics, showed substantial sensitivity to selected E/M variables. Nutrient (pollutant) losses were particularly sensitive to most of the studied farm variables. Some of these variables are management-related ones and thereby, could be easily altered by the farmer. Others, such as soil type, show a large effect on pollutant losses from the soil (N2O, NO3

-, NOx and P), but are not readily manipulated by the farmer. SPECIFICATION OF NOVEL SYSTEMS USING SIMSDAIRY A major scenario testing exercise was carried out running SIMSDAIRY to explore how UK dairy farming systems could become sustainable. Clearly, to explore the whole of the model response surface would involve many hundreds of thousands of runs and would not be sensible to carry out: ways of constraining the extent of and focusing the exercise were required. This was done by first defining a ‘conventional’ baseline dairy system, in

SID 5 (Rev. 3/06) of 25

terms of management, site conditions and also by model outputs of all indices of sustainability. For example, quantity and timing of fertiliser applications were based on RB209 (Defra, 2000), whilst slurry application timings were based on the survey data provided by Smith et al. (2001). Then we defined a set of 13 management improvements that the model could simulate, but recognising that the sequence by which they are introduced may influence the outcomes on the indices, we planned to assess the effects of this possibility. Thus, the scenario testing was done in two parts. First the effect of the 13 improvements was evaluated by introducing them in 4 different sequences of steps (‘trajectories’). Secondly, the effects of imposing a single optimised trajectory (best sequence of management improvement steps) on 3 typical UK dairy typologies (conventional 180-day housing/grazing; extended grazing with no silage; and fully housed/high concentrate use) on the ease and feasibility of achieving sustainability was carried out. In all over 10,000 model runs were completed. We defined a typical conventional dairy farm as the baseline farm on which we would be testing the different steps comprising 4 different trajectories towards sustainability. The characteristics of the farm are given in Table 4 below and in Appendix 4. Table 4. Main characteristics of the typical dairy farm used as baseline.

Farm management Site

Milk yield (litres/ cow yr) 6570 Location Devon (UK)

Fat in milk (g/kg) 40 Soil type clay loam Protein in milk (g/kg) 34 Drainage status moderate Dairy cows (number) 88 Replacement rate (%) 27 Followers (number) 60 Calving pattern All-year Breed Holstein Silage management Average quality Housing time-Dairy cows (days/year) 185 Housing time-Followers (days/year) 160 Diet During housing grass silage, maize silage, concentrates During grazing grazed grass, maize silage, concentrates

Annual Fertiliser management grass maize

cut grazed (dairy) grazed

(followers) Fertiliser N (kg N/ha) 290 240 180 40 Fertiliser P (kg N/ha) 35 25 40 40 Manure management Type of manure slurry (60 g /kg DM*) cut-grass grazed-grass maize % of total applied to land 45% 50% 5% Storage slurry tank: open Application technique Broadcast Grassland management

cut-grass grazed-grass young grazed-grass

History Long term grassland Sward age (years) 11-20 4-6 >20 *DM: dry matter The management steps and trajectories simulated are given in Table 5. In these simulations the each successive step incorporated all previous steps. SIMSDAIRY’s optimisation procedure, for most parameters, was set to adjust the hectares needed for forage instead of adjusting the number of cows that certain reductions in plant production per hectare could trigger. This is not the case for mineral fertiliser, by which changes were obtained by adjusting the fertiliser rate and supposing the same amount of forage surface and CP plant production per hectare. SIMSDAIRY was also set up to assume that when forage area needed is smaller than that simulated in the baseline scenario, any surface that is no longer needed for production, compared with the baseline scenario hectare, is used for enhancing biodiversity. These assumptions do not intend to emulate the most probable decision made by farmers but only be used as an example of the most probable way to achieve sustainability.

SID 5 (Rev. 3/06) of 25

Table 5. Order of steps introduced and optimised by SIMSDAIRY within the 4 trajectories to improve the sustainability of the baseline dairy farm.

Trajectory Steps 1 2 3 4

1 Mineral N fertiliser distribution

Manure application method

Milk properties Manure system (FYM)

2 Diet manure properties (%DM)

Gmilk* Manure distribution

3 Manure application method

Manure distribution Gmilk* +GFERT

* Gmilk*

4 manure properties (%DM)

Manure storage Gh* Manure storage

5 Manure distribution Mineral N fertiliser distribution

Gurinedung* Milk properties

6 Manure storage Gmilk* +GFERT

* Manure storage Welfare+landscape +reseeding

7 Manure system (FYM)

Gh* Welfare+landscape

+reseeding Silage quality

8 Milk properties Welfare+landscape +reseeding

Diet Manure application method

9 Gurinedung* Diet Silage quality Gmilk

* +GFERT*

10 Gmilk* Silage quality Manure application

method Mineral N fertiliser

distribution 11 Gmilk

* +GFERT* Manure system

(FYM) manure properties

(%DM) Gh

*

12 Gh* Milk properties Mineral N fertiliser

distribution Gurinedung

*

13 Welfare+landscape+ reseeding

Gurinedung* Manure distribution Diet

*Changes at the genetic level: recovery of soil available N by the plants (Gh), and at the animal genetic level: diet N partition into milk (producing the same amount of milk) (Gmilk), animal fertility (GFERT) and urine: dung ratio (Gurinedung). Figure 5 shows a summary of the final results after introducing the 13 steps in the 4 trajectories. The following steps were introduced at the plant genetic level: (i) recovery of soil available N by the plants (Gh), and at the animal genetic level: (ii) diet N partition into milk (producing the same amount of milk) (Gmilk), (iii) fertility (GFERT) and (iv) urine: dung ratio (Gurinedung). All the trajectories improved all the environmental losses and attributes of sustainability except for that related to net farm income. The 4 trajectories resulted in a modest increase (5-12%) in surface requirements for forage; thus, implying a certain extensification to improve the sustainability of the farm. The differences were small between trajectories 1, 3 and 4. However, trajectory 2 required more forage area than the rest of the trajectories. In terms of environmental pollution per hectare (Fig 5.b), all trajectories improved the index of overall environmental sustainability by more than 100%. The scope to reduce individual losses by following any of the trajectories was different for the different forms of pollution. The trajectory 3, however, resulted in about 50% more improvement that any of the other trajectories, which, in fact, were very similar and around 120% improvement. Most of these differences were caused by the fact that trajectory 3 reduced losses of NH3 and N2O losses further than the rest of the trajectories. Trajectories 1 and 2 improved more than 10 % N and P concentrations in the leachate compared with 3 and 4. In terms of results related to attributes of sustainability for scores of quality of milk, animal welfare, level of biodiversity and landscape aesthetics, and soil quality, results changed substantially from trajectory to trajectory. Whereas soil quality, milk quality and landscape aesthetics were improved similarly for the 4 trajectories, biodiversity reduction was much larger in trajectory 1 than in the other trajectories, which resulted in similar improvements and animal welfare was greatest in trajectory 1, followed by trajectory 3 and 4 (similar to each other) and those followed by trajectory 2, which resulted in much smaller improvements than the rest of trajectories. All trajectories except for the trajectory 4, resulted in decreasing net farm income. Trajectory 1 was particularly poor in cost-effectiveness as a decrease in about 100% in net farm profit resulted after all the steps were implemented.

SID 5 (Rev. 3/06) of 25

Figure 5. Comparison between the results after completing the 4 trajectories in terms of % change compared with baseline dairy farm of: (a) ha, total volume of milk, (b) losses per ha of CH4, N2O, NH3, global warming potential (GWP), NOx and index of environmental sustainability (Sust), (c) average N and P in the leachate and (d) net farm income and attributes of sustainability of milk quality (Milk Q), biodiversity (Biodiv), landscape aesthetics (Landscape) and animal welfare (Anim. Welfare).

0%

5%

10%

15%

20%

25%

Bas

elin

e

1-1

1-2

1-3

1-4

1-5

1-6

1-7

1-8

1-9

1-10

1-11

1-12

1-13

trajectory-step

Cha

nge

(%)

ha-1 ha-2

ha-3 ha-4

Figure 6. Change in forage area required on trajectories towards sustainable dairy systems In the second part of the scenario testing we defined 3 typical dairy farm typologies as baseline farms on which to test a common trajectory towards sustainability. These represented (i) extended grazing, (ii) conventional half-year grazing and (iii) fully-housed intensive dairy farms. Each typology was subdivided into 3 different farm sub-types to cover some range of management variability in terms of, for example, herd size and fertiliser rate. These farms were simulated for more than one location to also cover some climatic and soil variability. Thereby, in total, 27combinations defined by farm typology, sub-types and location were simulated (Table 6). Tables 1-3 in Appendix 4 show the details of management inputs for each typology baseline farm scenario. The variations in management of each farm typology are described in Table 7. Additional assumptions made in the simulations are given in Appendix 4.

0%

2%

4%

6%

8%

10%

12%

14%

1 2 3 4trajectory

% c

hang

e co

mpa

red

with

bas

elin

e ha

L of milk (total)

-200%

-180%

-160%

-140%

-120%

-100%

-80%

-60%

-40%

-20%

0%1 2 3 4

trajectory

% c

hang

e co

mpa

red

with

bas

elin

e

CH4/ha N2O/ha NH3/ha GWP/haNox/ha Sust

b

-80%

-70%

-60%

-50%

-40%

-30%

-20%

-10%

0%1 2 3 4

trajectory

% c

hang

e co

mpa

red

with

bas

elin

e

Average N mg/L P

c

-200%

-100%

0%

100%

200%

300%

400%

500%

1 2 3 4

trajectory%

cha

nge

com

pare

d w

ith b

asel

ine

Mil QBiodivLandscapeSoil QAnim. Welfarenet farm incomed

a

SID 5 (Rev. 3/06) of 25

Table 6. Soil type and locations studied for each type of baseline farm typology. Id Typology Location soil texture drainage class 1 extended Somerset loam moderate 2 extended Staffordshire clay loam poor 3 conventional Devon loam moderate 4 conventional Cornwall loam good 5 conventional Cheshire loam poor 6 conventional Lancashire loam poor 7 conventional Wales loam good 8 fully housed Yorkshire clay loam moderate 9 fully housed Cheshire loam moderate Table 7. Variations of subtypes studied (summary) including the main changes to baselines. Typology Subtype 1 Subtype 2 Subtype 3 Extended grazing As in table FYM-based decrease grazing days Conventional As in table less herd size (30%), less herd size (100%) 185 grazing days more concentrates more maize silage /less maize /less grass silage less N mineral fertiliser Fully housed As in table less herd size (20%), less herd size (20%), intensive More milk/cow (12%) More milk/cow (12%) more N mineral fertiliser less N mineral fertiliser longer housing period longer housing period more concentrates/ more concentrates/ less grass silage less maize silage In this second part of the scenario testing exercise 11 steps towards sustainability were simulated in one trajectory defined in Table 5, by which management and plant and animal genetic (G) changes were optimised as an additive process (each step incorporated all changes being previously made) in each step. This trajectory was not intended to represent the most successful possible trajectory but the most probable trajectory in terms of feasibility of introduction of steps and also taking account of the results of the first part of the exercise. Each management or genetic-based step was optimised by the SIMSDAIRY optimisation procedure. It aimed first to minimise the N efficiency ratio (defined as N losses/N in herbage) by predicting the best mineral N distribution and rate for each forage area, and subsequently, to maximise the overall environmental index (index adapted from Van Calker et al., 2006) by tuning the rest of management or genetic-based factors. The steps of this trajectory are given in Table 8. Table 8. Steps towards sustainability used in this study. The steps are introduced in an additive process.

Steps Steps towards sustainability (additive steps)

1 Baseline

2 Mineral N fertiliser (optimised)

3 Diet +silage (optimised)

4 Manure application +storage+ %DM (optimised)

5 Manure timing + areas (optimised)

6 Milk properties (optimised)

7 Animal welfare + landscape+ buffer + reseeding

8 Gmilk+GFERT

9 Gh

10 G%N

11 Gurinedung *Optimisation of functional trait that controls the crude protein (CP) content of the plant (G%N). Remaining terms are defined in Table 5.

SID 5 (Rev. 3/06) of 25

‘Overall sustainability’ was defined by 7 indices as the criteria of SIMSDAIRY optimisation. One of these is a composite environmental index (adapted from Van Calker et al., 2006) and accounts for the impacts of farming on eutrophication potential per ha, acidification potential per ha, global warming potential per t of milk and water use per ha and comprises predicted outputs of CH4, N2O, NOx, NH3, NO3 leaching and water use. The rest of the indices are as follows:

(i) economic: farm net profit (pence /L of milk) (ii) quality of milk: defined as milk with enhanced PUFA composition (iii) animal welfare status (iv) level of biodiversity (v) landscape aesthetics (vi) soil quality.

The sustainability targets were set to those defined above for in Table 3 above. It must be pointed out that, although not explored in this study, our results do not intend to cover the whole range of possible conditions in the UK. Our results cannot be extrapolated to different climatic, soil conditions in the UK as sustainability indicators, especially environmental losses, are likely to vary, not only in total but also in different forms of pollutants. Economic results are also very sensitive to herd size, for example, and thereby, farms with larger herds than those simulated in this study would be likely to result in greater net farm margins. Various general trends occurred with optimisation along the trajectories towards greater sustainability. These are detailed in Appendix 4 and involve changes to mineral fertiliser, diet, manure use, interchange of mineral fertiliser and manures, reduced butterfat but greater protein content of milk, increased farm costs for biodiversity and landscape improvements and the increased use of better plant and animal varieties. All the results corresponding to the individual environmental pollutants, CO2 equivalents of global warming potential (CH4 + N2O) and to the different indices of other sustainability attributes: economics (pence/L of milk), milk quality, biodiversity, landscape, soil quality and animal welfare were plotted and shown in tables in Appendix 4. The expected values for the sustainable scenario were transformed to be on a 1unit basis and the predicted results from the different combinations were proportionally transformed to this value of 1. Values < 1 implied meeting the threshold of individual sustainability indices. For clarity of visualisation (to avoid distortion of the graphs), all values >2 and >4 were assumed to be 2 and 4 for pollutants and other sustainability attributes, respectively. Graphs were plotted as radar graphs (also known as a spider plot, a polar plot or a spiral space diagram). This is a two dimensional polar graph that enables us to simultaneously display many variables. It does this by plotting each variable along a different radial axis emanating from the origin of the polar plot. We plotted for a farm scenario 2 different graphs: one with 7 variables for pollutants (average N in leachate, CH4/ha, N2O/ha, NH3/ha, GWP/ha, NOx/ha, average P in leachate) and another one with 6 variables for other sustainability attributes, (net farm income, milk quality, biodiversity, landscape, soil quality and animal welfare). Small values are near the centre of the polar plot and large values near the outer circumference. Thereby, the closer the value is to the centre of the plot, the more sustainable the farm scenario is for that particular variable (sustainable when it reaches the value1). The variable values on each radial axis are linked with a straight line creating a distorted star-like pattern. One can then visually compare the patterns created by the lines representing the different variables for the experiment. A scenario-example is provided below (Fig. 7). The complete set of 27 scenarios is shown in Appendix 4 (Figs. 1-9). An alternative expression of the results (Figs. 8 and 9) shows the overall scope for achieving the sustainability targets of each of the 9 farm scenarios. These show more clearly that there were differences in scope between each farm system and for each of the studied desired targets. Most trajectories met the desired targets for pollution sustainability on the last step. Those targets related to potential eutrophication (average N and P in the leachate/run-off) were met in all cases for NO3

– leaching and for all cases, except for the extended system in Staffordshire, for P leaching/run-off. Predicted P losses were greater in systems under higher grazing pressure (i.e. extended) and heavy soils (where large losses of P from sediments are to be expected). In areas where rain is more limited, as in Staffordshire, P concentrations may well exceed the limits of 100 μg l-1. Very intensive fully-housed systems, due to the higher pressure to produce large yields of herbage/maize per unit of area require higher rates of fertiliser, and the large amounts of manure produced, will in most cases give rise to larger losses of the more soluble and mobile forms of N (e. g. NO3

-). Those targets related to potential impacts on the climate change by production of GHG were met in all cases for CH4 and in almost all cases for N2O, except for the fully housed intensive system in Cheshire. All farm scenarios met the targets If the global warming potential (as CO2 equivalents) from CH4 +N2O is computed as a whole (data not shown). Farms on heavier soils (e.g. clay loam soil in Yorkshire) generally showed greater scope to reduce N2O emissions from soils than those on lighter ones. However, lighter soils (e.g. loam soil in Cheshire) resulted in

SID 5 (Rev. 3/06) of 25

much lower absolute values of N2O losses per hectare than heavier soils. Although changes in the diet were sufficient in all cases to reduce CH4 losses to the targeted values, these reductions were in all cases more modest that those predicted for N2O emissions and thereby, measures to reduce GHG emissions should be preferable aimed at reducing N2O losses. Those targets related to potential impacts on acidifying gases were met in all cases for both NH3 and NOx losses. Ammonia losses were generally reduced through technical improvements in the management (e.g. through capital investment on especial machinery like manure injectors) and NOx losses reduced through either technical improvements in the silage management process and/or through adjustment of N mineral fertiliser applications in terms of does and timing distribution. The scope for reduction in NOx was particularly large in fully-housed intensive farm systems in Yorkshire. This was mainly due to greater scope for reducing NOx soil emissions from fertiliser N applications under these particular edapho-climatic conditions (NOx predicted emissions are very sensitive to those conditions that control nitrification processes).

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0average N

CH4/ha

N2O/ha

NH3/haGWP/ha

Nox/ha

average P

1-Baseline

2

3

4

5

6

7

8

9

10

11

Sustainable

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0Mil Q

Biodiv

Landscape

Soil Q

Anim. Welfare

£/ milk

1-Baseline

2

3

4

5

6

7

8

9

10

11

Sustainable

Figure 7. Examples of ‘spider plots’ showing approach to sustainability possible by the introduction of a trajectory of management steps. Milk quality targets were met in almost all cases. Landscape, biodiversity and soil quality targets were seldom met and the scope for animal welfare and net farm income targets to be met varied with and within systems and sites. It is worth mentioning that although some targets were not met in this last step towards sustainability, in all cases the attributes were greatly improved and possibly some of the targets were no doubt very ambitious. It is also

SID 5 (Rev. 3/06) of 25

worth mentioning about the huge variability in terms of net farm income results within farm systems where farm size was explored as a factor to change. Figure 8. Median and range (defined by minimum, maximum and Quartiles 1 and 3) values for the final step towards sustainability in terms of average N in the leachate, CH4, N2O, NH3, average P in the leachate/run-off and NOx . Values are normalised on a 1-basis. Only values<1 meet the target for sustainability. Figure 9. Median and range (defined by minimum, maximum and Quartiles 1 and 3) values for the final step towards sustainability in terms of landscape aesthetics, soil quality, animal welfare, net farm income/L milk, milk quality and biodiversity. Values are normalised on a 1-basis. Only values<1 meet the target for sustainability. The larger the herd was the larger the net farm income would be. It is clear that unless the public sector subsidizes smaller enterprises (e.g. multifunctional and low intensity farms), the ones with greater scope of reaching overall sustainability without losing money are large farms. Animal welfare was able to be improved up to the desired standards only in cases where cows were able to graze for longer than half a year. There is however a large scope to explore genetic traits that improve both animal nutrient use efficiency and fertility problems. For larger farms, although the target was not met for those farms that rely on winter-autumn housing, in most cases, improvements through more veterinary control and/or improvement in general management of the herd would be potentially ways to overcome this. Unfortunately, both measures are difficult to quantify and moreover they are likely to increase complexity in terms of management and thereby, more costs on staff and more costs related to veterinary care. Soil quality was generally improved but heavier soils under wet conditions had a smaller scope to reach the desired target for sustainability than other soils under drier conditions. Biodiversity was enhanced through making capital investment for field strips for wildlife and wild vegetation and also by replacing the extra existing surface of land (land that is transformed from agricultural production to a field for enhancing biodiversity due to greater nutrient use efficiency of plants) to surface dedicated to biodiversity enhancement. Unfortunately, in most cases it would still require extra surface to buffer the negative impacts of intensive production on the biodiversity levels of the farm land area. Landscape aesthetics was very sensitive to the amount of extra land that was not required for production and could actually be used for enhancement of biodiversity and also landscape quality. Despite the efforts made to objectively define and quantify this landscape

Exte

nded

-Som

erse

t

Exte

nded

-Sta

fford

shire

Con

vent

iona

l-Dev

on

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Wal

es

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Hou

sed-

inte

nsiv

e-C

hesh

ire0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Ove

rall

sust

aina

bilit

y

q1 min median max q3

Average N in leachate

Exte

nded

-Som

erse

t

Exte

nded

-Sta

fford

shire

Con

vent

iona

l-Dev

on

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Wal

es

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Hou

sed-

inte

nsiv

e-C

hesh

ire

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Ove

rall

sust

aina

bilit

y

q1 min median max q3CH4/ha

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Ove

rall

sust

aina

bilit

y

q1 minmedian maxq3 sustainable

N2Oha

Exte

nded

-Som

erse

t

Exte

nded

-Sta

fford

shire

Con

vent

iona

l-Dev

on

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Che

shire C

onve

ntio

nal-L

anca

shire

Con

vent

iona

l-Wal

es

Hou

sed-

inte

nsiv

e-Yo

rksh

ire Hou

sed-

inte

nsiv

e-C

hesh

ire

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Ove

rall

sust

aina

bilit

y

q1 min median max q3

NH3/ha

Exte

nded

-Som

erse

t

Exte

nded

-Sta

fford

shire

Con

vent

iona

l-Dev

on

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Wal

es

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Hou

sed-

inte

nsiv

e-C

hesh

ire

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Ove

rall

sust

aina

bilit

y

q1 min median max q3NO x/ ha

Exte

nded

-Som

erse

t

Exte

nded

-Sta

fford

shire

Con

vent

iona

l-Dev

on

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Wal

es

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Hou

sed-

inte

nsiv

e-C

hesh

ire

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0O

vera

ll su

stai

nabi

lity

q1 min medianmax q3 sustainable

Average P in leachate/run-off

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Ove

rall

sust

aina

bilit

y

q1 min medianmax q3 sustainable

milk quality

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ove

rall

sust

aina

bilit

y


biodiversity

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ove

rall

sust

aina

bilit

y


landscape

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ove

rall

sust

aina

bilit

y


soil quality

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0O

vera

ll su

stai

nabi

lity


animal welfare

Hou

sed-

inte

nsiv

e-C

hesh

ire

Hou

sed-

inte

nsiv

e-Yo

rksh

ire

Con

vent

iona

l-Wal

es

Con

vent

iona

l-Lan

cash

ire

Con

vent

iona

l-Che

shire

Con

vent

iona

l-Cor

nwal

l

Con

vent

iona

l-Dev

on

Exte

nded

-Sta

fford

shire

Exte

nded

-Som

erse

t

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ove

rall

sust

aina

bilit

y

q1

min

median

max

q3

net farm income/L

SID 5 (Rev. 3/06) of 25

aesthetics in this and other studies, it is recognised that this attribute may still be too challenging for objective assessment. The scale of the SIMSDAIRY’s approach, does not at present allow exploring the effects at the larger scale of either polarisation of farms in one specific area or scattering them in small enterprises across a region. It is hence needed to upscale these results appropriately in larger scales (catchment, landscape and regional). ENGAGEMENT WITH THE INDUSTRY AND KNOWLEDGE TRANSFER An important objective of this project was to engage with the Dairy Industry to determine how best the management options for greater sustainability explored with the model framework could be implemented and how barriers to implementation could be overcome. This was done by (i) identifying representatives of the key stakeholders and inviting a selection of them to a Stakeholders Workshop; and (ii) conducting a comprehensive and effective Knowledge Transfer campaign using all modern media methods and outlets. The Stakeholders Meeting was held on 26 April 2006 at IGER, North Wyke. The details of the people invited and the proceedings of the day are detailed in Appendix 3. The Stakeholders were divided into 3 equal-size groups. Each group was asked to consider sets of questions relating to achieving sustainability with one of 3 dairy farm typologies:- conventional, extended grazing and fully housed, in order to focus their discussions. These typologies were defined for the stakeholders in terms of managements and modelled outputs of all sustainability indices. The typologies were very similar to those defined as baselines for the scenario testing exercise reported above The groups were first asked to consider the quantitative accuracy and appropriateness of these baseline definitions and whether the model simulations were realistic. They were then asked to consider trajectories towards sustainability in 3 phases and at each phase comment as to whether each phase was implementable in 5 years, 10 years or never. Examples of the information given to Stakeholders are shown for the conventional typology only in Tables 9 and 10, Fig. 10 and (in full) in Appendix 3. Table 9. Predicted results (environment, milk quality, biodiversity and economics) of the baseline scenario (conventional system)

Figure. 10 Improvements to each index required to meet sustainability with a conventional dairy farm

Environment Waters NO3 g/L of milk 6.1 average N ppm 18.0 average P ppb 7.1 C Change CH4 g/L of milk 13.8 N2O g/L of milk 0.3 GWP g/L of milk 0.4 Acidification NH3 g/L of milk 6.1 NOx g/L of milk 0.00011

Matrices Milk MilkQ Scores 2.6 Biodiversity BioDiv 1.9 Economics NetMargin £ 96526

-40-20

020406080

100120140160

NO3average N

average P

CH4N2O

GWP

NH3Nox

MilkQBioDiv

NetMargin

% im

prov

emen

t

SID 5 (Rev. 3/06) of 25

Table 10. Specifications for the different stages of % change required to achieve a sustainable systems starting from the conventional typology baseline Specifications stage 1 stage 2 stage 3 Final Plant improvement 2 4 6 10 Animal improvement 2 4 6 10 Fertilizer reduction 3 7 12 20 Milk increase/cow 2 4 6 10

Manure application method Surface,

slurry

Slurry (Surface)

50 %; FYM 50 %

Slurry (Injection)

50 %; FYM 50 % Surface, FYM

Manure storage capacity 3 months 5 months 8 months 9 months Improvements in diet(i.e: oils, energy...) none none energy energy, oils Silage making normal normal more efficient more efficient It was generally agreed that the baseline systems were appropriately chosen and that they were adequately specified and simulated. However, there was some dissent shown by individuals over the levels chosen for milk output per cow, stocking rate, use of maize, scope for biodiversity and fertiliser levels, particularly for the extended and fully housed typologies. Several people thought that the model was not sufficiently sensitive and versatile, but others thought that it was too complex to be used on farm. It was pointed out that we had deliberately set up rather simplistic simulations to suit the purpose of the day and that the model was capable of very complex simulations at the specific site level. The conventional and extended typology groups were positive about the feasibility of sustainability being achievable in the ways and times specified during the 10 year period, given appropriate incentives. However, individuals in the fully housed group doubted that much progress would be achievable after stage 1 (5 years) with this typology, but pointed out that there was probably much more scope for promoting biodiversity with this typology than had been specified in the simulations. It was thought generally that the best ways of progressing the results of this project included further model development on the one hand, to include a wider range of crops and legume species, but on the other, to develop a simpler version for on-farm use. It was thought that farmers should be offered financial incentives for introducing the steps towards sustainability that incurred profit reductions and these may be linked to carbon (C) and/or energy saving initiatives, or to biodiversity and wildlife habitat improvements schemes. The need for validation of the model against real farm performance was pointed out and once this had been one the model should be utilised at catchment/regional scales. There was some support for the idea of optimising the sustainability of milk production by phasing out activity in unsuitable localities/sites, where, for either climatic or soil factors, achievement of sustainability would be too costly. Information from this project has been disseminated to a range of stakeholders at various stages throughout the project. Initially, an article was published in the Defra Newsletter (September, 2004) to raise awareness of the project. In addition, a briefing document was written in association with key stakeholders, LEAF and Velcourt, and disseminated via the stakeholder group and placed on the IGER website (http://www.iger.bbsrc.ac.uk/research/departments/sees/teams/MFR/SIMS_Dairy.htm). One of the roles of the stakeholder group was to assist in disseminating information and seek feedback from industry. The SIMSDAIRY modelling framework was demonstrated to farmers and advisors at the North Wyke 25th Anniversary Open Day in 2006 (which over 250 farmers and advisors attended), and was also featured at the 2005 Grassland/Muck Event at Stoneleigh. This generated much interest (particularly from advisors), with queries about when the model would be available for use. It is clear that the model is most suitable for policy makers and advisors and would require a simplified version for farmers to use. There has also been much interest in the SIMSDAIRY model from researchers both in the UK and overseas. During a recent visit to New Zealand, we were requested to demonstrate the model’s capabilities to quantify potential ‘trade-offs’ to agriculture and greenhouse gas experts from both New Zealand (Agresearch, Landcare, Pastoral Greenhouse Gas Research Consortium- PGGRC) and the USA. Scientific interest in SIMSDAIRY is evidenced by the written outputs and invitations to participate and present at numerous national and international conferences (see the list of outputs in section 9 and appendixes 7-13). Investment in this unique modelling framework has since resulted in SIMSDAIRY being modified and used in other Defra projects addressing whole farm sustainability. For example, it was used in project CC0270, The implications of farm-scale methane mitigation measures for long-term national methane emissions, and is currently being used in projects AC0209, Ruminant nutrient regimes to reduce methane and nitrogen emissions; AC0307, Assessment of the environmental consequences of adaptation of the livestock sector to climate change; and work package 4 of WQ0106, Underpinning evidence and new model frameworks for mitigating multiple diffuse water pollutants from agriculture.

SID 5 (Rev. 3/06) of 25

CONCLUSIONS This project has delivered the basic knowledge for the definition of all important indices of sustainability for UK dairy systems and an extremely sophisticated and versatile modelling framework for simulating and optimising the effects of farm managements and input parameters on those indices within realistic farm scenarios. We have comprehensively explored the relevant fractions of the response surface of this framework to quantify the sustainability of our current systems and to specify novel sustainable systems and how these might be achieved by stepwise changes in managements and the quality of the main plant and animal components. This work demonstrates that there is high variability in both the position of current systems with regard to sustainability and ease with which each one could achieve sustainability by implementing known techniques and managements. Farm location and size were shown to be all important in determining the likely degree of success with achieving both economic and environmental sustainability targets:- smaller farms in drier areas on freely-draining soils would need greater incentives and financial aid. The study has also pointed to the essential contribution that genetically improved plants and animals must make to the achievement of sustainability. The approach has been extensively challenged as regards both it’s scientific rigor and practicality for guiding the dairy industry, by way of:- (i) the composition of the research team; (ii) ongoing discussion with Defra; (iii) very high extent of publication at all levels and (iv) focused discussion by Stakeholders. Our work indicates that the UK dairy sector could be almost totally environmentally and economically sustainable within a 10 year period given the appropriate incentives by policy makers and sufficient confidence in the value of the modelling approach by farmers. It is clear though that such success will demand sweeping changes in attitudes to implementation of pollution mitigation measures and the ways that such implementation is regulated. There is a danger though that imposition of inflexible and/or unfeasible ways of improving the dairy industry may result in catastrophic reductions in farming activity. In view of the rapid changes in global commodity and food prices currently being witnessed due to increased cropping for energy, increasing incidence of crop failure due to changing climates and increasing world demand for milk and meat due to population growth, it will be extremely important to ensure that the mainstream UK dairy industry survives to make its contribution. INTELECTUAL PROPERTY This project has developed a unique and novel modelling framework (SIMSDAIRY) which has clear opportunities for commercial exploitation. A simplified and more user-friendly version of the current SIMSDAIRY would have an enormous potential to be widely used within the UK milk-based industry (e.g. to indicate on-farm site-specific sustainability impacts, sustainability footprint...) or farmers themselves (e.g. to optimise their farm management). REFERENCES Brown, L., Scholefield, D., Jewkes, E.C., Lockyer, D.R. and del Prado, A. 2005. NGAUGE: A decision support

system to optimise N fertilization of British grassland for economic and environmental goals. Agriculture Ecosystems and Environment 109: 20-39.

Butler, A. J. and M. M. Turner. 2007. Modelling integrated dairy systems in the UK: towards economic and environmental sustainability, Agricultural Economics Society, 81st Annual Conference, Reading, 2nd – 4th April, 2007.

Chadwick, D.R. and Pain, B.F. 1997. Methane fluxes following slurry applications to grassland soils: laboratory experiments. Agriculture Ecosystems and Environment 63: 51-60.

Chambers, B.J., Lord, E.I., Nicholson, F.A. and Smith, K.A. 1999. Predicting nitrogen availability and losses following application of organic manures to arable land: MANNER. Soil Use and Management 15: 137-143.

Colman, D. and Y. Zhuang. 2005. Changes in England and Wales dairy farming since 2002/03: a resurvey. University of Manchester.

Colman, D., J. E. Farrar and Y. Zhuang. 2004. Economics of Milk Production: England and Wales 2002/03. University of Manchester.

Competition Commission, 2007. The supply of groceries in the UK market investigation: provisional findings report. London: Competition Commission, October 2007.

Curry Commission (2002). Farming and Food: a sustainable future. Report of the Policy Commission on the Future of Farming and Food. London: Cabinet Office

Davison, P.S., Withers, P.J.A., Lord, E.I., Betson, M.J., Stromqvist, J., 2008. PSYCHIC - A process-based model of phosphorus and sediment mobilisation and delivery within agricultural catchments. Part 1: Model description and parameterisation. Journal of Hydrology 350, 290-302.

Defra. 2000. Defra publication, Fertiliser recommendations for agricultural and horticultural crops (RB209), 7th edition. 2000. http://www.defra.gov.uk/environ/pollute/rb209/

Defra. 2002. The Strategy for Sustainable Farming and Food: Facing the Future. London: Defra Pubications Giger-Reverdin, S., Morand-Fehr, P. and Tran, G. 2003. Literature survey of the influence of dietary fat

composition on methane production in dairy cattle. Livestock Production Science 82: 73-79.

SID 5 (Rev. 3/06) of 25

Goodlass, G., Halberg, N., Verschuur, G., 2003. Input output accounting systems in the European community - an appraisal of their usefulness in raising awareness of environmental problems. European Journal of Agronomy 20, 17-24.

Grime, J.P. 1979. Plant strategies and vegetation processes. Chichester: John Wiley and Sons. Haygarth, P.M., Chapman, P.J., Jarvis, S.C., Smith, R.V., 1998. Phosphorus budgets for two contrasting

grassland farming systems in the UK. Soil Use and Management 14, 160-167. Jarvis, S.C. 1999. Accounting for nutrients in grassland: challenges and needs. Accounting for Nutrients: A

Challenge for Grassland Farmers in the 21st Century, British Grassland Society Occasional Symposium No 33, BGS Conference, Great Malvern, UK, 22-23 November 1999 Corrall, A. J., ed, 3-12. BGS.

MAFF. 2001. Guidelines for farmers in NVZs. MDC, 2003. Prices and profitability in the British dairy chain. Report by KPMG to the Milk Development Council.

Cirencester: MDC, 2003 NFU/RABDF, 2007. British Milk: What Price 2007? Report published by the National Farmers Union and the

Royal Association of British Dairy Farmers. Stoneleigh: NFU/RABDF, May 2007 Noordhuizen, J. and Metz, J.H.M. 2005. Quality control on dairy farms with emphasis on public health, food

safety, animal health and welfare. Livestock Production Science 94: 51-59. Ondersteijn, C.J.M., Beldman, A.C.G., Daatselaar, C.H.G., Giesen, G.W.J., Huirne, R.B.M., 2002. The Dutch

mineral accounting system and the European nitrate directive: implications for N and P management and farm performance. Agriculture Ecosystems and Environment 92, 283-296.

Scholefield, D., Jewkes, E.C. and Bol, R. 2007. Nutrient cycling budgets in managed pastures In: Marschner P. and Rengel, Z. (eds.), Nutrient cycling in terrestrial ecosystems, Soil Biology Volume 10: 215-256. Springer-Verlag, Berlin and Heidelberg.

Schroder, J.J., Aarts, H.F.M., ten Berge, H.F.M., van Keulen, H., Neeteson, J.J., 2003. An evaluation of whole-farm nitrogen balances and related indices for efficient nitrogen use. European Journal of Agronomy 20, 33-44.

Schroder, J.J., Scholefield, D., Cabral, F., Hofman, G., 2004. The effects of nutrient losses from agriculture on ground and surface water quality: the position of science in developing indicators for regulation. Environmental Science and Policy 7, 15-23.

Smith, K.A., Brewer, A.J., Crabb, J. and Dauven, A. 2001. A survey of the production and use of animal manures in England and Wales. III. Cattle manures. Soil Use and Management. 17(2), 77-87.

Stromqvist, J., Collins, A.L., Davison, P.S., Lord, E.I., 2008. PSYCHIC - A process-based model of phosphorus and sediment transfers within agricultural catchments. Part 2. A preliminary evaluation. Journal of Hydrology 350, 303-316.

Thomas, C. 2004. Feed into Milk (FiM): A new applied feeding system for diary cows. Nottingham University Press, UK.

UNECE. 1999. Protocol to the 1979 convention on long-range transboundary air pollution to abate acidification, eutrophication and ground-level ozone. United Nations Economic Commissions for Europe (UNECE), Geneva.

Van Calker, K.J., Berentsen, P.B.M., Giesen, G.W.J., Huirne, R.B.M., 2005. Identifying and ranking attributes that determine sustainability in Dutch dairy farming. Agriculture and Human Values 22, 53-63.

Van Calker, K.J., Berentsen, P.B.M., Romero, C., Giesen, G.W.J., Huirne, R.B.M., 2006. Development and application of a multi-attribute sustainability function for Dutch dairy farming systems. Ecological Economics 57, 640-658.

Webb, J. and Misselbrook, T.H. 2004. A mass-flow model of ammonia emissions from UK livestock production. Atmospheric Environment 38: 2163-2176.

Webb, J., Ryan, M., Anthony, S.G., Brewer, A., Laws, J., Aller, M.F. and Misselbrook, T.H. 2006. Cost-effective means of reducing ammonia emissions from UK agriculture using the NARSES model. Atmospheric Environment 40: 7222-7233.

Winter, M. 2002. Rural policy: new directions and new challenges. CRR research Report No. 3, University of Exeter, April 2002.

NOTE ON APPENDIXES Appendix 1. Definitions and use of score matrices for modelling framework. Appendix 2. Structure of the SIMSDAIRY framework. Appendix 3. Stakeholders meeting. Appendix 4. Runs with SIMSDAIRY. Appendix 5. SIMSDAIRY input data. Appendix 6. Documents as the project progressed. Appendixes 7-13. Outputs [summary (7), peer reviewed articles in international journals (8), edited conferences

(9), book chapter (10), popular articles (11), oral international presentations (12) and other oral presentations (13].

SID 5 (Rev. 3/06) of 25

References to published material

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

OUTPUTS

Peer-reviewed international journals

Del Prado A. and Scholefield D. 2008. Use of SIMSDAIRY modelling framework system to compare the scope on the sustainability of a dairy farm of animal and plant genetic-based improvements with management-based changes. Journal of Agricultural Science. 146 (2): 1-17.

Schils R.L.M., Olesen J.E., del Prado A. and Soussana J.F. (2007) A farm level approach for mitigating GHG emissions from ruminant livestock systems. Livestock Science. 112 (3): 240-251.

Del Prado A., Cardenas L. and Scholefield D. 2006. Impact of NO3 leaching abatement measures on N2O and CH4 emissions from a UK dairy system. International Congress Series. 1293: 359-362.

Del Prado A. and Scholefield D. 2006. Use of SIMSDAIRY modelling framework system to specify sustainable UK dairy farms. Aspects of Applied Biology. 80: 73-80.

Del Prado A., Misselbrook T., Chadwick D., Hopkins A., Dewhurst R., Davison P., Butler A., Turner M., Schröder J. and Scholefield D. (in advanced preparation.) SIMSDAIRY. A modelling framework to identify new integrated dairy production systems (description and sensitivity analysis). Agriculture, Ecosystems and Environment.

Del Prado and Scholefield D. (at early stages.). Use of SIMSDAIRY modelling framework system to specify trajectories towards sustainable UK dairy farms. Agriculture, Ecosystems and Environment.

Peer-reviewed international edited meeting/workshops/conferences Butler, A. J. and M. M. Turner. 2007. Modelling integrated dairy systems in the UK: towards economic and

environmental sustainability, Agricultural Economics Society, 81st Annual Conference, Reading, 2nd – 4th April, 2007.

Del Prado A., Scholefield D., Chadwick D., Misselbrook T., Haygarth P., Hopkins A., Dewhurst R., Davison P., Lord E., Turner M., Aikman P. and Schröder J. 2006. A modelling framework to identify new integrated dairy production systems. EGF: 21st General Meeting on 'Sustainable grassland productivity", Badajoz, Spain, 3-6 April 2006.

Schils R.L.M., Olesen J.E., del Prado A. and Soussana J.F. 2006. Keynote paper: A farm level approach for mitigating GHG emissions from ruminant livestock systems. 12th RAMIRAN International Conference: "Technology for recycling of manure and organic residues in a whole-farm perspective", Aarhus, Denmark, 11th - 13th September, 2006.

Book chapter Del Prado A., Scholefield D. and Brown L. 2006. A model to simulate the effects of different dietary

strategies on the sustainability of a dairy farm system. In: Kebreab et al. (eds), Nutrient Digestion and Utilization in Farm Animals: Modelling Approaches. CAB International.

Popular articles Murray P.J., Chadwick D.R., del Prado A., Hopkins A., MacLeod C.J., Misselbrook T.H. and Scholefield D.

2006. Improving Environmental Quality. In: Gordon A.J. (eds.), IGER Innovations. 10: 37-41. Rook A., Misselbrook T and Scholefield D. 2006.Multifunctional uses for pastures. In: Gordon A.J. (eds.),

IGER Innovations. 10: 43-47. Chadwick D. 2005. New integrated dairy production systems: specification, practical feasibility and ways of

implementation. IGER Website. Scholefield D. 2004. New integrated dairy production systems: specification, practical feasibility and ways

of implementation. Defra Newsletter, September, 2004.

SID 5 (Rev. 3/06) of 25

SID 5 Research Project Final Report -...

Documents

Transcript of SID 5 Research Project Final Report -...