Climate change impacts on the livestock sector AC0307randd.defra.gov.uk/Document.aspx?Document=AC0307_8527... · Climate change impacts on the livestock sector AC0307 DRAFT FINAL

Climate change impacts on the livestock sector

AC0307

DRAFT FINAL REPORT

LEAD AUTHORS: Dominic Moran, Kairsty Topp, Eileen Wall, Anita Wreford

CONTRIBUTING AUTHORS:

David Chadwick, Clare Hall, Mike Hutchings, Malcolm Mitchell, Agustin Del Prado, Bert Tolkamp, Lianhai Wu

With further contributions from

Ross Davidson, Cathy Dwyer, Marie Haskell, Davy McCracken, Alistair McVittie, Vicky Sandilands, Anita Shepherd, Nick Sparks, Steven

Thompson

SAC COMMERCIAL LTD SAC RESEARCH

22 JULY 2009

  i

Executive summary This report considers the issue of climate change adaptation as it applies to the UK livestock sector comprising the beef, dairy, sheep, pigs and poultry industries. Beyond an understanding of direct impacts from warming, a key policy consideration relates to indirect or ancillary costs that might arise incidental to any accelerated private adaptation decision-making. That is, how changing patterns of autonomous or private on-farm adaptation responses might exacerbate external impacts in terms of animal welfare from heat stress, disease and diffuse pollution to water and air. In relation to the latter, an important consideration is how adaptation responses potentially undermine other environmental objectives including greenhouse gas mitigation strategies. In this report, UKCIP 02 climate projections are applied across a representative set of UK regions to determine potential impacts across the main livestock categories. This information is developed with a view to quantifying a range of significant impact categories that can be expected under a central warming scenario. These impacts are necessarily limited to those that we currently consider most significant in terms of the potential cost to the industries and wider society. Economic damage estimates are then derived for these impacts under a business as usual scenario, which assumes that animal numbers remain stable and that autonomous adaptation does not occur. While there is potential for under-estimating autonomous adaptation, and therefore over-estimating damage costs, this analysis provides an initial estimate of the value at risk in livestock sectors and beyond, and will be indicative of the likely adaptations required. The relevant impact categories include pastoral/grassland productivity, waste generation (by livestock), heat stress, and disease. In-field management decisions can also be expected to impact upon biodiversity and potentially water quality. However, identifying the potential impacts of adaptations is beyond the scope of this project. In sketching out the impacts resulting from warming scenario, we also have to assume certain behavioural responses by producers. In some cases, the distinction between an impact and an adaptation is blurred. For example, we assume that increased grass availability leads to an increase in days the animal spends outside, and consider this within the impacts section; however it could be argued that increasing the days outside is actually an adaptation. The validity of behavioural responses to warming are not analysed in this report and we highlight this as a specific research gap. The effects of climate change on pastoral systems and regions are likely to increase grass production. Increased forage availability could increase the annual grazing period by a maximum of five weeks for cattle systems, and seven weeks for sheep systems. In most regions this will allow animals to be kept outdoors for longer, and could mean a potential reduction in the proportion of the year that animals require housing and/or access to conserved forages. The changes in length of the grass growth season are generally more substantial in northerly than in southerly regions. Changes in husbandry in response to changes in forage availability may lead to greater potential gains in terms of private productivity. However, any change will also have potential welfare and disease exposure consequences. For example, a prolonged

grazing season extends the exposure to environmental populations of parasites and pathogens. Increased grass/grazing can also imply an increase in acidifying gases and, in the case of the beef and dairy sectors, diffuse pollution loading. The increases in acidifying gases are mainly due to increases in ammonia emissions from slurry. The net global warming impact of the two gases (methane and nitrous oxide) will decrease: on the one hand we predict a reduction in the emissions of nitrous oxide for all sectors. But this will be offset by increased methane emissions due to ruminants being fed more grass and forage. There are some variations between regions for all the livestock types, and the magnitude of the change tends to increase with time for all the environmental pollutants. Climate change, including extreme events, will also affect animal systems and transportation, leading to production and functional losses and increased potential mortality and morbidity costs. We predict that by 2080 mortality costs related to heat stress could amount to around £34 million (present value). These represent private losses and we can expect industry to adapt if guided by appropriate regulatory reform and surveillance. The addition of extreme events to gradual climate change may provide a “shock” to livestock systems. The results show that the damages due to a warming climate will be amplified with the additional stress of extreme events, such as a heat wave. Also, some of the impacts of climate change on animal production and functionality (e.g., health and welfare) may be further exacerbated by adaptations farmers make in other areas of the farming system, such as taking advantage of a longer growing season and keeping animals outdoors for longer periods, thus increasing livestock exposure to prevailing weather conditions. Some of the private adaptations that livestock keepers will put in place to minimise the impact of climate change will not have a large cost (e.g., changing grazing patterns, introducing shelter belts in fields). However, some adaptations may require larger scale interventions and therefore higher costs (e.g., new buildings, introduction of mechanical ventilation/heating) which some livestock keepers may not be able to afford. Climate change has the potential to drive outbreaks of highly infectious exotic disease in the UK livestock industries. Our understanding of the disease risks and impacts (both endemic and exotic) of climate change are limited by data that are required to inform predictive modelling, although the climate signal is likely to increase the risk of exposure of livestock to a range of pathogens and parasites (some may decline). Climate matching techniques offer immediate returns in highlighting diseases of immediate concern for future research, however, process based modelling techniques offer the potential for increased predictive power. Despite the lack of more specific climate-epidemiology modelling, some existing evidence suggests that episodes of certain diseases can lead to significant economic impacts within and beyond the sector. There is also an increased potential for accelerated drug resistance. For each of the preceding impact categories we consider the nature of any adaptation responses and whether there are unanticipated external costs occasioned by private adaptation choices. Overall, we do not yet find compelling evidence on social costs in terms of increased greenhouse gases from management changes, though the impacts of diffuse pollution to water may warrant further investigation in terms of linking waste generation to water quality objectives in specific receiving waters. Nevertheless we consider that public adaptation needs are often convergent with the aims of existing

  ii

national and EU environmental regulations (e.g. climate change mitigation plans or Nitrates Directive). We suggest a need for further examination of extreme event information, and for detailed modelling of new disease threats. The broad conclusion we reach at this point is that while there is a need to adapt to climate changes, the extent of required adaptation is largely within the capacity of the livestock industry and can be motivated by information to provoke an attitudinal shift and increased awareness that climate change presents certain risks that need to be factored into farm and production planning. Relevant information includes clear data on the potential frequency and magnitude of extreme events. Information provision is simply a basis of facilitating informed private action. In addition to improved scenario information, some regulatory reform may also be required to accommodate changing housing and transportation requirements. In the case of disease, the returns to anticipatory disease surveillance investments already yield demonstrably good rates of return, and will therefore continue to be a good adaptation strategy. Rapid identification of emergent disease challenges is often a critical factor affecting the effort required for control. Consideration should be given to more pro-active early warning systems and the potential use of fast throughput screening technologies deployed outside of and within the UK. Instigation of long term parasite and pathogen monitoring schemes would greatly improve our capacity to accurately predict emergent disease challenges to the livestock industries that are driven by climate change. Proposed cost-sharing instruments can also be considered part of a rational public adaptation strategy that attempts to modify producer behaviour and re-allocated responsibility for disease costs. The agenda suggested here does not currently represent a radical change from the gradual autonomous adaptation that characterises the adjustments different parts of the industry have had to undertake in response to market liberalisation and other changes over the last two decades. Some attention also needs to be given to the synergies and interactions between mitigation efforts and adaptation goals. Considering the two climate change responses in isolation may lead to trade-offs and maladaptations, undermining the response of the agricultural sector as a whole. Specifically, ways in which a more immediate greenhouse gas mitigation agenda will influence adaptive capacity, and conversely, how adaptation actions may affect mitigation efforts. The project used input from industry experts to gain insights into priority actions in relation to adapting to climate change. This information suggested that adaptation decisions are a low priority relative to more immediate business re-structuring in the face of changing market conditions. To the extent that climate change figures in decisions, it is the more immediate mitigation obligations that are gradually being factored in as potential business costs. Adaptations may be considered only to the extent that: a) climate scenarios and damage information can be made more specific; and b) that these are incidental to, or convergent with, improving financial rates of return. To date there is little evidence suggesting a clear case for the latter. Information from the same experts also assisted in drawing up a range of potential adaptation options and their costs. Qualitative and quantitative surveys were then used to identify options and how these are ranked by industry participants. This report can be viewed as a contribution to the national requirement to develop an adaptation economic analysis (AEA) within the context of the Climate Change Risk Assessment (CCRA) (See Watkiss et al 2009). This exercise presents a first attempt

  iii

to integrate impact and valuation information as a basis of deciding whether livestock impacts are significant and where these impacts fit within (agricultural) sector and national priorities.

  iv

  v

Contents Executive Summary i 1 Introduction ............................................................................................................. 1 2 Characteristics of UK livestock systems ................................................................. 5 3 Climate change projections for the UK ................................................................... 7

3.1 Main messages for the UK relevant to livestock production and health.......... 7 3.2 UKCP09 .......................................................................................................... 9

4 Livestock and climate change impacts ................................................................. 10 4.1 Existing research of impacts on agriculture .................................................. 10

5 Climate change responses ................................................................................... 13 5.1 Adaptation..................................................................................................... 13 5.2 What is the aim of adaptation? ..................................................................... 14 5.3 Adaptation in agriculture ............................................................................... 15 5.4 Public/private interest.................................................................................... 16 5.5 Adaptation in the livestock sector ................................................................. 16 5.6 Adaptation and impact assessments appraisal............................................. 17 5.7 Costs of adaptation ....................................................................................... 19

6 Cost-benefit analysis (CBA) under uncertainty ..................................................... 21 6.1 Defining efficient adaptation under uncertainty............................................. 21 6.2 Project approach to adaptation appraisal ..................................................... 22 6.3 Impacts storylines ......................................................................................... 24

7 Impacts of climate change on grazing systems .................................................... 25 7.1 Introduction ................................................................................................... 25 7.2 Models .......................................................................................................... 25 7.3 Scenarios ...................................................................................................... 27 7.4 Results .......................................................................................................... 27 7.5 Discussion..................................................................................................... 39 7.6 Conclusions .................................................................................................. 40

8 Modelling the effect of climate change on environmental pollution losses from grassland based systems (beef, sheep and dairy) ....................................................... 41

8.1 Introduction ................................................................................................... 41 8.2 Methodology ................................................................................................. 41 8.3 Description of current versions for IGER models. ......................................... 42 8.4 Description of case study: farm typologies, scenarios. ................................. 43

8.4.1 Dairy...................................................................................................... 43 8.4.2 Beef....................................................................................................... 45 8.4.3 Sheep.................................................................................................... 45

8.5 Results and discussion: SAC, NWRes/IGER................................................ 46 8.5.1 Dairy...................................................................................................... 46 8.5.2 Beef....................................................................................................... 53 8.5.3 Sheep.................................................................................................... 57

8.6 Discussion of results ..................................................................................... 62 9 Climate impacts on welfare: expert review, transport and thermal challenges. .... 64

9.1 Introduction ................................................................................................... 64 9.2 Impacts of climate change on livestock welfare............................................ 64 9.3 Impacts of climate change on dairy cattle..................................................... 66 9.4 Impacts of climate change on beef cattle...................................................... 67 9.5 Impacts of climate change on sheep ............................................................ 68 9.6 Impacts of climate change on pigs and poultry ............................................. 69 9.7 Impacts of climate change on livestock in transport...................................... 71 9.8 Examples of thermal stress in transport........................................................ 73 9.9 Impacts of thermal challenges for livestock in transit.................................... 76 9.10 Quantitative assessment of the impact of thermal challenges on livestock. . 79

9.11 Summary....................................................................................................... 85 10 Impacts of climate change of disease risk to livestock ..................................... 88

10.1 Introduction ................................................................................................... 88 10.2 Background................................................................................................... 88 10.3 Qualitative predictions of future disease risk to the UK................................. 89 10.4 Quantitative approaches to predicting disease risk....................................... 90 10.5 Changing management in response to a changing climate .......................... 94 10.6 Conclusions .................................................................................................. 95

11 Estimating the cost of impacts .......................................................................... 97 11.1 Costs from each impact category ................................................................. 98 11.2 Impacts on grazing systems ......................................................................... 99

11.2.1 Dairy.................................................................................................... 101 11.2.2 Beef..................................................................................................... 101 11.2.3 Sheep.................................................................................................. 102

11.3 Impacts from environmental pollution losses from grass-based systems... 103 11.4 Heat stress.................................................................................................. 104

11.4.1 Dairy cows........................................................................................... 105 11.4.2 Beef cows............................................................................................ 106 11.4.3 Pigs and Poultry .................................................................................. 107

11.5 Total losses from mortality .......................................................................... 108 11.6 Losses from transport ................................................................................. 109 11.7 Conclusion .................................................................................................. 109

12 Identification of adaptation options and attitudes............................................ 110 12.1 Possible adaptation options ........................................................................ 110

12.1.1 Potential adaptation to climate change in dairy systems .................... 110 12.1.2 Potential adaptation to climate change in beef systems ..................... 111 12.1.3 Potential adaptation to climate change in sheep systems .................. 111 12.1.4 Potential adaptation to climate change in pig and poultry systems..... 111 12.1.5 Potential adaptation to climate change for livestock transportation .... 112

12.2 Refined adaptation options ......................................................................... 118 12.2.1 Workshop discussions ........................................................................ 118 12.2.2 Further survey research on potential adaptation choices ................... 119 12.2.3 Q-methodology.................................................................................... 121

12.3 Conclusions ................................................................................................ 122 13 Assessment of key public responses.............................................................. 123

13.1 Placing livestock in context ......................................................................... 124 13.2 Analysis of shortlisted adaptation options................................................... 124

13.2.1 Uncertainty .......................................................................................... 126 13.2.2 Public/Private ...................................................................................... 126

13.3 Interaction with mitigation ........................................................................... 128 14 Conclusions and recommendations................................................................ 130 15 References...................................................................................................... 133

  vi

Appendices Appendix One

Appendix to section 1

Appendix Two

Appendix to section 7

Appendix Three

Appendix to section 8

Appendix Four

Appendix to section 9

Appendix Five

Appendix to section 10

Appendix Six

Appendix to section 11

Appendix Seven Appendix to section 12 List of tables Table 3.1 Differences in the methodology employed in UKCIP02 and UKCP09

(taken from ukclimateprojections.defra.gov.uk) ...................................... 9 Table 6.1 Robustness of adaptation options (from Hallegatte 2009) .................... 23 Table 7.1 Average start day (since 1st January) of grazing season for dairy system

.............................................................................................................. 29 Table 7.2 Average annual biomass silage (t/ha) for dairy system ........................ 30 Table 8.1 Main characteristics of the typical dairy farm used as baseline ............ 44 Table 8.2 Definition of variables used for beef baselines (farm level)................... 45 Table 8.3 Definition of variables used for sheep baselines (farm level) ................ 45 Table 8.4 Results for the different environmental pollutants in dairy farms .......... 50 Table 8.5 Projections of annual drainage volume (mm) for dairy systems ........... 52 Table 8.6 Projections of NO3 leaching (kg NO3-N/ha) for dairy systems............... 52 Table 8.7 Projections of milk production per hectare in dairy systems ................. 52 Table 8.8 Projections of land area (grass + maize) required in dairy systems...... 52 Table 8.9 Projections of net farm income in dairy systems................................... 52 Table 8.10 Projections of biodiversity level in dairy systems .................................. 52 Table 8.11 Projections of soil quality level in dairy systems ................................... 53 Table 8.12 Results for the different environmental pollutants in beef farms ........... 55 Table 8.13 Projections of meat production per hectare in beef systems. ............... 57 Table 8.14 Projections of grassland area required in beef systems. ...................... 57 Table 8.15 Results for the different environmental pollutants in sheep farms ........ 60 Table 8.16 Projections of meat production per hectare in sheep systems.............. 62 Table 8.17 Projections of grassland area required in sheep systems..................... 62 Table 9.1 Summary of climate change in summer in dairy regions of England and

Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on mature dairy cows............................................................... 66

Table 9.2 Summary of climate change in spring in dairy regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on mature dairy cows............................................................... 66

Table 9.3 Summary of climate change in summer in dairy regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on dairy calves ......................................................................... 67

  vii

Table 9.4 Summary of climate change in summer in beef regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on beef cattle ........................................................................... 67

Table 9.5 Summary of climate change in winter in beef regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on out wintered beef cattle....................................................... 67

Table 9.6 Acceptable temperature limits for productivity of pigs. Below the lower limit, animals will use feed energy to keep warm and feed efficiency will suffer. Above the upper limit feed intake will reduce and weight gain suffer (taken from McGlone & Pond 2002)............................................ 69

Table 9.7 Potential heat and cold stress problems affecting the UK..................... 75 Table 9.8 Mortality losses in slaughter pigs in transport over a range of

temperatures (Dewey et al 2005 ........................................................... 77 Table 9.9 Table showing risk of cold stress during spring (March-May) during rain

for categories of beef cattle likely to be grazing during these periods. Main beef-production areas are shown................................................. 80

Table 9.10 Summary of the classes of animals, assumed THI threshold and the biological traits modelled in studying the impacts of heat stress on livestock ................................................................................................ 82

Table 9.11 Annual impacts of heat stress on the duration of heat stress (hours/cow/year) and selected production (Milk_loss: kgs of milk lost/cow/year), reproduction (DO_loss: increase in the number of days open/cow/year) and mortality (PDeath: number of deaths in 1000 cows due to heat stress) in dairy cows by region in 2050 and 2080 .............. 83

Table 9.12 Annual impacts of heat stress on the duration of heat stress and selected production (Gain_loss: reduction in kgs of live weight gain/animal/year) and mortality (PDeath) in finishing beef animals by region in 2080 (Further results in Appendix 4) ...................................... 84

Table 9.13 Annual impacts of heat stress on the duration (Dur) of heat stress and selected production (Gain_loss; Egg_loss: reduction in kg eggs/animal/year), and mortality (PDeath) in pigs (sows and growers/finishers) and poultry (layers) by region in 2050 and 2080 ..... 85

Table 9.14 Impacts of heat stress in the month of August with and without and heat wave on dairy cows by region in 2020 and 2080 .................................. 86

Table 11.1 Impacts costed from each chapter ........................................................ 99 Table 11.2 Total forage cost per cow over four points in time under climate change

............................................................................................................ 100 Table 11.3 Discounted cumulative regional dairy grazing and forage costs under

climate change .................................................................................... 101 Table 11.4 Discounted cumulative regional beef grazing and forage costs under

climate change .................................................................................... 102 Table 11.5 Discounted cumulative regional sheep grazing and forage costs under

climate change .................................................................................... 103 Table 11.6 Projections of net farm income in dairy systems................................. 104 Table 11.7 Present value of mortality, reproductive and production losses due to

heat stress in dairy animals (cows and heifers) .................................. 105 Table 11.8 Present value of beef finishing loss in body weight gain..................... 106 Table 11.9 Present value of body weight gain loss in grower/finishers................. 107 Table 11.10 Present value of egg loss.................................................................... 108 Table 11.11 National level present value of costs arising from mortality from major

animal categories ................................................................................ 108 Table 12.1 Climate change impacts and adaptation options for the livestock sector

............................................................................................................ 114 Table 12.2 Three response categories with greatest number of comments, for each

question............................................................................................... 120

  viii

Table 13.1 Adaptations assessed as to their robustness against uncertainty, their effect on mitigation, and who would be adapting (++ = this adaptation always has the characteristic of the column it is in; + = in some cases the adaptation has this characteristic, but not all) ............................... 125

List of figures Figure 3.1 Change in thermal growing season (2020) ............................................. 8 Figure 3.2 Change in thermal growing season (2050) ............................................. 8 Figure 5.1 Costs and benefits of adaptation........................................................... 20 Figure 6.1 Widening uncertainty............................................................................. 21 Figure 7.1 Predicted usable dry matter yield per ha of dairy grass land in 10

regions at four points in time ................................................................. 32 Figure 7.2 Predicted changes in stocking rate on dairy farms in 10 regions in four

points in time. ........................................................................................ 32 Figure 7.3 Predicted grass DM consumption as a proportion of total forage DM

intake on dairy farms in ten region at four points in time....................... 33 Figure 7.4 Predicted N-excess (i.e. difference between N imported onto the farm in

the form of concentrate minus the N exported with milk) per ha for ten regions at four points in time ................................................................. 33

Figure 7.5 Predicted usable dry matter yield (DMY) per ha of beef grass land in 10 regions at four points in time ................................................................. 34

Figure 7.6 Predicted changes in stocking rate on beef farms in 10 regions at four points in time. ........................................................................................ 35

Figure 7.7 Predicted grass DM as proportion of total forage DM consumed on beef farms in ten region at four points in time ............................................... 35

Figure 7.8 Predicted week in the year of the last cut of grass on beef farms in 10 regions at four points in time ................................................................. 36

Figure 7.9 Predicted usable dry matter yield (DMY) per ha of sheep grass land in 10 regions at four points in time ............................................................ 37

Figure 7.10 Predicted changes in stocking rate on sheep farms in 10 regions at four points in time. ........................................................................................ 38

Figure 7.11 Predicted grass DM consumption as a proportion of total forage DM intake on beef farms in ten region at four points in time ....................... 38

Figure 8.1 General modeling framework with the main input-output data linking SAC and NWRes/IGER models. (Where MH stands for medium high emission profile scenarios from UKCIP02). .......................................... 42

Figure 8.2 Relationship between annual drainage volume (mm) and average NO3-N concentration in the leachate for the different RDP areas within each time-slice. .............................................................................................. 48

Figure 9.1 Schematic representation showing the relationship of thermal zones and temperatures. ........................................................................................ 65

Figure 9.2 Chart of the severity of heat stress in dairy cattle (Wiersma, 1990)...... 65 Figure 10.1 Sheep fluke outbreaks in Scotland during 2003-2006 from the SAC

veterinary records ................................................................................. 93

  ix

  1

1 Introduction This report considers the issue of climate change adaptation as it applies to the UK livestock sector comprising the beef, dairy, sheep, pig and poultry industries. Climate change has climbed steadily up the list of government priorities with the Stern Review (2006) providing a compelling economic case for advancing spending on mitigation. Climate projections from the Intergovernmental Panel on Climate Change (IPPC) indicate warming trajectories of global temperature ranges between 1.4 and 5.8 degrees centigrade by 2100. There is considerable uncertainty attached to these projections, as well as to the expected impacts and responses to them. While the UK economy is being managed for an emissions trajectory consistent with a global 2 degrees target, most realistic adaptation plans must now be based on scenarios that at least allow for the possibility of 4 degrees of warming. A consensus among climate scientists1 suggests that emissions stabilisation consistent with 2 degrees centigrade above pre-industrial levels is increasingly unlikely, and that significant preventative and remedial adaptation efforts will be required across a range of industrial sectors. The UK is economically less vulnerable to climate change relative to some other regions of the world, but its policy response to warming needs to consider both mitigation and adaptation strategies. The Climate Change Act (2008) sets out the UK government’s responsibilities and commitments in terms of mitigation and adaptation to climate change. The adaptation strategy includes a statutory commitment to undertake a periodic climate change risk assessment (CCRA), part of which entails an attempt to update information on the impacts of climate change across vulnerable sectors of the economy. This impact assessment can be the basis of a national prioritisation process that can eventually be informed by a cost-benefit analysis (CBA) of interventions. Due to weather dependence, agriculture is probably the sector most vulnerable to the effects of climate change in the UK. This matters because while the financial contribution of the sector may be small, its role in the provision of public goods2 (e.g. ecosystem goods and services) is significant. How agriculture adapts to maintain productivity, and how this in turn affects the provision of public goods, potentially has wider consequences for social well-being. The impacts need to be factored into any public policy response and associated investment decisions. Adaptation is therefore an important public policy priority, although there are other clear distinctions between policy on adaptation and mitigation. The first is that although climate science is becoming more categorical in terms of the relationships between greenhouse gas concentrations and dangerous climate change, there is still much uncertainty about global scenarios on emissions and the consequences of climate impacts at regional and local scales. This uncertainty does not affect efficient policy on meeting mitigation targets for compliance with international agreements. That is, mitigation policy decisions do not directly depend on damage outcomes, and can instead be based on an appraisal of where

1 http://www.guardian.co.uk/environment/2009/apr/14/global-warming-target-2c 2 Public goods in economic jargon, describes a good that is non excludable and that can be enjoyed

by all once it is provided. Agriculture affects many rural public goods, e.g. landscapes, water quality, biodiversity and clean air.

the lowest cost emissions reductions can be found as soon as possible (i.e. a cost -effectiveness appraisal)3. But impact uncertainty is much more problematic in relation to cost-benefit decision making on adaptation. Such decisions would ideally compare the costs of measures to the benefits, expressed in terms of the damages they are helping to avoid. Uncertainty about these benefits (i.e. impact scenarios), means that decision making is typically restricted to cost-effective spending on adaptations that are flexible, reversible or entail no regrets4. Despite the uncertainties, determining this level of public adaptation requires some understanding of both the costs of relevant adaptations and the benefits they bring. Notwithstanding the problems mentioned above, this suggests that some element of cost-benefit analysis ought be applied to the problem; a comparison between the cost of adaptation investments and the benefit of avoided climate related impacts including extreme events. A rational public adaptation strategy ought therefore to attempt to map out both sides of this comparison. This analysis accounts for the fact that impacts have a time dimension, which has a bearing on how adaptation should be prioritised. Even where costs and benefits are identifiable, it is important to identify the goals and aims of adaptation and to be clear where the responsibility for action lies. While in some cases adaptation may be about preserving the status quo, in other cases an effective adaptation might be to accept that some changes will occur and decide to live with them. A decision to take no action by government may imply some level of costs to society. No action, i.e. living with change, is also likely to be a rational strategy in the face of some impacts (e.g. disease incursion) that may prove impossible to contain at low cost. Similarly in some private cases an optimal decision may be to do nothing in the face of uncertain outcomes or even perceived net gains. It is the role of government to determine whether private decisions add up to increased or reduce overall welfare to society. Agricultural gains (e.g. increased crop yields) are likely to be a reality for some agricultural producers in the UK, and ‘no action’ on adaptation may be a short term rational choice. A short term gain narrative may be attractive for many private producers who are typically unable to make business plans for events beyond the current year, let alone across decades or a century. But gains are unlikely to accrue to all agricultural producers, and the stance of government must act as a palliative to private behaviours that will lead the market to under-invest or to delay low cost investment in the face of new evidence on potential longer term impact costs. This is most important where private decisions might accentuate damage to agricultural public goods. Overcoming this market failure presents key challenges in terms of relevant information needed to influence behaviours, and the identification of the best strategy to share the cost burden of a socially efficient adaptation strategy. The latter implies the need for a better understanding of both private and social costs and benefits of adaptation. Conventional economic appraisal of adaptation is therefore not straightforward, although policy can be informed by an attempt to determine the damage costs of projected warming

3 Note that a cost-benefit appraisal is possible using the shadow price of carbon 4 No regrets options will deliver benefits that exceed their costs, whatever the extent of climate change, see

http://www.ukcip.org.uk/index.php?Itemid=253&id=128&option=com_content&task=view

  2

scenarios. In agriculture, potential adaptation decisions also need to be considered against a backdrop of early mitigation obligations, other sector reforms, and regulations that are changing the environment in which agriculture operates. For example, as part of its ‘public good’ provision role, agriculture is increasingly implicated in national food security objectives. The research reported here aimed to identify and, to the extent possible, appraise the extent of private and public investment required for adaptation in the livestock sector. Livestock production is a significant percentage of agricultural output, and there is some incentive for private adaptation investments to safeguard productivity. Under many warming scenarios, it is also possible to anticipate a number of livestock related impacts that have public good implications (e.g. greenhouse gas emissions, disease spread and the effects of grazing on ecosystems and water quality), which may therefore warrant some level of public intervention to address market failure and to maintain environmental security. In this report we use projected climate information to identify climate change impacts on the UK livestock industry and as a basis for informing an efficient adaptation response. The emphasis of this project is on gaining an understanding of the nature of the risks and trade-offs that might arise, their extent and their value, and the policy interventions that are then required. To reach this point, the report considers likely private responses and the costs and benefits of adaptation options. Our analysis is based on the premise that Defra wishes to distinguish between market-led adaptations and those where the market will fail to protect public good supply. This report is set out in the following sections that correspond with pre-defined project objectives (Appendix 1). Section two sets out the characteristics of the UK livestock sector, as a basis for considering the potential private and public vulnerabilities to climate change. Section three provides a summary of main climate change information as relevant to the UK. The section sets out the geographical boundaries of analysis for later assessment in the report. Section four provides a background to the main climate impacts affecting livestock. Section five introduces the two main responses to climate change – mitigation and adaptation, focussing on the concept of adaptation in more detail. Section six considers the nature of an efficient adaptation response. The section draws on the theoretical literature on adaptation strategies in the face of future uncertainties, and clarifies the appropriate perspectives to be taken on investment decisions. This section also details the impact categories that have been selected in this report under the headings of grazing systems, pollution and waste, welfare, and health (disease), which are then described in sections seven – ten. Each section sets out the impacts relevant to that category. Section 11 provides an analysis of the costs of the impacts identified in the previous four sections, distinguishing between private and public costs.

  3

Section 12 identifies a range of adaptation options available to producers. This options appraisal draws on industry informants and qualitative and quantitative survey information. Section 13 provides an assessment of key public responses. Section 14 develops conclusions and policy recommendations.

  4

2 Characteristics of UK livestock systems Livestock production generates more gross revenue than any other single output in UK agriculture5. The value of livestock output (i.e. livestock production and products) in 2003 was £9.2 billion, of which £5.9 billion was livestock production. The four major livestock groups (cattle, sheep, pigs and poultry) cover different geographical distributions6 and so can be expected to be affected differently by shifting temperature and precipitation. Dairying, with 2.1 million dairy cows, is generally concentrated in the lowland grass-growing areas of the west. The average herd size has increased in recent years to 83 cows/herd, with 57% of all cows kept in herds of 100 or more, although the size of the national herd has declined. Most commonly, cows will be housed during the winter and grazed during the summer, with the length of the grazing season dictated by the location’s climate. A minority of farmers house their cows throughout the year. Cows are fed a combination of ensilage forage, mostly grown on the farm, and concentrate feed. The impact of climate change may be to increase the length of the grazing season, but some farmers may opt to reduce vulnerability to weather conditions by adopting a continuous housing system. Changes in climate will also affect the availability of the cereals and other crops that make up the concentrate portion of the diet. Farmers may investigate the possibility of breeding cows for specific farming systems and conditions, or using another breed of dairy cow. The national beef cattle herd numbers 1.7 million (in 2009). Suckler cows are either a specialised continental or British beef breed or a cross between dairy and beef breeds. Suckler cows are found in both the lowlands and the uplands. Traditional systems produce finished cattle from grass with minimum concentrate feed at 24 months of age. These systems continue on the poorer quality land but there has been a gradual change to shorter indoor finishing systems elsewhere, requiring the use of higher levels of concentrate feed (Renwick and Reader, 2004). As for dairy cattle, changes in climate will affect the housing systems used and the type of forage and feed-crops that can be grown or purchased. There are approximately 18 million breeding ewes in the UK with sheep on about 84,000 farms. The diverse geography of the UK has resulted in a number of different sheep systems. In broad terms, a stratified system exists, whereby specialised hill breeds are kept in mountain and upland areas and first-cross ewes of these breeds are mated to specialised meat breeds to produce lambs for market. The hill breeds, such as Swaledale and Scottish Blackface, are very hardy, surviving on poor land and lambing outdoors, whereas the specialised meat breeds, such as Suffolk and Texel, are often lambed indoors and perform best when supplementary concentrate feed is provided. Meat is the primary product, with wool representing only about 3% of output and falling (Renwick and Reader, 2004). Climate change may alter the grazing ranges of sheep across the country and the breed of sheep used in each type of system.

5http://www.Defra.gov.uk/science/GeneticResources/ABS%20Review_final%20text_Feb%2005.pdf 6 http://www.Defra.gov.uk/animalh/diseases/vetsurveillance/reports/pdf/rp6051.pdf Source: Defra: quoted at

http://www.mdcdatum.org.uk/Farm%20Data%20&%20Prices/producernumbers.html

  5

Increased international competition in the intensive pig and poultry sectors has placed downward pressure on prices, in the face of rising costs of production associated with increased commitment to animal welfare and food safety. Pig and poultry production has become increasingly specialised. Smaller producers have gone out of business in the face of falling or variable market prices whilst remaining units have tended to get larger. Welfare considerations have had a significant effect on production. In the pig sector, total pig numbers fell by 9.8% in 2001, although the breeding herd fell by only 2%. There has also been a distinct trend towards outdoor pig production. In poultry, there has been a rapid rise in the number of table birds with a corresponding decline in the laying flock, and the popularity of free-range systems continues to increase (Renwick and Reader, 2004)). For flocks and herds that are housed, the need to provide adequate climate control within the housing will be a major issue as the British climate warms. As for the other species, sourcing of appropriate feeds may also be an issue.

  6

3 Climate change projections for the UK The current (at the time of writing this report) working scenarios of climate change impacts for the UK are provided by UKCIP, known as UKCIP02 (Hulme et al., 2002). They provide four alternative descriptions of how the UK climate might evolve over the course of this century. Differences between scenarios result from uncertainty regarding future trends and responses, such as population growth, socio-economic development and technological progress, and how these might affect future global emissions of greenhouse gases. UKCIP02 scenarios are generated from a climate model developed by the Hadley Centre in the UK. The latest climate change scenarios for the UK were released in mid-June 2009, unfortunately too late to be incorporated into this report. The UKCIP09 scenarios are discussed briefly at the end of this section. The four different UKCIP02 scenarios are based on different levels of future emissions (low emissions, medium-low emissions, medium-high emissions, and high emissions). The analysis in this report uses the medium-high scenario as the basis for assessing the impacts.

3.1 Main messages for the UK relevant to livestock production and health The thermal growing season for plants has increased by up to 30 days since 1900 and is expected to lengthen, but soil moisture levels in summer and autumn are expected to decrease. By the middle of this century, average annual temperature for the UK is expected to increase by 0.5 and 1 degree C, depending on region. However average annual temperature masks seasonal differences. In the UK, there is expected to be greater warming in the summer and autumn than in winter and spring. By 2050, average summer temperature for the UK is expected to have risen by between 0.5 and 2˚C, depending on region, while the average winter temperature in the UK is expected to have risen by between 0.5 and 1˚C. Winters are expected to become wetter, with rain occurring in more intense bursts. The south east is expected to experience changes earlier than other regions, and the north of England and Scotland are not expected to experience changes until the middle or end of the century. Figure 3.1 and Figure 3.2 illustrate the projected change in the thermal growing season by 2020 and 2050 respectively, which has important implications for livestock production. By the 2020s, there will be some changes in growing season days, depending on scenario. By the 2050s, the increase in growing season days is projected to be much more pronounced, with some areas increasing the length of the growing season by 50 – 60 days under a high emissions scenario. The term “growing season” in this context only refers to temperature and does not account for water availability or day-length. Drier summers together with the same daylight hours as today’s may mean that many plants will not be able to take advantage of the longer theoretical growing season.

  7

Figure 3.1 Change in thermal growing season (2020)

Figure 3.2 Change in thermal growing season (2050) More detail and more graphs can be obtained on the UKCIP website (www.ukcip.org.uk) and the UKCIP02 report (Hulme et al., 2002) The greatest impacts of climate change in the short term are likely to be from extreme weather events such as floods, droughts, heat waves and windstorms. These are expected in increase in both frequency and intensity; they are more difficult to map and therefore more difficult to adapt to than gradual warming.

  8

The scenarios provide information at 50x50km grid squares, and highlight (among other things) the regional variation of impacts in the UK. This variation indicates that adaptation will also need to be local and appropriate to the impacts experienced. A one-size-fits-all approach to adaptation is not likely to be very constructive.

3.2 UKCP09 As discussed earlier, where necessary this project has used the UKCIP02 Climate Projections (medium-high scenario) in the modelling. The UK Climate Projections team has recently released a fifth generation of climate projections, UKCP09. The methodology used for UKCP09 differs fundamentally from that used for UKCIP02, which makes a direct comparison between the two difficult. The climate projections in UKCP09 are based on probabilistic modelling compared to deterministic modelling in UKCIP02. Also the UKCP09 and UKCIP02 projections use different emission scenarios, three versus four respectively. Table 3.1 highlights some of the differences between the two projection scenarios. This project could not anticipate the challenge presented by these differences, and so we have been unable to incorporate these new data. Table 3.1 Differences in the methodology employed in UKCIP02 and UKCP09 (taken from ukclimateprojections.defra.gov.uk)

Feature UKCIP02 UKCP09 Time periods 2020s, 2050s, 2080s 7 decadal overlapping 30 year time slices from 2010-2099 Spatial resolution 50km grid 25 km grid Temporal resolution Daily, monthly seasonal,

annual Daily, monthly, seasonal, annual Weather generator for daily output

Aggregation UK level UK English administrative regions Devolved administrations River based regions

Modelling approach and quantification of uncertainty

Single UK model (HadCM3) 4 emissions scenarios

Ensemble approach: model parameter uncertainty explored in HadSM3, using perturbed physics experiments (PPE). PPE used to construct a statistical emulator from which a greater range of variability in the climate model parameters can be considered. Additional IPCC climate models included (MME) (model uncertainty). 3 emissions scenarios: low (B1), medium (A1B), and high (A1F1). Statistical framework used to produce probabilistic projections, users will have a probability associated with each projection

Outputs Single mean value of projections of 15 climate variables

Probabilistic projections of same climate variables as UKCIP02. Observed trends in recent climate. Marine projections

Delivery UKCIP02 report Maps, underlying model output

3 science reports: land, marine, historical trends. Web-based user interface to provide customisable output Climate model output and associated probability Dedicated website

  9

4 Livestock and climate change impacts The impacts of climate change on livestock systems will be spatially and temporally diverse. Broad categories of impacts7 include drought, and variations in the length of growing seasons for grasses and forage crops. Further, altered ranges for pathogens and pests are likely to increase overall disease burdens, and present associated challenges in terms of animal welfare impacts. On a longer time scale, livestock producers are likely to need to contend with a higher incidence of extreme events such as fluvial and pluvial flooding. It is possible to distinguish between direct and indirect effects of climate on livestock production. The direct effects on livestock include impacts on animal health, welfare, growth and reproduction, while the indirect effects are due to the impact of climate change on the productivity of pastures and forage crops. A more complex indirect impact may result from the impacts of climate change on the economic cost of inputs, e.g. feedstocks, that are imported into UK systems from global markets. These effects are not considered further in this report, which focuses primarily on the direct effects on animal health and welfare and indirect impacts via grasslands. Both forms of impact can be addressed by a range of adaptations that in turn imply further impacts that may be external to the agent undertaking the adaptation. This ancillary adaptation impact is of interest in this project, although its identification clearly requires a robust scenario of what private adaptation will take place.

4.1

Existing research of impacts on agriculture Considerable research has been carried out on the primary effects on, and interactions of climate change with, plant growth and yield. Much of this research is summarised in the ‘Food, Fibre and Forest Products’ chapter of the Fourth Assessment Report (AR4) of the IPCC (Easterling et al 2007). Less research has been carried out on the impacts of climate change on pastures and livestock, which is again summarised in Easterling et al (2007). A number of studies have been carried out assessing the impact of past extreme weather events (particularly heat waves) on the agricultural sector in the UK and Europe (Ciais et al., 2005; COPA COGECA 2003., Orson, 1996; Subak, 1997; European Commission, 2009; Hunt et al., 2006). These provide useful indicators of what might occur in the future. In the UK, the warming associated with climate change during historically cooler periods (i.e winter) is likely to reduce feed requirements, increase survival and lower energy costs (Maracchi et al., 2005). However, warming in the warm periods of the year may result in heat stress, which can result in reductions in the fertility of cows and sows, and can also result in animal welfare issues (MAFF project CC0325, Furguay, 1989). The impact on grassland-based livestock production is expected to be larger than on pigs, who are more adaptable to changes in temperature. Results of a previous study (MAFF CC0325) showed that climate change would also impact on the seasonality of parasites.

7 There are positive impacts associated with increased CO2 in the atmosphere (CO2 fertilisation),

which will influence yield projections via improved grazing, and need to be considered as they may offset the negative impacts on overall productivity.

  10

Within the UK, grasslands vary from intensively managed monocultures to species-rich communities, and the response of these grasslands to climate change will be species dependent (Olsen and Bindi, 2002). This will have implications for animal nutrition. The response of intensively managed grasslands to climate change will be affected by the grazing and cutting management regime adopted by the farmer (ibid). In nitrogen-poor systems and species-rich communities, experimental studies have shown little increase or even a reduction in productivity with CO2 enrichment (Korner, 1996). On the other hand, using a simulation model Cannell and Thornley (1998) predict that the response observed by Korner (1996) is transitory, and that the long-term response is relatively larger compared to the N-rich systems. The Defra commissioned study CC0359 shows that there would be a net increase in dry matter yields, typically in the region of 10-25% for the 2020s scenarios for grass production for dairy, lowland beef and sheep, and upland beef and sheep farms. However, the increases in yields are likely to be skewed towards the beginning of the season, with a reduction in the yield later on. Legume swards tend to show a larger increase than fertilised swards, which may result in an increase in the use of alternative forages. Although the increase in the temperature arising from climate change is likely to increase the length of the growing season and thus the productivity of the sward, the length of the grazing period (as modelled) tends to be reduced overall. This is mainly due to the increase in soil moisture and hence the risk of poaching damage (the soil structure becoming damaged and soil becoming muddy). This study also reveals a relatively small effect of climate change on silage quality as measured by digestibility (D value) of water soluble carbohydrates. Muriel et al (2000) summarise the main impacts of climate change on agriculture in the UK. The report highlights that intensive grazing systems are likely to be quite sensitive to climate change, particularly through their sensitivity to water supply and drainage. Extensive systems are likely to respond more slowly to impacts, but the differences in response between species are likely to lead to a change in species composition. A more favourable climate in the uplands and a longer growing season in general may lead to an expansion of intensive grazing, with the associated increases in fertiliser application. This in turn suggests potential problems in terms of diffuse pollution to water and air. Climate change is also expected to increase the number of extreme meteorological events, including high temperatures, heavy storms and droughts. Defra CC0360 showed that the cutting and grazing management of the grassland pasture is critical when extreme meteorological conditions are experienced. That study found that hot dry summers had little impact on the grazing days, but wet springs and autumns could reduce the number of grazing days substantially. Hot dry conditions during the grazing period can reduce the quality and consistency of the feed, which may lead to animal welfare and productivity problems. The health and welfare of farmed livestock are affected by management systems, the type of housing, transport and the feeding system on the farm. The predicted levels of climate change will have implications for all of these factors, and how farmers respond to the changes will determine the overall impacts. Adaptation is discussed in detail in subsequent sections of this report. IPCC (2007) highlights that thermal stress reduces productivity, conception rates and is potentially life-threatening for livestock in conventional production systems and even more

  11

so for animals being transported. High temperatures put an upper bound on milk production, regardless of feed intake. While animals can adapt to gradual changes in temperature, rapid increases in air temperature and/or humidity, such as in a heat wave, may affect conception rates and production (Amundson et al 2005). There are species specific health, welfare and environmental issues associated with each of these climate change impacts. Some will be the responsibility of producers, others, due to the external costs they impose, may warrant government intervention. Such intervention may be in terms of new regulation or adjustment to existing regulation. Information on the welfare of animals in farming systems used in countries whose current climates are similar to those predicted for the UK can be used to determine the impacts of the predicted changes on the welfare of livestock. However, all of these impacts are subject to considerable uncertainty, which is compounded by the ways in which society is anticipating these changes and responding to them.

  12

5 Climate change responses Two distinct but related responses strategies have been adopted in the face of climate change. To prevent the worst impacts of climate change and to minimise future impacts, mitigation strategies have been developed, which aim to reduce the emissions of greenhouse gases (GHGs) into the atmosphere. Adaptation on the other hand, is concerned with coping with the unavoidable changes that will occur, even if mitigation is successful at minimising or avoiding the worst impacts. Mitigation obligations are becoming more clearly defined by national compliance with externally determined emissions reductions. In contrast, there is no similar obligation to adapt, although under the Climate Change Act (2008), the UK government has made provision for a periodic risk assessment that will inform priorities for intervention. These is also some recognition of the links between adaptation and mitigation actions. Adaptation was previously regarded as ‘giving-up’ on mitigation and accepting that greenhouse gas abatement targets would not be met. However, due to historical emissions and inertia in the climate system, the world is now committed to a certain degree of change, regardless of the level of mitigation, and impacts are already being experienced. Both mitigation and adaptation are important and complementary tools for tackling climate change, and in some areas the distinction between the two is less clear.

5.1 Adaptation Adaptation has been defined broadly as the ‘adjustment in natural or human systems in response to actual or expected climatic stimuli or their effects, which moderates harm or exploits beneficial opportunities’ (IPCC 2001). Adaptation can focus either on managing the impacts of the climate-related hazard, or reducing the vulnerability of the elements at risk, or both. Further definitions of adaptation based on their timing and responsibility are listed below: Anticipatory adaptation – adaptation that takes place before the impacts of climate

change are observed Reactive adaptation – adaptation that takes place in response to an event, after it

has occurred Autonomous adaptation – adaptation that does not constitute a conscious

response to climatic stimuli but is triggered by ecological changes in natural systems and by market or welfare changes in human systems. Also referred to as spontaneous adaptation

Planned adaptation – adaptation that is the result of a deliberate policy decision, based on an awareness that conditions have changed or are likely to change and that action is required to return to, maintain, or achieve a desired state.

Adaptations can also be defined as public or private, depending on the actors involved. Planned adaptations tend to be made in the public sector, but not exclusively. Distinguishing between the process (building adaptive capacity) and the outcome (the delivery of actual adaptation measures) of adaptation can also be useful and relates to the private/public domains of responsibility. In a private economic activity like livestock

  13

production, there is a private incentive to invest in outcome measures. But there is a public interest in the way this adaptation takes place, and so a public role might be restricted to capacity building through the provision of information targeted at fostering appropriate voluntary private action. An anticipatory strategy requires a certain amount of voluntary action by private agents. This can best be fostered through a process of information provision, which can:

Build adaptive capacity through information and R&D policy Increase resilience to reduce sensitivity to climate change Reduce exposure to climate-related risks Promote understanding and acceptance that there will be impacts and

corresponding losses for which the sector should take responsibility for.

5.2 What is the aim of adaptation? At the outset, the goals and aims of adaptation should be identified. While in some cases adaptation may be about preserving the status quo, in other cases an effective adaptation might be to accept that some changes will occur and decide to live with them. However, while for some impacts this might be a decision left to individuals, other decisions affecting agriculture will be made at a higher level. Who makes these decisions and what they are based on (economics, values, societal expectations) can have an important impact on individuals and the sector generally. The potential entry of livestock diseases is a case in point. For some diseases, a distinction between anticipatory (ex ante) and reactive (ex post) responses needs to be informed by an assessment of which diseases are worth stopping - i.e. some diseases will be exacerbated by climate change but their prevention may not be economically worthwhile. While surveillance is likely to be the strategy in both cases, recent experience suggests that ex post responses can be very costly involving containment and wider economic and social disruption to the industry and in the countryside. There is therefore a premium on improving anticipatory capacity as a climate change adaptation. Adaptation is often referred to as being as much a process as an outcome, and the ongoing nature of adaptation is one of the reasons for standard economic techniques for estimating costs and benefits no longer being appropriate. As the climate continues changing, adaptation must continue evolving. We are unlikely to reach a state where we are “adapted” (or “climate-proofed”); indeed, we are not adapted to the current climate, as evidenced by continual damage associated with flooding, droughts and heat waves. The level of damage we are prepared to accept is a societal decision, and will vary spatially and temporally. As mentioned, there is a public role in making this clear and in delineating responsibilities for resulting costs, and fostering the development of better adaptive capacity; the ability of individuals, systems, or society to adapt to changes in climate. Adaptive capacity is shaped by available resources, institutions, skills and knowledge, and refers specifically to the ability to adapt, rather than manage risks (this assumes that the impacts will occur). In faming, this comprises the product of a managed biological system, socio-economic characteristics of farmers and land owners, and a given set of institutions.

  14

Adaptation in agriculture 5.3 The literature on assessing the adaptive capacity in agriculture is scarce, but tends to concentrate on the limited capacity in the developing world. In theory, agriculture in the UK has a high adaptive capacity since it is financially strong, is supported by government and industry institutions, has a large research sector behind it and actors who are relatively well educated. In the UK, as in much of Europe, the sector has historically developed adaptation strategies in response to variable market conditions, including the need to alter practices in response to changing subsidy regimes. But having a high adaptive capacity does not necessarily translate into action, especially if there are significant information or attitudinal barriers to adaptation in the sector. Research on the extent of these barriers in the context of climate change is currently very limited, but see Adger et al (2009) for examples across sectors. These possible barriers may undermine the potential adaptive capacity of the sector. Some barriers may be structural or even physical, such as being locked-in to certain production practices because of the physical characteristics of the land or because of restrictions on land use. Other barriers may lie in the behavioural characteristics of producers and their attitudes to risk. Producers have many competing demands on their resources and are unlikely to spend time and money on actions that have uncertain future benefits. There appear to be very few, if any, adaptations that have been undertaken solely in response to expected climate change. This is in clear contrast to reported mitigation actions such as investment in biofuels as a contribution to renewable energy. This result is common throughout the world, although particular events have sometimes prompted adaptation responses. In Canada, most individual farmers respond primarily to extreme events such as prolonged drought and unseasonal or excessive rainfall. In a survey in Ontario, 80 percent of respondent farmers judged extreme events to be the most significant impact to which adaptation was required, rather than changing growing season length or heat stress (Smit et al., 1996). The study also documented barriers to adaptation, including lack of knowledge of water supplies and water use. An individual’s perception of the risks associated with natural hazards declines over time (Wheaton et al 2007). The more time that elapses since the last drought, the less likely it is that producers will adopt measures to reduce their exposure to the drought. Agriculture is continually adapting to changing conditions, in response to weather, policy, market or social conditions, and there may be no reason to expect that climate change will be any different. However the concern with regard to the emerging evidence on climate change is the rate and magnitude of that change. In the UK however, the projected changes are likely to have generally positive effects on agriculture, with a longer growing season, possible CO2 fertilisation effects, and the possibility to expand productive areas. Probably the greatest threat to UK agriculture from climate change in the shorter term lies in the increased incidence of extreme weather events.

  15

Public/private interest 5.4

5.5

An important component in adaptation research is identifying the role for the public and private sectors respectively. This will partly be determined by an assessment of the costs and benefits involved, but also by the sector and the aims and objectives of adaptation as defined by the decision makers. In some senses mitigation delivers a global public good, while it is often argued that adaptation delivers local benefits and is therefore much more a private good. However while there is a private incentive to adapt in many cases (i.e. to avoid damage to one’s own property, or to avoid production losses resulting from climate change), there are other necessary adaptations that would not be undertaken without public support. Public goods and issues around property rights are likely to hinder private action. Aside from providing adaptation actions directly, government can aid private actors by providing information about the likely changes, the impacts, and options for adaptation. It has been suggested (Berkhout, 2005) that there are seven objectives for public policy adaptation: to inform the potentially vulnerable; to assist in the provision of disaster relief; to provide incentives for and enable adaptation; to mainstream climate-proofing of public policy; to plan and regulate long-term infrastructural assets to reduce future vulnerabilities; to regulate adaptation ‘spillovers’; and to compensate for the unequal distribution of climate impacts. In the context of agriculture the aforementioned objective of regulating adaptation spillovers is an area where public involvement in agricultural adaptation may be appropriate. Collective private actions may lead to unintended public good consequences, such as impacts on biodiversity or water quality. These may be positive or negative. If positive then government may have a role to play in providing incentives, and if negative then regulation and information about alternatives may be appropriate.

Adaptation in the livestock sector This section reviews the recent literature relating to adaptation in the livestock sector in the UK. Research relating to livestock sector adaptation can be broken down into practical management options for producers, and higher level policy strategies. The IPCC summarises recent adaptation research, both in Easterling et al (2007) and Adger et al (2007). There is however, relatively little relating specifically to UK livestock systems, although many of the principles are applicable in a range of settings. The IPCC reports do not provide any information on the costs of adaptations. In the UK, Hossell et al. (2002), (MAFF CC0357) identify and cost selected responses of agriculture to climate change. For the livestock sector they assess adaptation for grass systems for dairy production, and pig production. Where quantitative information is available, adaptations are analysed using cost-benefit analysis. This is based on representative farm units and then aggregated up to the industry level. This approach provides a useful indication of the likely costs and benefits of certain adaptation options over a specific time period, but does not account for differing uptakes of adaptation, timings, or uncertainty in impacts.

  16

More recently, also in a project for Defra, Hughes et al. (2008) (CC0361) assess adaptation to extreme events in UK agriculture, and provide cost estimates of selected adaptation options for representative farming enterprises. They include several options relevant to the livestock sector, and use farmer focus groups to identify the impacts that extreme weather has already had on agriculture. They also use an expert workshop to identify a shortlist of adaptation options. The results from CC0361 could be usefully combined with the findings from this study to form a more complete picture of adaptation options, covering both gradual climate change as well as extreme weather events. AEA (2007) provide an assessment of adaptation options for agriculture in Europe, including an assessment of the role of policy in facilitating adaptation, although they do not provide cost estimates. In terms of developing a consistent policy framework the EU has recently released a White Paper on Adapting Europe to Climate Change, with a supplementary working document on Adapting to climate change: the challenge for European agriculture and rural areas (European Commission, 2009). In the document it is stressed that planned public adaptation strategies can enhance farmers’ awareness of the projected changes, encourage early action and facilitate appropriate responses and solutions with long-term viability. It is also highlighted that the Common Agricultural Policy (CAP) has an important role to play in supporting adaptation, and that currently there may be perverse incentives encouraging production practices which may be inappropriate under a changing climate. Ideally all policies, but particularly the CAP, should have adaptation integrated or “mainstreamed” throughout, in order to ensure there are no mixed signals for producers. This is particularly true of the measures contained under the climate change headings allowable under the Rural Development Programme (RDP) regulations. Currently RDP measures are focussed predominantly on mitigation actions rather than adaptation. This focus leads to potential conflicts between mitigation and adaptation.

5.6

Adaptation and impact assessments appraisal The basic impact uncertainties in agriculture help explain why there are few cost-benefit studies of adaptation plans. This section considers the nature of cost-benefit analysis applicable in the context of investment uncertainty8. Prior to this however, it is worthwhile considering the assumed adaptation strategies implicit in existing impact or damage assessments. Existing impact assessments use assumptions about what adaptations to build into future damage trajectories. These assumptions amount to behavioural reactions that assume agents either do nothing to head off damages, or that they demonstrate more foresight in their anticipation of impacts. The relevance of these strategies lies in the extent to which the assumptions under- or over-estimate costs of impacts, and by extension under- or over-estimate the levels of autonomous adaptation. Impact estimates have traditionally included adaptation in one of four main ways:

8 Investment uncertainty in this context relates to the nature of benefits or avoided costs as a result

of adaptation spending. However, climate change is not unique in presenting this problem.

  17

No adaptation ‘dumb farmer’ Many estimates of the impacts of climate change in agriculture (Parry et al., 2004; Parry et al., 2005; Rosenzweig & Parry, 1994) do not include adaptation, and have become known as the “dumb farmer” approach. Ignoring adaptation can lead to an over-estimation of the damage of climate change. In these studies, agronomic research is typically used to estimate the impact of climate change parameters on particular crops in order to extrapolate to wider environments and situations with an altered climate. As well as over-estimations of damages, the approach also conveys the message that there are no actions available in the face of climate change and the only option is to mitigate emissions or suffer serious consequences. Complete adaptation ‘clairvoyant farmer’ The clairvoyant farmer contrasts to the dumb farmer perspective. Here, near perfect climate foresight is implicit in the ways in which farmers configure inputs (e.g. crop choice) to deliver maximum output value. This value in turn shows up (or is capitalised) in the value of land. Under the so-called Ricardian approach, variation in capital values is taken as a reflection of the economic costs of climate change. All else being equal, the implicit climate damage is shown in the price variation between two identical parcels of land. While this approach is useful for providing empirical impact cost estimates, these are largely abstract from specific adaptation options, their cost, or effectiveness, or any constraints to their implementation (Rosenzweig & Tubiello, 2007). Arbitrary adaptation Other studies have assumed that some level of adaptation will take place and have attempted to incorporate it into the analysis at some arbitrary level, such as including “no adaptation, some level of adaptation, and full adaptation” as various options for comparing costs. This type of analysis can be very useful and may provide important information regarding the upper and lower bounds of costs. In the agricultural sector, Easterling et al (1993) and Rosenzweig and Parry (1994) provide estimates of the potential of various arbitrary levels of adaptation to reduce damages of climate change at a global level. Observed adaptation Observed adaptation uses examples from other situations as predictions of adaptation in the current situation. These analogues may be spatial or temporal. Spatial analogues use the experiences and actions in one location as examples or predictions of possible action in another, similar location. Doing so can provide valuable insights but in terms of economics, the process and hence the costs of changing state are ignored (Tol et al., 1998). Mendelsohn et al. (1994) use this method to estimate climate change impacts on US agriculture. Temporal analogues examine how adaptation has occurred historically (Wreford et al 2009). Recent large scale analyses of the costs of adaptation to the global economy include a report by the UNFCCC (UNFCCC, 2007), which studies the investment and financial flows required to address climate change. This report found that around USD 14 billion in

  18

investment and financial flows would be needed to adapt agriculture, forestry and fisheries to climate change. In summary, it is clear that adaptation is incorporated to differing extents in the existing literature on impact assessment. The same literature is much less specific in providing a useful categorisation of adaptation options that might be transferred to the UK context. The differing treatment of adaptation in impact costing is also problematic for deriving a clear picture of the benefits (or costs of inaction).

5.7 Costs of adaptation Impacts that occur after adaptation has taken place are known in climate change literature as residual impacts, the impacts that society on some level has decided are acceptable. Residual impacts make costing adaptation difficult as these must somehow be netted out of impact costing; i.e. not all the impacts will be avoided, therefore the cost of inaction does not necessarily translate directly into the benefits of adaptation. Further, because of these different adaptation behaviours, the inaction baseline may in fact be changing over time as climate change impacts are routinely absorbed into management practices (or adapted to). Impact estimates assume that no adaptation occurs at all (the dumb farmer approach), which is an over-estimation of damages if farmers will adapt autonomously to changes over time. However an estimate of these impacts without any form of adaptation is necessary in order to compare with impacts after adaptation, so that the benefits of adaptation can be quantified and highlighted to illustrate the importance of adaptation. Furthermore, many adaptation actions may have non-climate ancillary benefits, which may need to be taken into account in the valuation of net impacts. These further complicate any notion of efficient adaptation. Figure 5.1 illustrates the costs of climate impacts over time, for no adaptation (dashed line), with adaptation (solid line), and the baseline scenario of impacts with no climate change (dotted line). The baseline is increasing because the value of production and assets is assumed to increase over time. The difference between the solid and dashed lines represents the benefits of adaptation, while the difference between the dotted and solid lines represents residual impacts, which will not be able to be adapted to. Residual impacts will vary both temporally and spatially.

  19

Baseline scenario of flooding costs

Residual impacts

Future impacts less adaptation

Benefits of adaptation

Costs £

Unregulated impacts

n n+10 n+20 n+30 time

Figure 5.1 Costs and benefits of adaptation Literature on the costs of agricultural adaptation is limited. This may in part be because the focus of adaptation is on farm-level adjustments such as changes in timing of planting, or crop choices that are low cost. It is also the case that the lack of a distinction between public and private responsibilities has given rise to inertia in defining cost data and deriving overall estimates. In summary, adaptation comes in many forms, and may take place ex ante or ex post, by private or public agents, and may have a tangible outcome or consist more of building capacity. Assumptions made about levels of adaptation can have important implications for the assessment of impacts arising from climate change. While many adaptations may occur autonomously, the existence of barriers may impede effective adaptation. Ideally these should be identified and removed. Information on costs of adaptation is currently limited, and reasons for this are explored in subsequent sections.

  20

6 Cost-benefit analysis (CBA) under uncertainty

6.1

Defining efficient adaptation under uncertainty Uncertainty is inherent in almost all dimensions of climate change. There is uncertainty in the rate, timing and magnitude of climate change itself, and around the biological response to these changes. There are tiers of uncertainty up to the costing of climate impacts that complicate specific adaptation planning (figure 6.1). Beyond the impact costing there is further uncertainty surrounding how society will respond or its capacity to adapt to the projected impacts.

Figure 6.1 Widening uncertainty Many decision-makers are unwilling to make adaptations until they see evidence of a changing climate, or can be convinced that the climate projections are “accurate” and reliable. However while some uncertainties can be quantified, others cannot, and waiting for more certain projections, or evidence of change, may mean that decisions made in the meantime lead to greater vulnerability to future changes, and possibly higher future costs as well. Decision errors can be made in both directions with regard to adaptation to climate change; both over- and under-adaptation may occur depending on the relative significance accorded climate change and non-climate factors. Uncertainty presents a major challenge to adaptation. The benefits of adaptation are uncertain in terms of when (if) the benefits (avoided damages) will occur, and how large they will be. On the other hand, the costs of adaptation are relatively observable, may be high, and are likely to need to be paid now. Watkiss et al (2009) suggest that these conditions handicap the use of traditional cost-benefit analysis, which is typically dependent on short term horizons and observable cost and benefit streams. In these cases techniques for decision-making under uncertainty must be employed. UKCIP (Metroeconomica (2004) and Willows and Connell (2003)) offer guidelines for carrying out these techniques. Beyond cost-benefit analysis, other decision support tools exist. These include cost-effectiveness analysis (CEA), and multi-criteria analysis (MCA). CEA can be used to

global climate sensitivity ⇒

emission scenarios

carbon cycle response

regional climate change

scenarios

range of possible impacts

⇒ ⇒ ⇒

  21

identify the ‘least cost’ adaptation response to achieve a specific level of climate risk management, and can be used to consider cost-based approaches to valuing the economic benefits foregone through damage caused by climate change (Metroeconomica 2004). It cannot be used to compare options that provide different outputs, and because the benefits are usually measured in non-monetary units, it does not work so well when different options yield outcomes that are measured in different units. MCA is an alternative to monetary appraisal methods and is typically used to screen options defined by non-monetary criteria or attributes. The analysis is not limited to economic efficiency criteria, and it allows climate impacts to be measured in non-monetary units. MCA is not an alternative to CBA, but it allows broader criteria to be accounted for in addition to the CBA framework.

6.2 Project approach to adaptation appraisal There are clearly many challenges in presenting an accurate ex ante cost-benefit appraisal of livestock sector adaptation to climate change. While such an exercise is instructive in highlighting likely information needs, as an evidence-base the resulting figures will inevitably be debatable. How then should we approach appraisal rationally in the context of uncertainty? Adaptation decisions need to be robust in the face of uncertainty, that is, they should be relevant for a range of future climate impacts. Ideally, actions will create win-win (no regrets) situations, in terms of decreasing vulnerability to current variability while simultaneously increasing (or maintaining) farm profitability. Actions that increase the adaptive capacity of the sector generally, rather than specific actions in anticipation of uncertain events, are likely to be more cost-effective. Actions that allow flexibility for further action or reversibility when impacts become more certain are also important. Hallegatte (2009) identifies six attributes that can contribute to the robustness of adaptation options. In addition to the win-win and flexible actions, he stresses the importance of “safety margin” strategies that allow enough flexibility when designing new structures or systems to allow for the worst case scenario. Alternatively, so called “soft” strategies are those that do not involve engineering or infrastructure but involve institutional or financial tools, which generally have greater flexibility or reversibility than hard adaptation options. Reducing the time horizons in decision-making is also a strategy such that decision-makers are not locked on to a path that proves to be inappropriate in a changing climate. Finally, identifying synergies and conflicts with mitigation are important considerations for making adaptation actions robust. This is likely to be an important focus in the livestock context. Examples of adaptation options in the agricultural sector with an evaluation of their robustness in the face of uncertainty are shown in Table 6.1:

  22

Table 6.1 Robustness of adaptation options (from Hallegatte 2009) Examples of adaptation options

No regret strategy

Reversible/ flexible

Existence of cheap safety margins

Soft strategy

Reduced decision horizon

Synergies with mitigation

Developing agricultural insurance tools

+ + +

Irrigation + - + Forestry with shorter rotation time

- - +

Development of resistant crops and animals

++

Section 13 of this report will use these criteria to screen specific livestock adaptations.

  23

Impacts storylines 6.3 An appraisal of adaptation needs is determined by a survey of key risks and impacts plus the availability of relevant data to quantify these and assign monetary values to expected damages. Table 2-1 in Watkiss et al (2009) identifies key gaps across significant sectors of the UK economy. While some qualitative information is available for impacts on livestock there are notable gaps in relation to valuation and adaptation options. For this study it is possible to identify a range of impacts affecting private production and the potential generation of external impacts from private adaptation responses. The aim of this analysis is to determine whether these impacts are significant. This report divides the analysis into the following sections

i) Grassland potential under alternated climatic conditions and the effects on the length of grazing periods ii) Alternative private productivity and waste generation associated with increased grassland consumption

iii) Welfare from in-field, housed and transportation conditions

iv) Disease

Consideration of i) and ii) provides a basis for considering external costs in terms of greenhouse gas emissions, water quality impacts and potentially to biodiversity. However, the extent of this analysis remains largely qualitative. Each impact category uses climate data differently; categories i-iii) are informed by downscaled climate information and the impact of extreme events. Uncertainty in relation to disease impacts means that this section is largely qualitative.

  24

7 Impacts of climate change on grazing systems

7.1

7.2

Introduction Grassland is the main productive driver of cattle and sheep systems and is sensitive to climate and hence any changes in the climatic conditions. The climate change scenarios will determine the productive potential for grassland systems of the future and so will drive any change in ruminant livestock distribution and production. The effects of climate change on grassland systems are potentially complex with effects on forage yields and quality, which may affect the relative suitability of grasses and legumes for grazing and fodder and thus their utilisation. Climate change is likely to have impacts on the length of the growing and grazing seasons, and animal production, which will have consequences for environmental pollution. In addition, changes in fertiliser strategies resulting from changes in growth and utilisation patterns, and the possibility of planting shelter / shade belts will have implications for biodiversity. The objective of this section is to explore the response of ruminant production from different regions within the UK to climate change. The impacts are explored for 2020, 2050 and 2080.

Models To assess the impact of climate change on livestock systems an assessment has been carried out of how grassland production would be affected. A grassland model was used to predict the changes in forage production and length of grazing season for dairy, beef and sheep systems in the UK. The output from this model was used as the input to farm system models, and the models of environmental pollution, which are described in section 8. Grassland production A SAC grassland systems model (Molle et al. 2006) has been used to predict the impact of climate change on grassland production used for livestock production within the UK. The model of the sward developed allows both legume and grass-legume mixtures to be explored, and is described in Appendix 2. Nutrition The outputs of the grass production models were used as inputs to the farm models. These outputs were programmed in Excel, were based on the means of 50 simulations per region per time point (i.e. baseline, 2020, 2050 and 2080) when grass cuts would occur (day of the year), the dry matter yield (DMY) of the grass, the grass quality characteristics, digestibility (D), crude protein (CP) content and the variation in these characteristics as indicated by their standard deviations. Details of the nutrition model are described in Appendix 2. Dairy farms All dairy models are based on a standard cow with a body weight of 600 kg and an annual milk yield of 8000 kg (peak yield 44.6 kg) with a negative energy balance at the start of

  25

lactation and positive energy balance later in lactation (no net change in liveweight during the year). Standard rules on milk composition, net energy (NE) requirements for maintenance, pregnancy, milk yield, body weight change and efficiency of ME utilisation were used to predict cows’ ME requirements. Simulated farms were assumed to have year-round calving with a lactation period of 44 weeks and 8-week dry period. For each week the ME requirements of each cow were calculated. Standard rules on concentrate requirements, based on stage of lactation and type of forage available, were used to formulate appropriate diets for each cow in each week. From the availability of grass during each week of the grazing season, grass ME content, cows’ ME requirements and concentrate to forage ratio in the cows’ diets, the minimum number of hectares under grass was calculated for a standard 52-cow herd (one cow calving each week). This number was used to predict the amount of silage produced from this area alone (i.e. grass grown but not required in the cows’ diets during the grazing period for this area). From the difference between this number and total yearly silage requirements, the additional acreage required for silage production was calculated. Stocking rate was calculated as 52 divided by the total area required under grass. Silage quality was derived from the (weighted) mean grass quality. Using standard rules, the requirements for CP of the herd was calculated weekly. The required CP content (with a minimum of 12%) of the concentrates was calculated from animal requirements and the predicted CP content of the forage available in each week. This allowed calculation of the difference between the yearly import onto the farm of concentrate nitrogen (N) and the yearly export of N with milk as an indicator of efficiency of N utilisation (but note that this does not include N in fertiliser, which should be an outcome of the grass growth model). Given time constraints, the models were limited to the actual herd of standard dairy cows because a preliminary exercise suggested that including heifer and follower modules would not essentially change the conclusions. Beef farms All beef models are based on a standard cow with a body weight of 600 kg and an annual milk yield of 1400 kg (peak yield 9 kg) in 210 days with a small negative energy balance at the start of lactation and positive energy balance later in the lactation (no net change during the year). Standard rules on milk composition, net energy (NE) requirements for maintenance, pregnancy, milk yield, body weight change and efficiency of ME utilisation were used to predict cows’ ME requirements. All cows were assumed to produce one 40 kg calf per year in spring (end March) that were weaned at 210 days weighing 210 kg. Standard rules on milk composition, NE requirements for maintenance and growth and efficiency of ME utilisation were used to predict calves’ ME requirements from grass until weaning. It was assumed that calves left the farm after weaning. It was assumed that the systems would be entirely forage based, i.e. the very small quantities of concentrates fed (mainly in the form of mineral supplements) were ignored in the modelling. Sheep farms All sheep models are based on standard ewes with body weights of 75 kg and an annual milk yield of 210 kg (peak yield 3.5 kg) in 16 weeks with a small negative energy balance at the start of lactation and positive energy balance later in lactation (no net change during the year). Standard rules on milk composition, net energy (NE) requirements for maintenance, pregnancy, milk yield, body weight change and efficiency of ME utilisation

  26

were used to predict ewes’ ME requirements. All ewes were assumed to produce two lambs (birth weight 4 kg each) per year in spring (first week March) that were weaned at 16 weeks (weighing 32 kg) and removed from the farm at 28 weeks of age, with a weight of 42.5 kg. Standard rules on milk composition, NE requirements for maintenance and growth and efficiency of ME utilisation were used to predict lambs’ ME requirements from grass until weaning. It was assumed that lambs consumed forage only from weaning until they left the farm. It was assumed that the systems would be entirely forage based, i.e. the very small quantities of concentrates fed (mainly in the form of mineral supplements) were ignored in the modelling.

7.3 Scenarios Selection of the regions for the modelling and baseline data The impact of climate change on grazing systems may differ between regions of the UK. Hence, the grassland and the nutrition models were run for each of the Rural Development Programme (RDP) regions in England plus Scotland and Wales. The grassland model requires daily weather data whereas UKCIP02 only produces monthly means of the data for 50*50 km2. In addition to temporal downscaling, June agricultural census data (Defra) for 2003 was used to identify separate dominant areas within each RDP for dairy, beef and sheep (10*25km2) in each of the RDP regions (see Appendix 2). Using the UKCIP02 medium-high scenario, the weather generator Earwig (Kilsby et al., 2007) was used to generate daily climate data for a period of 50 years for each of the livestock types zones within the RDP and for each of the climate change time slices. This data was then used as the climate data for the grassland model. The annual rainfall data for the dairy regions in each of the RDPs, and the monthly means for the south of the country are illustrated in Appendix 2, Figure a.1 and Figure a.2. For the grassland model, the underlying soil type influences the nutrient flows and hence uptake and growth of the grassland. Thus, for the selected areas in England and Wales the Soilscapes data (http://www.landis.org.uk) was used to define the soil types. In Scotland an expert at SAC was consulted. (Dominant and associated soils within the same area are distinguished and are described in Appendix 2, Table a.1. The soil types are linked to the pre-defined soil physical properties.

It has been assumed that the livestock systems are conventional and the fertiliser for each of the livestock systems was based on expert opinion and is described in Appendix 2 Table a.2. Cutting managements are used to mimic grazing systems. First cutting in a year (equivalent to turnout) occurs when both simulated canopy height and aboveground biomass are greater than pre-defined criteria for various enterprises (Appendix 2 Table a.3). Monthly cutting is made afterward until standing biomass on the scheduled date is less than a pre-defined amount.

7.4 Results Grassland model The model is run under three time-slices plus the baseline climate for 50 years for each of the selected sites. The output of the model focuses on grazing turnout, the period of grazing season and biomass available for grazing. Simulation results show that average

  27

biomass available and start day of grazing in a year for each livestock category vary with the climate change scenario. Average start dates for grazing season for the future time-slices are getting earlier compared with those in baseline (Table 7.1, and Appendix 2. Table a.4 and Table a.5). The impact of the start dates vary with different time-slices and livestock systems. For dairy system the most affected areas are WA and EE in 2020 but the areas shift to NW and WA in 2080. The most affected areas in 2020 are SC, NW and SW while the areas are limited to SC and NW areas in 2080. The trend in SW area in beef system (Appendix 2, Table a.4) where the date of turnout is delayed compared with the baseline, is different from the other areas. In general, the change in climate would produce more biomass from grassland in a year, although the response of grass growth to the changes is different (Table 7.2, Appendix 2. Table a.6 and Table a.7). For the sheep system, annual biomass removal will increase with the time-slices. The largest absolute increase in biomass removal is in the SW area. There is the same trend for the beef system. Some areas (EE, SE and SW) for dairy systems will reach the maximum output in 2050 and then decline; and the rest areas will increase biomass removal until 2080.

  28

  29

Table 7.1 Average start day (since 1st January) of grazing season for dairy system Site Name Advanced days compared

with baseline

Site Baseline 2020 2050 2080 2020 2050 2080

Eastern England EE 95

(7.75)* 86

(9.08) 80

(6.96) 73

(5.49) 9 15 22

East Midland EM 101 (8.92)

98 (8.75)

90 (7.08)

82 (5.92) 3 11 19

North E NE 106 (7.70)

98 (7.13)

91 (6.81)

86 (6.13) 8 15 20

North W NW 113 ( 7.05)

105 (8.86)

99 (8.74)

89 (5.76) 8 14 24

Scotland SC 108 (6.96)

102 (6.49)

96 (6.11)

87 (5.92) 6 12 21

South E SE 88 (9.01)

81 (8.99)

77 (9.04)

71 (8.44) 7 11 17

South W SW 78 (7.30)

74 (6.88)

70 (5.60)

69 (9.60) 4 8 9

Wales WA 103 (8.53)

93 (9.34)

87 (5.73)

79 (6.47) 10 16 24

West Midland WM 96 (8.24)

89 (6.54)

83 (6.54)

77 (5.87) 7 13 19

Yorkshire YH 98 (6.44)

92 (7.49)

86 (6.10)

80 (5.84) 6 12 18

* numbers in parentheses are standard deviations.

Table 7.2 Average annual biomass silage (t/ha) for dairy system Increased biomass

compared to baseline Site Baseline 2020 2050 2080 2020 2050 2080

EE 11.78 (0.79)

12.85 (1.01)

13.51 (1.24)

12.93 (2.08) 1.07 1.73 1.15

EM 9.66 (0.52) 10.42 (0.73)

11.09 (0.69)

12.12 (1.15) 0.76 1.42 2.45

NE 9.73 (0.53) 10.74 (0.69)

11.42 (0.98)

11.50 (1.15) 1.01 1.68 1.77

NW 9.02 (0.51) 9.75 (0.56)

10.24 (0.60)

10.68 (0.80) 0.72 1.22 1.66

SC 8.95 (0.43) 9.68 (0.54)

10.05 (0.52)

10.34 (0.81) 0.72 1.10 1.39

SE 12.59 (0.87)

13.69 (1.30)

14.62 (1.37)

13.71 (2.65) 1.11 2.04 1.13

SW 12.75 (0.72)

13.77 (0.90)

14.54 (1.25)

13.89 (2.39) 1.02 1.79 1.14

WA 9.14 (0.62) 10.40 (0.65)

11.19 (0.82)

12.56 (1.01) 1.26 2.05 3.42

WM 10.60 (0.45)

11.26 (0.67)

12.12 (0.99)

12.80 (1.59) 0.66 1.52 2.20

YH 10.41 (0.51)

11.17 (0.86)

11.79 (0.92)

12.69 (1.25) 0.76 1.37 2.27

* numbers in parentheses are standard deviations.

  30

  31

Impacts on nutrition General The predictions of grass growth for cattle systems allowed, in all cases, application of the rules that, first, minimum acreage under grass was calculated from grass requirements and availability during the grazing season and that, subsequently, the final acreage was calculated on the basis of additional silage requirements. For sheep farms, however, the predicted grass production was very variable during the grazing season and application of the first rule mentioned above resulted in very large areas required under grass and excessive silage production levels. An ad hoc rule was therefore formulated that, first, calculated the amount of forage required from the total yearly ME requirements of the flock and the average ME content of the grass. This gave a first prediction of the acreage required and the resulting ME availability. Because silage ME was assumed to be slightly lower than the ME content of the grass it was produced from, this always resulted in under prediction of the acreage. The additional acreage required was then estimated from the additional ME required and the total ME availability was again calculated, this time taking account of differences in ME content of different cuts. In all cases, the resulting predicted ME availability was within 1% of ME requirements and, therefore, no further iterations were considered necessary for sheep farms for adequate predictions of required acreage and associated characteristics. Dairy The usable forage production per ha increases in all areas (with the exception of three areas in the south of the UK). This is likely related to the increase in light intensity and temperature predicted by the climate change models, although this might be partially mitigated by a decrease in rainfall. The predicted decrease in yearly rainfall in dairy areas is not large (see Appendix 2 Figure a.1). There seems to be, however, a gradual shift towards wetter winters and dryer growing seasons (see Appendix 2 Figure a.2 for an example). The latter change is probably the cause of the decrease in grass growth in the south of England during the last year of the predictions (Figure 7.1 ). The increased forage production was at least partly the result of the extended growing/grazing season, which was caused almost entirely by the earlier predicted week at first cut/turnout (Table 7.1). The increase in forage production per ha is directly related to the increase in stocking rate with time in most areas (again with the exception of three RDPs in the south of England, Figure 7.2). As a result of the extended grazing season, grass DM as a proportion of total forage DM intake by cows tended to increase (Figure 7.3). An increase in stocking rate in itself results in an increase in the difference between the N-input with concentrate and the N-removal with milk. This is, however, slightly mitigated by the generally small reduction in concentrate requirements per cow as a result of the increased grass availability and further affected by the predicted changes in average CP content of the grass, which differs per region. These three factors resulted in either a decrease (especially in the north) or a small increase or very little systematic change in N-excess per ha with time (Figure 7.4).

Figure 7.1 Predicted usable dry matter yield per ha of dairy grass land in 10 regions at four points in time

Figure 7.2 Predicted changes in stocking rate on dairy farms in 10 regions in four points in time.

  32

Figure 7.3 Predicted grass DM consumption as a proportion of total forage DM intake on dairy farms in ten region at four points in time

Figure 7.4 Predicted N-excess (i.e. difference between N imported onto the farm in the form of concentrate minus the N exported with milk) per ha for ten regions at four points in time

  33

Beef Figure 7.5 shows that the predicted usable forage production per ha increases on beef farms in all areas with the exception of one area (EE). This is likely related to the increase in light intensity and temperature predicted by the climate change models, although this might be partially mitigated by a predicted decrease in rainfall during the growing season. The latter phenomenon is probably the cause of the decrease in grass growth in the East of England during the last year of the predictions (see Appendix 2). The increased forage production was at least partly the result of the extended growing/grazing season, which was caused almost entirely by the earlier predicted week at first cut/turnout; Appendix 2 Table a.1). The reason for the strongly deviating pattern for the South West of England, for which region a gradual increase in day of first cut was predicted, is unclear. The increase in forage production per ha is directly related with the increase in stocking rate with time in most areas (again with the exception of the East of England; Figure 7.6). As a result of the extended grazing season, grass DM as a proportion of total forage DM intake tended to increase in most regions (Figure 7.7). A notable exception is Scotland, where the proportion decreased. This is a direct result of a predicted shortening of the grass growing period in Scotland between 2008 and 2080; this is mainly the result of the early occurrence of the last cut in Scotland while there is very little systematic change in the last cutting week for most other regions (see Figure 7.8).

Figure 7.5 Predicted usable dry matter yield (DMY) per ha of beef grass land in 10 regions at four points in time

  34

Figure 7.6 Predicted changes in stocking rate on beef farms in 10 regions at four points in time.

Figure 7.7 Predicted grass DM as proportion of total forage DM consumed on beef farms in ten region at four points in time

  35

Figure 7.8 Predicted week in the year of the last cut of grass on beef farms in 10 regions at four points in time Sheep Figure 7.9 shows that the predicted usable forage production per ha increases markedly on sheep farms in all areas. The increased forage production is at least partly the result of the extended growing/grazing season, which is caused almost entirely by the earlier predicted week at first cut/turnout; Appendix 2 Table a.3). The increase in forage production per ha is directly related with the increase in stocking rate with time in all areas (Figure 7.10). As a result of the extended grazing season, grass DM as a proportion of total forage DM intake tends to increase, especially in northern regions (Figure 7.11).

  36

Figure 7.9 Predicted usable dry matter yield (DMY) per ha of sheep grass land in 10 regions at four points in time

  37

Figure 7.10 Predicted changes in stocking rate on sheep farms in 10 regions at four points in time.

Figure 7.11 Predicted grass DM consumption as a proportion of total forage DM intake on beef farms in ten region at four points in time

  38

Discussion 7.5 As a result of the predicted climate change, notably temperature and light intensity, the length of the growing season and forage DM production per hectare is expected to increase with time on most farm types in most regions. This increase in grass DM production per hectare allows an increase in stocking rate and an increase in the proportion of forage DM intake that consists of grass. In almost all cases, the original values observed in more northerly regions are predicted to increase in the future to values that are currently observed in more southerly regions. In the southern-most regions of England, by 2080 the forage DM production per hectare and stocking density is predicted to be similar to those predicted for 2020. This is most likely the result of a decrease in rainfall during the growing season, even when annual rainfall changes very little. In all other areas, the growing season, forage production per hectare and stocking rate are predicted to continue to increase throughout the simulated time span. As a result, the proportion of conserved forage and concentrate in the dairy cow diet and the N excess per hectare tends to decrease. Predicted changes in length of the growing season, forage production per hectare and stocking rate on beef and sheep farms follow very similar patterns to those predicted for dairy farms as a result of similar effects of climate change. For sheep farms the models predict that, primarily as a result of the extension of the grazing season, the reliance on forage conservation continues to decrease, especially in more northern areas. Caveats To show effects of climate change, as opposed to adaptation to climate change, the models used to predict grass growth and farm outcomes were based on constant other characteristics. For the interpretation of the results presented above it is, however, important to keep in mind that other changes are expected to occur in the considered time span that will affect the magnitude or even the direction of the changes that were predicted. Below we list, very briefly, a number of factors that need be considered before valid conclusions about likely changes can be drawn.

1. In the grass growth modelling a fixed yearly application of N-fertiliser was assumed. In reality, an increasing length of the growing season and increased grass production could lead to increased fertiliser application by farmers, which could lead to yet longer grazing seasons, higher DMY and, probably, D and CP contents of grass than predicted by the current models. The effect of an increase in fertiliser application is, therefore, expected to exaggerate the changes in grass yield, stocking rate, grass as a proportion of forage consumed and N excess per ha in the direction as predicted by our current models. In contrast other farmers may reduce their application of fertiliser, which may result in no or very little change in the farming system. In addition, other forages (notably forage maize) is expected to play an increasing role at the expense of grass in farming systems where at present these are not favoured because of climatological conditions (especially temperature limitations).

2. The dairy farm model is based on constant milk yields per cow per lactation. It is, however, very likely that milk yields will increase further in the future in response to genetic selection in breeding programmes. Then the amounts of forage and (especially) concentrates required per cow per year can be expected to increase. The effect of an increase in average yield per cow is,

  39

therefore, expected to reduce (or even reverse) the changes predicted in this report in terms of stocking rate and N excess per hectare.

3. Similarly, dam and birth weights and offspring growth rates were kept constant between time points in our models. In reality, further increases in at least growth rates of offspring can be expected in response to genetic selection in breeding programmes. This will lead to an increase in forage requirements and, therefore, reduce, or even reverse, the changes predicted by our models.

4. Without much more detailed modelling of the effects of such expected changes in farming characteristics, in combination with effects of climate change proper, the consequences for specific farming systems in specific areas are not easy to assess.

7.6 Conclusions The effects of climate change on farms that rely on grass production will, in almost all systems and regions, lead to a potentially longer grazing period. The predicted increase in annual grazing period between the baseline and 2080 differs between regions and varies from 3.2 weeks on dairy farms (range: 1 to 5 weeks), to 1.8 weeks on beef farms (range:-2 to +5 weeks) and 2.9 weeks on sheep farms (range: 1 to 7 weeks). In most regions this will allow animals being kept outdoors for longer and means a potential reduction in the proportion of the year that animals require housing and/or access to conserved forages. Assuming an increase in fertiliser use, there is likely to be further increases in the length of the grazing season. Proportionally, the changes in length of the grazing season are generally more substantial in northerly than in southerly regions. Climate change is also predicted to lead to increased grass growth per hectare. The predicted increase in annual usable grass DMY per hectare between the baseline and 2080 differs between regions and varies from +1.2 tonnes on dairy farms (range: 0.4 to 2.4 tonnes), to 1.5 tonnes on beef farms (range:1.0 to 2.3 tonnes) and 2.3 tonnes on sheep farms (range: 1.7 to 2.9 tonnes). Proportionately, the increase is predicted to be strongest on sheep farms, and the increase tends to be higher in northerly compared to southerly regions on all farm types. The trend of increasing DMY/ha is expected to be strengthened by a likely increase in fertiliser-N application with more favourable climatic growing conditions. This would allow for an even more pronounced increase in stocking rate, i.e. a larger herd/flock size for a given farm acreage or a reduction in farm acreage for a given herd/flock size. For dairy farms, the effects of climate change on predicted changes in stocking rate, concentrate use and N excess per hectare is likely to be limited, or reversed, as a result of increases in feed intake per cow, especially in the form of concentrates, that will likely be associated with expected increases in milk yield per animal.

  40

8 Modelling the effect of climate change on environmental pollution losses from grassland based systems (beef, sheep and dairy)

8.1

8.2

Introduction The predicted increases in grass production and the lengthening of the grazing season will lead to changes in the farming system, its economic performance and the emissions of environmental pollutants including nitrogen (N) losses, methane (CH4). However, there are competing factors regulating this process. Firstly, there are different processes competing for the available N in the soil, and secondly, the changes in climate will influence the soil water conditions, which regulate the oxidative processes. The changes in N availability will also have an influence on plant biodiversity. The objective of this section is to explore the impact of climate change on the potential environmental pollution from ruminant production for the 10 RDP regions in the UK. The SIMSDairy model, which was used to explore the impact of climate change on the dairy sector also has the capability to make an assessment of the knock-on effects for biodiversity and the economic performance of the farming systems, and hence these were explored. The predictions of grass production for this exercise were taken from the grassland model described in the previous section.

Methodology

Outputs from the SAC models of grass growth and animal performance, described in the previous section, were used as inputs into NWRes/IGER models (NGAUGE and SIMSDAIRY). . It should be noted that both NGAUGE and SIMSDAIRY models have the capacity to simulate variables that SAC models simulate. However, it was decided to use SAC models as they have been developed to simulate that part of the soil-plant-animal system with a greater specificity. The main links of information between models are shown in Figure 8.1 (general modeling framework).

  41

Figure 8.1 General modeling framework with the main input-output data linking SAC and NWRes/IGER models. (Where MH stands for medium high emission profile scenarios from UKCIP02).

Description of current versions for IGER models. 8.3 NGAUGE: NGAUGE (Brown et al., 2005) is an empirical system-based model. It simulates monthly N flows within and between the main components of grazed or cut grassland system (with/without clover), according to user inputs describing site conditions and farm management characteristics (e.g. monthly fertiliser and manure application). NGAUGE is an improvement on existing N fertiliser recommendation systems, in that it relates production to environmental impact and is therefore potentially valuable to policy makers and researchers for identifying pollution mitigation strategies and blueprints for novel, more sustainable systems of livestock production. Main losses and flows simulated are as follows: Losses of N per ha in different forms: N2O (both from nitrification and nitrification), NH3, NOx, N2, NO3 leaching. Concentrations of N in the water-associated losses (average and peak NO3-N in the leachate). Grass offtake/uptake and total N and DM per ha (hence %N content in herbage). NGAUGE has successfully been applied to different desk-top studies, which are listed in Appendix 3.

  42

SIMSDAIRY Sustainable and Integrated Management Systems for Dairy Production (SIMSDAIRY) (del Prado et al., 2006b, Defra project IS0214: final report del Prado et al., 2009) is a modelling framework at the farm level that integrates existing models for N and P, equations to simulate NH3 losses from manure application, predict CH4 losses and cows’ nutrient requirements, ‘score matrices’ for measuring attributes of biodiversity, landscape, product quality, soil quality and animal welfare and an economic model. The model follows the main principles of mass balance conservation. SIMSDAIRY is capable of simulating farm N flows (and CH4 outputs) for a given combination of management strategies, soil types and new technologies (e.g. new plant and animal traits) in UK agroclimatic areas. Appendix 3 also describes the current version of the NGAUGE and SIMSDAIRY models and the main processes that are affected by climate. Description of IGER models used and climate-main interactions with the model components Three types of changes had to be carried out within the existing versions of NGAUGE and SIMSDAIRY to enable suitable interfacing and time-step compatibilities. The specific changes can be summarised as:

• Changes to link the existing versions of NGAUGE and SIMSDAIRY with the data simulated from SAC (e.g. the effect of climate change on grass growth at the field scale and effect of climate change production on animal performance); and changes to relate the effect of climate change on certain biogeochemical processes within both models.

• Development and linking of new submodels (e.g. soil water balance submodel) to enable future climatic conditions to be simulated.

A more detailed description of these can be found in Appendix 3.

8.4 Description of case study: farm typologies, scenarios. We defined baseline typical dairy, beef and sheep farms in the UK. A selection of different UK sites was chosen to represent a wide range of soil (Appendix 3 Table a.8) and climatic conditions currently and in the future.

8.4.1 Dairy This baseline farm had typical application rates for each grassland area and a different proportion of total area used for grazed grass, cut grass (different number of cuts was also taken into account) and production of maize silage. The main characteristics of the farm are shown in Table 8.1, which describes key components of a conventional dairy farm that typically relies on on-farm grass and maize production and bought-in concentrates for sustaining animals. Dairy cows graze for about 180 days a year (from April to September) and are housed during the rest of the year. Some assumptions were made about the dairy system in order to simplify the simulations. For example, any cattle other than lactating dairy cows was simulated with the assumption that it would be represented by an average follower of a bodyweight size of 300 kg with no body-weight change during the year. The grassland area was split into cut-only fields (three cuts for silage) and grazed fields (with

  43

one small cut generally). The timing and percentage of annual mineral fertiliser applied per month was designed to follow the UK fertiliser recommendations for agricultural crops (RB209), (MAFF, 2000) and timing for the manure applied to land followed the distribution patterns described by Smith et al. (2001), in which the proportion of manure applied of the annual total is as follows: from February-April (40%), May-July (10%), August-October (25%) and November-January (25%). Seasonal milk production had substantial effects not only on the economics of the farm but also on the seasonal requirements for the herd. Thereby, it affected the needs of feed supply from varied sources (e.g. grazed grass vs silage). For this study we used an all-year round pattern. Table 8.1 Main characteristics of the typical dairy farm used as baseline Farm management Milk yield (litres/ cow yr) 7600 Fat in milk (g/kg) 40 Protein in milk (g/kg) 34 Dairy cows (number) 200 Replacement rate (%) 31 Followers (number) 190 Calving pattern All-year Breed Holstein

Silage management Average quality

Housing time-Dairy cows (days/year) 180 Housing time-Followers (days/year) 150 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) 230 270 230 40 Manure management Type of manure slurry (60 g /kg DM*)

cut-grass grazed-grass Maize

% of total applied to land 35% 35% 30%

Storage slurry tank: open

Application technique Broadcast Grassland management

cut-grass grazed-grass

young grazed-grass

History Long term grassland Sward age (years) 2-3 11-20 >20

  44

*DM: dry matter

8.4.2 Beef We defined a baseline beef farming system (lowland) following information from Defra project NT2511 (Table 8.2). Annual fertiliser N application was 175kgN/ha yr. The first application is applied at the beginning of April, and then once every six weeks. If an application date is close to or on a cutting date, fertiliser is applied the day after cutting. The total number of applications in a year is four. The diet profile is adjusted to the requirements of the herd in terms of energy density, fibre (ADF), starch and fat content. Table 8.2 Definition of variables used for beef baselines (farm level) Grazing days 300

Fertiliser N to grass (kg N/ha) 175

Cattle numbers

Number of beef 250

8.4.3 Sheep We defined a baseline lowland sheep farming system in England following information from Defra project NT2511 (Table 8.3). Annual fertiliser N application was 125 kg N/ha yr. The first application is by the end of March or early April, and then once every six weeks. The total number of applications in a year is three. Table 8.3 Definition of variables used for sheep baselines (farm level) Grazing days 300

Fertiliser N to grazed grass (kg N/ha) 125

Sheep numbers Number of sheep 600 The diet profile was adjusted to the requirements of the herd in terms of energy density, fibre (ADF), starch and fat content. Brief description of the model runs We ran NGAUGE and SIMSDAIRY for the combination of 10 RDP areas (on the dominant soil type for the region) x three livestock systems (dairy, beef and sheep) x four time slices x 50 years. The weather data generated by Earwig, which was derived from the UKCIP02 medium-high scenario and described previously, was used for the 4 time slices. Averages from 50 years of losses of N and C were recorded and analysed. Results were

  45

set to adjust the hectares needed for forage instead of adjusting the number of cows or level of milk/meat production that certain changes in plant production per hectare could trigger. In total over 6,000 model runs were completed.

Results and discussion: SAC, NWRes/IGER. 8.5

8.5.1 Dairy GHG Greenhouse gas emissions greatly varied depending on the specific gas, site and time-slice. For the baseline year, overall N2O emissions varied between 6.6 kg N2O-N/ha yr and 22.6 N2O-N/ha yr for the SE and SW regions, respectively (Table 8.4). Most future time-slices scenarios resulted in decreasing N2O emissions, expressed per ha and per unit of milk. The greatest reductions would occur in the WM region (40, 50 and 80% as N2O emissions per L of milk for 2020s, 2050s and 2080s, respectively). NW, SW, EE, SE and EM regions would also reduce substantially their emissions of N2O (up to 72, 68, 66, 62 and 60% reduction, respectively). These decreases were generally incremental in time. The most modest decrease took place in the YH region where a maximum reduction of N2O per L of milk was achieved in 2050s (25%). For the baseline year, overall CH4 emissions varied between 193 to 282 kg CH4/ha yr for the WA and SW regions, respectively (Table 8.4). Overall CH4 emissions per L of milk were similar at all sites and time-slices ca. 27 g CH4/L of milk. All future scenarios resulted in both an increase in CH4 emissions/ha and no change in CH4/L of milk. These values were completely related to the amount of forage area required for each scenario. For the baseline year, overall GHG emissions (CH4+N2O) expressed as GWP C equivalents varied between 7819 to 14056 kg C-eq/ha for the SC and SW regions, respectively (Table 8.4). Most future scenarios resulted in a decrease in C-eq emissions expressed per litre of milk and per hectare. C-eq emissions per litre of milk were always reduced with time. For example, C-eq emissions from the SW, WM and NW regions were reduced by up to 34, 32 and 31%, respectively. Reductions of C-eq emissions per litre of milk were smallest at WY and SC (up to 6% and 15% for 2080s, respectively). Acidifying gases Ammonia and NOx emissions greatly varied depending on the specific gas, site and time-slice. For the baseline year, overall NH3 emissions varied between 37.8 kg NH3-N/ha yr (WA) and 64.1 NH3-N/ha yr (EE) (Table 8.4). Simulated climate change scenarios resulted in varied trends in NH3 emissions. In most regions, NH3 emissions increased both expressed per ha and per L of milk. The largest increase of NH3 emissions per L of milk, for example, was found at NW and SC with increases up to 18 and 17%, respectively. A decrease in NH3 emissions per L of milk was only found in the SW and this only happened in the 2080s (up to 6% reduction). Very small changes in time were predicted in the SE and EE (NH3 emissions per L of milk). For the baseline year, overall NOx emissions varied between 1.2 to 2.7 kg NOx-N/ha yr for the WA and SW, respectively (Table 8.4). Future scenarios resulted in varied trends in NOx emissions (Table 8.4). Reductions in NOx/L of milk were found in the WM, YH, EE and SE regions values in the 2080s were reduced by up to 31%, 31%, 30% and 4% for the

  46

aforementioned areas, respectively. In the WA and SW regions, emissions of NOx/L of milk were reduced only in the 2080s. The rest of the regions showed increasing NOx emissions. Largest increases in the 2080s were in the NW and SC regions (up to 11% for both areas). Eutrophication losses Nitrate leaching losses varied between sites and time-slices. For the baseline year, average concentration in the leachate varied between 3.3 (SW) and 49 (EE) mg NO3-N/l. Figure 8.2 shows the clear relationship between annual drainage volume and average NO3-N concentration in the leachate. These values can in fact be fitted to a power equation with a high and positive correlation (r > 0.9 and S.E = 1.1, 95 % confidence: data not shown). The larger the annual drainage volume the smaller the average NO3-N concentration in the leachate.

  47

  48

Figure 8.2 Relationship between annual drainage volume (mm) and average NO3-N concentration in the leachate for the different RDP areas within each time-slice.

  49

Table 8.5 shows projections of annual drainage volumes for each RDP areas. For the baseline year, volumes range from 175.2 mm (EE) to 1050.8 mm (SW). Sites EE and SE showed much larger average N concentration in the leachate than the rest of the sites, which, in fact, is a reflection of their much smaller drainage volume values (<200mm) compared with the rest of the sites. Average N concentration in the leachate consistently increased with time. Large increases occurred at SW (up to 303% at 2080s), EM (up to 157% at 2080s) and WM (up to 135% at 2080s). The smallest increases were found at WA site (up to 10% at 2080s). These increases were consistent with the decrease in annual drainage volume in the future scenarios. Nitrate leaching results for dairy systems expressed as kg NO3-N/ha yr are shown in Table 8.6. For the baseline year, results range from 35.3 (SW) and 76.8 kg NO3-N/ha (EE). Nitrate leaching increased with time in almost all sites except for WA site, where a decrease of up to 50% was found in 2080s. The largest increase in NO3 leaching was at SW site (>150% in both 2050s and 2080s). The smallest changes in NO3 leaching were found at YH and EE areas with 0% and 6% increase in 2080s, respectively. Potential trade-offs on other pillars of farm sustainability The environmental indicators can be traded- off against other observable indicators of dairy farm performance and sustainability. These are shown in the following tables. The following indicators are shown: milk yield/ha (Table 8.7), land area (grass + maize) required in dairy systems (Table 8.8), net farm income (Table 8.9), biodiversity (Table 8.10) and soil quality (Table 8.11). According to results shown in Table 8.9, where net margin is shown for the different sites and time-slices´ scenarios, for the baseline year, values ranged between 48506 (SC) to 52709 £ (SW). The largest incomes were found in SE and SW sites, which coincided with the sites with greatest milk production per hectare (Table 8.8). Future scenarios brought about an increase in total net farm income possibly caused by the increase productivity of the land. For the baseline year, the biodiversity index varied between 0 (WA) to 1 (SE), which reflects the low level of biodiversity scope for intensive dairy farming systems. Future scenarios showed differences in trends between sites. Most areas resulted in an increase in biodiversity level in 2080s, however, some decreased the level up to 2050s (e.g. SW: up to 22% decrease) and some other increase the level of biodiversity for all time-slices (e.g. WM and YH). For the baseline year, the soil index varied between 1 (SW) to 2 (EE). Soil quality values were in relation to soil compaction, erosion and poaching. Most differences were related to the climatic changes and the rainfall patterns. All areas resulted in an increase in soil quality level in 2080s. The largest increase was at SW area (133%) and the smallest was at SE area (10%).

50

Table 8.4 Results for the different environmental pollutants in dairy farms GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha g/L milk

kg C/ha

g/L milk

kg C-eq/ha

g /L milk

kg N/ha g/L milk

kg N/ha

g/L milk

mg/L (mean concentration in leachate)

RDP AREA-EE*

current 7.2 0.7 270 27 8965 897 64.1 6.4 2.7 0.3 49 2020 -28% -32% 6% 0% -2% -8% 8% 1% -16% -21% 27% 2050 -32% -38% 9% 0% -2% -10% 9% 0% -20% -26% 16% 2080 -63% -66% 9% 0% -9% -17% 10% 0% -24% -30% 32%

RDP AREA-EM current 12.5 1.6 206 27 9104 1191 42.8 5.6 1.5 0.2 11 2020 -8% -16% 9% 0% 2% -7% 13% 4% 50% 37% 79% 2050 -26% -35% 14% 0% -3% -15% 20% 5% 32% 16% 109% 2080 -48% -60% 29% 0% -5% -26% 34% 4% 30% 1% 157%

RDP AREA-NE current 9.2 1.1 217 27 8343 1037 49.0 6.1 1.8 0.2 18 2020 0% -9% 9% 0% 6% -3% 14% 5% 51% 38% 82% 2050 -30% -42% 21% 0% 3% -15% 29% 6% 49% 23% 117% 2080 -41% -53% 24% 0% 1% -19% 38% 11% 26% 2% 99%

RDP AREA-NW current 12.2 1.6 205 27 8989 1183 42.9 5.6 1.5 0.2 16 2020 -10% -16% 8% 0% 0% -7% 14% 6% 20% 11% 2% 2050 -7% -19% 15% 0% 6% -8% 30% 12% 78% 54% 120% 2080 -65% -72% 25% 0% -14% -31% 47% 18% 38% 11% 87%

RDP AREA-SC current 8.8 1.2 201 27 7819 1052 43.2 5.8 1.5 0.2 12 2020 1% -4% 6% 0% 4% -2% 11% 5% 22% 15% 3% 2050 3% -8% 12% 0% 9% -3% 27% 13% 77% 57% 52% 2080 -28% -42% 24% 0% 5% -15% 45% 17% 11% 37% 74%

*See Table 7.1 for full name of RDP region

Table 8.4 (cont.). Results for the different environmental pollutants in dairy farms. GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha g/L milk

kg C/ha

g/L milk

kg C-eq/ha

g /L milk

kg N/ha g/L milk

kg N/ha

g/L milk

mg/L (mean concentration in leachate)

RDP AREA-SE current 6.6 0.6 277 27 8962 874 64.0 6.2 2.1 0.2 43 2020 -8% -15% 8% 0% 4% -4% 10% 2% 7% -1% 14% 2050 -19% -33% 20% 0% 11% -8% 19% -1% -1% -18% 20% 2080 -57% -62% 11% 0% -5% -14% 13% 2% 7% -4% 41%

RDP AREA-SW current 22.6 2.2 282 27 14056 1346 63.8 6.1 1.9 0.2 3.3 2020 -37% -36% -2% 0% -19% -18% 2% 5% 39% 42% 163% 2050 -57% -64% 18% 0% -20% -32% 19% 0% 59% 34% 245% 2080 -64% -68% 14% 0% -25% -34% 7% -6% -8% -19% 303%

RDP AREA-WA current 15.4 2.1 194 27 9692 1349 37.8 5.3 1.2 0.2 7.8 2020 11% -3% 14% 0% 12% -2% 19% 5% 60% 41% 28% 2050 -5% -24% 25% 0% 10% -13% 31% 5% 31% 5% 7% 2080 -6% -39% 54% 0% 23% -20% 62% 5% 12% -28% 10%

RDP AREA-WM current 12.2 1.4 237 27 9759 1109 55.8 6.3 2.6 0.3 22 2020 -33% -40% 11% 0% -6% -16% 14% 3% -17% -26% 51% 2050 -42% -50% 17% 0% -7% -20% 22% 4% -17% -29% 110% 2080 -76% -80% 22% 0% -16% -32% 29% 6% -16% -31% 135%

RDP AREA-YH current 7.6 0.9 230 27 8161 957 54.4 6.4 2.5 0.3 14 2020 12% -1% 13% 0% 12% -1% 15% 2% 5% -7% 9% 2050 -13% -25% 16% 0% 7% -8% 19% 3% -16% -28% 49% 2080 2% -17% 23% 0% 16% -6% 25% 2% -15% -31% 51%

  51

  52

Table 8.5 Projections of annual drainage volume (mm) for dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Mm current 175.2 570.1 353.4 414.0 529.4 193.2 1050.8 989.0 325.7 520.3

2020 -20% -18% -22% -11% -10% -14% -9% -13% -36% -8% 2050 -18% -33% -36% -40% -25% -3% -24% -24% -49% -26% 2080 -13% -46% -42% -39% -37% -16% -29% -34% -46% -31%

Table 8.6 Projections of NO3 leaching (kg NO3-N/ha) for dairy systems EE EM NE NW SC SE SW WA WM YH Scenario kg NO3-N/ha yr current 76.8 59.5 60.8 65.0 63.7 73.2 35.3 76.2 66.0 71.1

2020 0% 48% 19% -10% -7% 6% 135% 11% 1% -2% 2050 1% 44% 28% 27% 15% -10% 155% -19% 8% 1% 2080 6% 32% 20% 11% 1% 15% 153% -50% 18% 0%

Table 8.7 Projections of milk production per hectare in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario L milk/ha yr current 9996 7643 8045 7600 7431 10250 10441 7182 8797 8524 2020 6% 9% 9% 8% 6% 8% -2% 14% 11% 13% 2050 9% 14% 21% 15% 12% 20% 18% 25% 17% 16% 2080 9% 29% 24% 25% 24% 11% 14% 54% 22% 23% Table 8.8 Projections of land area (grass + maize) required in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Ha current 147 192 182 193 198 143 141 204 167 172 2020 -6% -8% -8% -7% -6% -8% 2% -12% -10% -12% 2050 -8% -13% -18% -13% -11% -17% -15% -20% -14% -13% 2080 -9% -22% -19% -20% -19% -10% -12% -35% -18% -18% Table 8.9 Projections of net farm income in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Net farm income (£) current 51139 48875 49376 49147 48506 52111 52709 48443 49733 491632020 3% 1% 2% 2% 1% 2% -1% 1% 5% 3% 2050 3% 4% 3% 2% 1% 1% 2% 5% 5% 5% 2080 4% 7% -5% 6% 6% 2% 2% 8% 6% 7% Table 8.10 Projections of biodiversity level in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Biodiversity index current 1 1 1 0 1 1 1 0 1 1 2020 4% -5% -5% 4% 2% 0% -7% -4% 11% 7% 2050 5% -1% 3% -1% -1% 2% -22% 10% 15% 15% 2080 7% 6% 9% 15% 13% 1% 2% 16% 18% 14% Biodiversity index: where the scores assigned reflect poor (0) to very satisfactory (6) sustainability.

Table 8.11 Projections of soil quality level in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Soil quality index current 2 1 2 2 2 2 1 1 1 1 2020 7% -9% 3% 14% 7% 5% 31% -29% 34% 13% 2050 10% 38% 15% 17% 1% 8% 81% 1% 43% 33% 2080 13% 49% 30% 34% 34% 10% 133% 51% 40% 38% Soil quality index: where the scores assigned reflect poor (0) to very satisfactory (6) sustainability.

8.5.2 Beef GHG Greenhouse gas emissions greatly varied depending on the specific gas, site and time-slice. For the baseline year, overall N2O emissions varied between 1.6 kg N2O-N/ha yr and 16.6 N2O-N/ha yr (0.5-13.9 g N/kg meat) for the SC and SW regions, respectively (Table 8.12), with greatest emissions in the SW region where current summer rainfall is greatest. Simulated N2O emissions, both expressed per unit area and unit of meat, for the 2020s, 2050s and 2080s were reduced in most regions as a result of the predicted drier conditions (compared with the baseline year) and also, indirectly due to differences in plant N use efficiency. The greatest reductions in N2O emissions per unit of meat would occur in the SW area (53, 58 and 68 % as N2O emissions per unit of meat for 2020s, 2050s and 2080s, respectively). N2O emissions would also be markedly reduced in the SW, NW, SE, EM and WM (up to 68, 55, 57, 44 and 57% reduction, respectively). These decreases were generally incremental in time. The most modest decrease in N2O emission per kg of meat (and also very constant from the 2020s) took place in YH region site where a maximum reduction was achieved in 2080s (10%). The only area where N2O emissions per unit of meat increased was in the WA region due to wetter conditions at the fertilisation application time. In the EE area N2O emissions increased until the 2080s, when emissions where substantially reduced (43%). For the baseline year, overall CH4 emissions varied between 202 to 267 kg CH4/ha yr for the SC and EE regions, respectively (Table 8.12). All future scenarios resulted in both an increase in CH4 emissions/ha and a decrease in CH4/kg meat. The largest reduction of CH4 per unit of meat product occurred in the SW where there was a reduction of 1%, 3% and 11% for the 2020s, 2050s and 2080s, respectively. Most scenarios, however, showed very similar reduction values and trend (CH4 emissions per unit of product decreased in time). For the baseline year, overall GHG emissions (CH4+N2O) expressed as GWP C equivalents varied between 4762 to 10604 kg C-eq/ha for the SC and SW regions, respectively (Table 8.12). Emissions expressed per unit of meat varied between 2879 to 4523 g C/kg meat for the SC and NW regions, respectively. Simulated climate change resulted in a decrease in C-eq emissions/meat in most areas, with the largest reductions in the SW (up to 39%) and NW (up to 28%). Exceptions, however, took place in the WA region with respect to this C-eq emissions/kg meat. WA increased C-eq emissions/kg meat emissions in 2020s and 2050s. Most CH4 emissions were caused by animal enteric fermentation. Although CH4 output from dung deposition may have varied from sites to sites and time-slices, the fact that the CH4 emissions from this source are much smaller than from enteric fermentation means that these potential differences do no effect the overall CH4 emissions at the farm level.

  53

Methane emissions per hectare were greatly affected by the hectarage required to support a given level of production; with fewer hectares resulting in more CH4 emissions per ha and per kg of meat product. Methane output per kg of product was inversely related to %N content as more protein in the diet generally increased the amount of kg of meat per unit of DM ingested by the animal. This study did not account for the potential changes in animal energy use in future time-slice scenarios. The quality of the diet was not included as a factor affecting CH4 emissions from enteric fermentation. Although there is some evidence that some factors (e.g. fat content) may have a strong effect on CH4 output from enteric fermentation, the SAC modelling of herbage production did not provide estimates of the parameters linked to enteric CH4 production. Acidifying gases Both NH3 and NOx emissions varied depending on the specific gas, site and time-slice. However, NOx emissions were very small (0.13 kg NOx-N/ha maximum) compared with NH3 emissions. For the baseline year, overall NH3 emissions varied between 6.8 kg NH3-N/ha yr (SC) and 19.6 NH3-N/ha yr (EE). In all RDP areas, NH3 emissions, both expressed per ha and per kg of meat, increased. The largest increase of NH3 emissions per kg of product was found at YH (up to 31% in 2080s) and SC (up to 30% in 2050s and 2080s). Predicted climate change led to changes in the grass composition. Increases in dietary protein generally leads to substantial increases in urinary loss (Van Soest, 1994) with almost all N ingested in excess of animal requirement excreted in urine (Peyraud et al.1995). The partitioning of excreted N between dung and urine reflects the concentration of N in the diet. For the baseline year, overall NOx emissions varied between 0 kg NH3-N/ha yr (e.g. NE) and 0.1 NH3-N/ha yr (e.g. EE). NOx emissions generally increased both expressed per ha and per kg of meat. In some cases, the increase was very large (e.g. EM area: 1776% in 2080s). However, as previously mentioned, absolute values were very small. Both NH3 and NOx increases were generally incremental with time. Eutrophication losses Nitrate leaching losses varied between sites and time-slices. For the baseline year, average concentration in the leachate varied between 1.7 (SW) and 10.6 (SC) mg NO3-N/l. Largest NO3 average concentrations in the leachate were associated with both freely drained soils (e.g. SC) and smaller drainage volumes (data not shown). In most sites, average NO3 concentrations in the leachate increased with time. Large increases were found at YH (13%, 183% and 165% for 2020s, 2050s and 2080s, respectively), SW area (175%, 262 % and 146% for 2020s, 2050s and 2080s, respectively), EE area (203%, 203 % and 235% for 2020s, 2050s and 2080s, respectively, SE area (32%, 55 % and 274% for 2020s, 2050s and 2080s, respectively), NE area (-12%, 102 % and 101% for 2020s, 2050s and 2080s, respectively) and WM (15%, 36 % and 91% for 2020s, 2050s and 2080s, respectively). Increases in NO3 concentrations in leachates in the WM, SE and NE regions were especially important as absolute values were substantially greater than at the rest of the sites. Decreases in average N concentrations in the leachate were only found in the SC and WA regions. For SC, this decrease may be potentially more relevant as absolute average N concentrations in the leachate values were in the upper range of absolute values. The predictions of meat predictions per hectare and the grassland area required for the beef are shown in Table 8.13 and Table 8.14.

  54

  55

Table 8.12 Results for the different environmental pollutants in beef farms GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha g/ kg meat

kg C/ha

g/ kg meat

kg C-eq/ha

g / kg meat

kg N/ha g/ kg meat

kg N/ha

g/ kg meat

mg/L (mean concentration in leachate)

RDP AREA-EE current 5.2 2.1 266.2 109.1 7198.9 2951.5 19.6 8.0 0.1 0.0 4.1 2020 34% 20% 8% -4% 13% 2% 23% 11% 14% 3% 203% 2050 20% 7% 8% -4% 11% -1% 26% 12% 10% -2% 203% 2080 -35% -43% 10% -5% 0% -13% 30% 13% 7% -7% 235%

RDP AREA-EM current 11.7 5.6 240.2 114.3 8685.9 4134.3 11.6 5.5 0.0 0.0 2.8 2020 -6% -8% 2% -1% -1% -4% 19% 16% 764% 742% 86% 2050 -34% -42% 10% -4% -8% -20% 15% 0% 1493% 1290% -1% 2080 -31% -44% 15% -6% -4% -22% 41% 15% 2210% 1776% 52%

RDP AREA-NE current 10.7 5.7 221.8 117.6 7976.8 4230.7 10.5 5.5 0.0 0.0 5.0 2020 -2% 1% -2% 1% -2% 1% 9% 12% 178% 186% -12% 2050 -14% -25% 11% -4% 1% -13% 29% 11% 519% 437% 102% 2080 -18% -33% 16% -6% 2% -17% 45% 18% 562% 440% 101%

RDP AREA-NW current 13.4 6.7 230.6 115.9 8996.5 4522.7 12.1 6.1 0.1 0.0 4.1 2020 -17% -20% 3% -1% -6% -10% 13% 8% -29% -31% 2% 2050 -24% -36% 13% -5% -4% -19% 22% 2% 48% 24% -19% 2080 -46% -55% 15% -6% -13% -28% 28% 5% 58% 30% 34%

RDP AREA-SC current 1.6 1.0 202.7 122.5 4762.8 2878.8 6.8 4.1 0.1 0.0 10.6 2020 12% 6% 3% -1% 4% -1% 31% 24% 59% 51% 3% 2050 8% -4% 9% -4% 9% -4% 47% 30% 140% 112% -15% 2080 7% -16% 19% -7% 18% -8% 66% 30% 154% 98% -63%

Table 8.12 (cont.). Results for the different environmental pollutants in beef farms. GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha g/ kg meat

kg C/ha

g/ kg meat

kg C-eq/ha

g / kg meat

kg N/ha g/kg meat

kg N/ha

g/ kg meat

mg/L (mean concentration in leachate)

RDP AREA-SE current 8.6 3.6 265.1 109.7 8234.7 3408.5 14.2 5.9 0.0 0.0 4.3 2020 6% 1% 3% -1% 4% -1% 6% 1% 73% 65% 32% 2050 -6% -14% 6% -3% 2% -6% 11% 2% 85% 69% 55% 2080 -49% -57% 12% -6% -8% -22% 31% 10% 87% 57% 274%

RDP AREA-SW current 16.6 7.1 259.8 110.8 10604.7 4522.6 14.7 6.2 0.1 0.0 1.7 2020 -51% -53% 2% -1% -24% -26% 10% 7% -73% -74% 175% 2050 -54% -58% 6% -3% -23% -30% 22% 11% -15% -23% 262% 2080 -56% -68% 23% -11% -16% -39% 41% 2% 50% 9% 146%

RDP AREA-WA current 11.9 5.9 234.6 116.2 8625.5 4272.9 9.8 4.8 0.0 0.0 2.0 2020 50% 50% 0% 0% 18% 18% 4% 4% -46% -46% -49% 2050 22% 26% -3% 1% 6% 10% 17% 22% 171% 181% -57% 2080 24% 5% 13% -5% 17% -1% 31% 11% 271% 213% -73%

RDP AREA-WM current 8.9 4.3 237.8 114.3 7751.8 3727.1 11.7 5.6 0.0 0.0 4.6 2020 7% -9% 12% -4% 10% -6% 26% 7% 283% 226% 15% 2050 6% -12% 14% -5% 11% -8% 37% 14% 411% 324% 36% 2080 -45% -57% 18% -7% -4% -25% 51% 18% 417% 306% 91%

RDP AREA-YH current 10.0 4.9 234.7 116.1 8020.5 3969.9 9.9 4.9 0.1 0.0 2.8 2020 -1% -9% 5% -2% 3% -5% 15% 6% -55% -58% 13% 2050 0% -9% 6% -3% 4% -5% 29% 18% 10% 0% 183% 2080 -11% -10% -1% 0% -5% -4% 30% 31% 36% 38% 165%

  56

57

Table 8.13 Projections of meat production per hectare in beef systems. EE EM NE NW SC SE SW WA WM YH Scenario Kg meat/ha yr current 2439 2101 1885 1989 1654 2416 2345 2019 2080 2020 2020 12% 3% -3% 4% 5% 5% 3% 3% 17% 8% 2050 12% 15% 15% 19% 13% 9% 10% 0% 21% 9% 2080 15% 23% 22% 21% 28% 19% 38% 22% 27% -1% Table 8.14 Projections of grassland area required in beef systems. EE EM NE NW SC SE SW WA WM YH Scenario ha current 82 100 112 96 118 81 80 102 95 101 2020 -8% -9% -7% -2% -2% -4% -1% -10% -15% -3% 2050 -15% -17% -19% -11% -4% -6% -8% -4% -16% -12% 2080 -11% -22% -23% -13% -13% -18% -27% -22% -26% -1%

8.5.3 Sheep GHG In sheep systems, greenhouse gas emissions varied greatly depending on the specific gas, site and time-slice. For the baseline year, overall N2O emissions varied between 0.9 kg N2O-N/ha yr (EE) and 16.3 kg N2O-N/ha yr (SW) (Table 8.15), with greatest emissions in the SW region where current summer rainfall is greatest. Simulated N2O emissions, both expressed in relation to land area and unit of meat, for the 2020s, 2050s and 2080s were reduced in all regions as a result of the predicted drier conditions (compared with the baseline year). Differences in N2O were mainly caused by differences in soil water conditions at the time of fertilisation and also, indirectly due to differences in plant N use efficiency. The greatest reductions would occur in the SW area (65, 77 and 82 % as N2O emissions per unit of meat for 2020s, 2050s and 2080s, respectively). N2O emissions would also be markedly reduced in the NW and WM areas (up to 77 and 71% reduction, respectively). These decreases were generally incremental in time. The most modest decrease in N2O emission per kg of meat (and also very constant from the 2020s) took place in the EE region where a maximum reduction was achieved in 2080s (18%). For the baseline year, overall CH4 emissions varied between 124 (SC) and to 214 (SE) kg CH4/ha yr (Table 8.15). This variation is due to mainly differences in hectares required for each farm scenario. Simulated climate change resulted in both an increase in CH4 emissions/ha and a decrease in CH4/kg meat in all regions as a result of both a decrease in surface required for forage (as grass production per hectare increased in time) and less protein in the diet generally decreased the amount of kg of meat per unit of DM ingested by the animal and thereby, indirectly decreased CH4 output per unit of meat. Methane output per unit of product was inversely related to %N content as more protein in the diet generally increased the amount of kg of meat per unit of DM ingested by the animal. The largest reduction of CH4 per unit of meat product occurred in the SW region, where there was a reduction of 3%, 5% and 8% for the 2020s, 2050s and 2080s, respectively. Most scenarios, however, showed very similar reduction values and trend (CH4 emissions

58

per unit of product decreased in time). In the baseline year, overall GHG emissions (CH4+N2O) expressed as GWP C equivalents varied between 4061 to 8229 kg C-eq/ha for the EE and SW regions, respectively (Table 8.15). Emissions expressed per unit of meat varied between 2810 to 7030 g C/kg meat for the EE and SW regions, respectively. Simulated climate change resulted in a decrease in C-eq emissions/meat in most areas, with the largest reductions in the SW (up to 54%) and NW (up to 40%). Reductions of C-eq emissions were smallest in the EE (up to 6%) and SE areas (up to 12%). Acidifying gases Ammonia and NOx emissions varied depending on the specific gas, site and time-slice. For the baseline year, overall NH3 emissions varied between 7.1 kg NH3-N/ha yr (SC) and 21.6 NH3-N/ha yr (SE) (Table 8.15), with greatest emissions in the SE region where changes in grass crude protein and also changes in temperature and rainfall patterns. Simulated NH3 emissions in relation to land area and unit of meat for the 2020s, 2050s and 2080s were increased in most regions as a result of the predicted drier conditions (compared with the baseline year) and an increase in grass crude protein, resulting in greater urine: dung ratios from excreted N and thereby larger pools of TAN excreted to the soil. The largest increase of NH3 emissions per unit of product was found in the NE (13, 24 and 39 % as NH3 emissions per unit of meat for 2020s, 2050s and 2080s, respectively) and NW (6, 16 and 38 % as NH3 emissions per unit of meat for 2020s, 2050s and 2080s, respectively). The largest decrease in NH3 emission per kg of meat was found at SE (-11, -8 and -9 % as NH3 emissions per unit of meat for 2020s, 2050s and 2080s, respectively) and SW (-17, -13 and -10 % as NH3 emissions per unit of meat for 2020s, 2050s and 2080s, respectively) regions. Climate changes led to changes in the grass composition. Increases in dietary protein generally leads to substantial increases in urinary loss (Van Soest, 1994) with almost all N ingested in excess of animal requirement excreted in urine (Peyraud et al.1995). The partitioning of excreted N between dung and urine reflects the concentration of N in the diet. It predicts that, with 1.5 % N in the diet, 45 % of the excreted N occurs in the urine whereas, with 4 % N in the diet, 80 % occurs in the urine (Scholefield et al., 1991). For the baseline year, overall NOx emissions varied between 0 to 0.5 kg NOx/ha yr for WA and EE regions, respectively. Overall NOx emissions per unit of meat varied in a similar way to those values expressed per unit of ha. Future time-slices´ scenarios varied in their trends in relation to increasing/decreasing NOx emissions. In most sites, NOx emissions increased both expressed per ha and per kg of meat. Wherever there was a decrease in NOx emissions per unit of meat, a smaller decrease or even an increase for NOx emissions per ha was found (e.g. SW). The largest increase of NOx emissions per unit of product were found at NE (13, 24 and 39 % as NOx emissions per unit of meat for 2020s, 2050s and 2080s, respectively) and YH (9, 22 and 33 % as NOx emissions per unit of meat for 2020s, 2050s and 2080s, respectively). The largest decrease in NOx emissions per unit of meat was found in the SE, SW, WA and EE regions. Eutrophication losses Nitrate leaching losses varied between areas and time-slices. Average concentration in the leachate varied between 1.2 (EM) and 14.5 (SC) mg NO3-N/l. Largest NO3 leaching losses were associated with freely draining soils (e.g. those of the SC, NW regions) and smaller drainage volumes (data not shown).

59

In most regions, concentrations of average NO3 in drainage reduced with time. In some sites, NO3 leaching losses were reduced in time for all time-slices (e.g. SC and WA site). This trend was not found at all sites though. Largest reductions were generally found at sites with lowest leaching losses (e.g. EM and SE). At site YH, whereas average N concentration in the leachate increased for 2020s and 2050s, it sharply decreased for 2080s time-slice. The predictions of meat predictions per hectare and the grassland area required for the sheep systems are shown in Table 8.16 and Table 8.17.

60

Table 8.15 Results for the different environmental pollutants in sheep farms GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha g/ kg meat

Kg C/ha

g/ kg meat

kg C-eq/ha

g / kg meat

kg N/ha g/ kg meat

kg N/ha

g/ kg meat

mg/L (mean concentration in

leachate) RDP AREA-EE

current 0.9 0.6 179.7 124.4 4061.0 2810.3 13.5 9.3 0.5 0.4 7.4 2020 -8% -16% 7% -2% 6% -3% 12% 3% -24% -30% -68% 2050 -1% -15% 13% -3% 12% -4% -3% -17% -23% -34% -82% 2080 3% -18% 20% -5% 19% -6% 6% -16% -25% -41% -98%

RDP AREA-EM current 4.3 2.8 189.1 122.8 5306.7 3445.8 17.3 11.3 0.0 0.0 1.2 2020 -13% -22% 9% -2% 3% -7% 23% 10% 105% 84% -100% 2050 -11% -27% 16% -4% 9% -10% 36% 12% 139% 97% -100% 2080 -24% -43% 26% -7% 13% -16% 38% 2% 141% 78% -100%

RDP AREA-NE current 5.1 4.2 157.5 128.0 4896.6 3978.9 11.0 9.0 0.0 0.0 5.9 2020 -35% -40% 7% -2% -6% -14% 23% 13% 55% 43% -99% 2050 -41% -51% 17% -3% -2% -19% 50% 24% 99% 65% -69% 2080 -36% -53% 29% -6% 8% -21% 89% 39% 85% 35% -84%

RDP AREA-NW current 8.6 7.9 141.4 130.8 5622.9 5203.8 9.2 8.5 0.0 0.0 10.9 2020 -29% -37% 10% -2% -8% -18% 19% 6% 158% 130% -46% 2050 -58% -67% 22% -4% -16% -34% 48% 16% 171% 114% -87% 2080 -67% -77% 37% -7% -12% -40% 103% 38% 159% 76% -86%

RDP AREA-SC current 9.6 10.3 124.2 134.2 5570.8 6017.8 7.1 7.7 0.5 0.5 14.5 2020 4% -6% 10% -2% 7% -4% 15% 3% -98% -99% -6% 2050 -10% -28% 20% -3% 4% -16% 34% 8% -92% -94% -31% 2080 -50% -63% 31% -5% -12% -36% 55% 13% -88% -92% -67%

Table 8.15 (cont). Results for the different environmental pollutants in sheep farms GHG Acidifying gases Eutrophication N2O CH4 Total NH3 NOx NO3 leaching Time-slice

kg N/ha

g/ kg meat

kg C/ha

g/ kg meat

kg C-eq/ha

g / kg meat

kg N/ha g/ kg meat

kg N/ha

g/ kg meat

mg/L (mean concentration in leachate)

RDP AREA-SE current 3.4 1.9 214.0 118.6 5550.0 3076.3 21.6 12.0 0.1 0.0 1.5 2020 -4% -13% 8% -3% 6% -5% -1% -11% -2% -11% -100% 2050 -17% -30% 14% -4% 8% -9% 10% -8% -7% -21% 24% 2080 -23% -39% 18% -6% 11% -12% 15% -9% -17% -34% -100%

RDP AREA-SW current 16.3 13.9 151.5 129.5 8229.2 7030.5 9.5 8.1 0.5 0.4 3.0 2020 -60% -65% 13% -3% -32% -41% -4% -17% -96% -97% -44% 2050 -70% -77% 26% -5% -33% -49% 16% -13% -89% -92% -81% 2080 -73% -82% 40% -8% -29% -54% 38% -10% -85% -90% -43%

RDP AREA-WA current 10.5 9.6 144.0 130.6 6286.8 5704.0 9.3 8.5 0.0 0.0 4.8 2020 9% 10% 0% 0% 4% 4% -1% -1% -62% -62% -57% 2050 -27% -35% 10% -2% -5% -15% 15% 3% 73% 54% -96% 2080 -29% -44% 22% -5% 2% -20% 4% -19% 125% 76% -100%

RDP AREA-WM current 7.3 5.7 163.9 127.1 5713.4 4430.0 11.6 9.0 0.0 0.0 5.4 2020 -45% -51% 10% -2% -12% -22% 18% 5% 122% 97% -74% 2050 -46% -57% 18% -4% -7% -25% 30% 5% 187% 132% -76% 2080 -60% -71% 28% -6% -7% -32% 11% -19% 163% 93% -100%

RDP AREA-YH current 4.5 3.7 156.3 128.1 4682.2 3839.3 12.0 9.8 0.0 0.0 1.5 2020 -16% -26% 11% -2% 3% -9% 24% 9% 293% 246% 46% 2050 -26% -43% 24% -5% 9% -16% 58% 22% 418% 299% 18% 2080 -23% -47% 35% -7% 18% -19% 94% 33% 369% 223% -100%

61

62

Table 8.16 Projections of meat production per hectare in sheep systems EE EM NE NW SC SE SW WA WM YH Scenario Kg meat/ha yr Current 1445 1540 1231 1081 926 1804 1170 1102 1290 1220

2020 9% 11% 9% 12% 12% 11% 16% 17% 13% 13% 2050 17% 21% 21% 27% 25% 19% 33% 31% 23% 30% 2080 26% 35% 37% 47% 38% 26% 52% 49% 36% 45%

Table 8.17 Projections of grassland area required in sheep systems EE EM NE NW SC SE SW WA WM YH Scenario Ha Current 57 53 74 80 98 45 70 74 63 68 2020 -9% -12% -15% -11% -11% -11% -15% -14% -12% -12% 2050 -17% -18% -24% -24% -21% -18% -26% -24% -20% -23% 2080 -22% -27% -33% -35% -33% -22% -36% -34% -27% -32%

8.6 Discussion of results As expected, changing climatic conditions greatly affected N losses. A combination of two factors were responsible for controlling such losses. First, different processes compete for the available N in the soil, and second, soil water conditions greatly regulate the oxidative level of the N lost (e.g. as NOx, N2O, N2). Nitrate may: (1) undergo denitrification to gaseous oxides of N and to N2, (2) be taken up by organisms (assimilatory reduction), (3) be used by microorganisms as an electron acceptor and become reduced to NH4

+ (dissimilatory reduction), (4) be lost in leachate or run off, or (5) accumulate in the soil. Ammonium may: (1) be taken up by plants, (2) be immobilised in microbial biomass, (3) nitrify to NO3

- and partially be lost as gaseous oxides of N (4) be leached, (5) accumulate in the soil (Paul and Clark, 1996) or (6) volatilised as ammonia (NH3). The most important factor affecting those pathways is the increase in plant production per unit of both land area (ha) and kg of animal product (data not shown). The increasing plant N uptake in future climate change scenarios causes a reduction in soil inorganic N available for the rest of the competing processes leading to loss of N to the wider environment. The form of N that is lost via denitrification and nitrification is greatly regulated by soil water content. The rate of NO emission is greater with drier soils and decreases as the soil moisture content approaches field capacity. As the soil atmosphere becomes more O2-limited, so, N2O and N2 emissions increase while emissions of NO decrease. Evidence suggests that the optimum soil % WFPS values for maximum NO emissions were between 30% and 40% WFPS (e.g. del Prado et al., 2006a). Nitrous oxide emissions increase to a level where simultaneous denitrification and nitrification are at their maximum (75% WFPS). Above this soil water content, denitrification is the main process producing N2O and, as the soil became more anaerobic, emissions of N2 became greater than those of N2O. Nitrate leaching losses (only shown for dairy systems) were affected by factors that regulate the competition for the available N in the soil during the period previous to autumn-winter drainage season. Among these factors, increasing plant N uptake would result in a decrease in leaching but decreasing denitrification losses through disfavouring soil anaerobic conditions would lead to an increase in soil leachable N for the autumn-winter drainage season. Smaller drainage volumes in future scenario

63

would result in a reduced amount of total N being leached and an increase in the concentration of that N actually leached. In this study, as land area required was adjusted to the amount of food produced, assuming a constant product (e.g total meat and milk) resulted in greater NO3 leaching per hectare but fewer per unit of product (or total) (data not shown). Most CH4 emissions were caused by animal enteric fermentation. Although CH4 emissions from dung deposition may have varied from region to region and time-slices, they are minor in comparison to enteric fermentation and would not influence the overall CH4 emission at the farm level. Methane emissions per unit hectare were greatly affected by the required number of hectares required to support the level of production. So, the fewer the land surface available, then the greater the CH4 emission per ha and also per kg of meat product. Methane output per unit of product was inversely related to %N content as more protein in the diet generally increased the amount of kg of meat per unit of DM ingested by the animal. This study did not account for the potential changes in animal energy use in future time-slice scenarios. The quality of diet was not included as a factor affecting CH4 emissions from enteric fermentation. Although there is some evidence that some factors (e.g. fat content) may have a strong effect on CH4 output from enteric fermentation, herbage parameters simulated by the SAC models did not include those needed to estimate the potential effect on rumen CH4 generation. The economic results were simulated by assuming that all input values would not change with geopolitical scenario and site. Due to this simplification, we did not intend to produce a robust assessment of the impact of climate change on the economic returns; however we have used these results as an example of possible differences caused only by climatic changes in different areas of the UK. As Glendining et al. (2009) indicated for sustainability-related studies, land area should also be included in any assessment. Land area was included in our study, but an improved methodology would require costing changes in land requirement to produce the same amount of food. Economically-based policies may in turn be in serious conflicts with those more related to environmental risks due to this mismatch of scale relevance. The effectiveness of mitigation methods to decrease any particular pollutant will depend for example on the farmer or land user´s response to any potential economic benefits or penalties due to implementation and in relation to market dynamics. Benefits of increasing the efficiency of N plant use included greater opportunities for improving the botanical diversity. This fact, however, needs further testing as the temporal variation of inorganic N flows in the soil within a year may have different implications for different potential plant species. The soil quality component in the SIMSDAIRY modelling framework refers only to soil structure and some general aspects of chemical fertility. Positive changes were observed for future climate change scenarios possibly due to a decrease in rain and thereby a decrease in soil poaching and erosion by grazing animals. This fact however needs further testing as although monthly rainfall rates could be smaller in future scenarios and for certain areas; the frequency of intensive storm events will increase. SIMDAIRY could not pick this up as it operates with monthly time-steps.

64

9.2

9 Climate impacts on welfare: expert review, transport and thermal challenges.

9.1 Introduction This section details the impact of, and potential adaptations to, climate change in the UK on the welfare of animals The format that this takes is the following: 1. an expert assessment of the impacts of predicted climate change on livestock

systems and livestock welfare, by livestock species 2. a review of the impacts of climate change on the welfare of animals in transport 3. an assessment of “thermal stress” that UK livestock will face in a changing

climate, including quantitative impacts on selected production and fitness traits using UKCIP02 data and modelling a selected extreme event.

Impacts of climate change on livestock welfare Weather and climate can directly and indirectly determine the welfare of livestock (Starr, 1980). Examples of direct influences include the heat balance of livestock and extreme meteorological events. Indirect influences are disease and parasites (section 10). Excessive heat or cold increases the metabolic energy required to maintain the animal’s body temperature, thus reducing the energy available for productivity and maintaining functional fitness of the animal. This requires an understanding of how environmental stressors (e.g., temperature, humidity, thermal, air speed) can directly and adversely affect animal performance, health, and well-being when coping capabilities of the animals are exceeded. The indirect consequences of weather episodes, such as feed quality and availability, must also be recognised, and are dealt with elsewhere in this report (section 7). Various indices based on meteorological measures (e.g., temperature, humidity, wind speed) have been developed and used to evaluate the thermal range for livestock (review: Hahn et al., 2003). All animals have a range of ambient environmental temperatures termed the thermo neutral zone (Figure 9.1). This is the range of temperatures that are conducive to health and performance. Many studies have explored the impact on livestock production and fitness traits of moving outside this zone and this type of information can be used to surmise the impact of climate change on the welfare of UK livestock. Specific examples will be given later. This section will focus mainly on the impacts on livestock of moving outside the thermo neutral zone. Setting aside the obvious impact of extreme weather events on livestock, there are other ways in which a changing climate can impact on the livestock systems and the welfare of animals. Section 3 describes that there will be stronger winds and higher rainfall in the west of the country in springtime, and this is where a high proportion of the UK’s dairy and beef cattle are farmed. This may have welfare consequences for animals as they begin to feel cold in wet and windy conditions. The lower critical temperature (LCT) is the point where the animal must increase its metabolic heat production to maintain homeothermy (“comfortable” body temperature) and therefore animals start to experience cold stress. The LCT of an individual animal is affected by its body size, condition score, metabolic state and hair cover.

Section 3 also describes how a changing summer climate is going to result in warmer summers. This is coupled with more favourable grass growing conditions in general and therefore grazing animals will have the opportunity to be at grass for longer period and will thus be more exposed to the changing climate. The upper critical temperature of the thermo neutral zone is the point at which heat stress effects begin to affect the animal. Heat stress can be simply defined as the point where the cow cannot dissipate an adequate quantity of heat to maintain body thermal balance. Impacts of heat stress on production, reproduction, health (including mortality), welfare and behaviour in many livestock species have been studied (review: Hahn et al., 2003). An example of the impact of temperature and humidity on heat stress in dairy cattle is given in Figure 9.2.

Figure 9.1 Schematic representation showing the relationship of thermal zones and temperatures.

Figure 9.2 Chart of the severity of heat stress in dairy cattle (Wiersma, 1990)

65

66

9.3 Impacts of climate change on dairy cattle Each of the following examples are based on the predicted changes in climate from UKCIP02 for England and Wales- why are we only considering England and Wales here when previous chapters have been looking at the RDP regions across the UK. Dairy cattle are generally housed during the winter and out at pasture from spring to autumn. The times of exposure to the changing climate will therefore be in the summer and in the spring. Table 9.1 summarises expert opinion and Table 9.2 summarises the impacts of a changing spring climate. Table 9.1 Summary of climate change in summer in dairy regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on mature dairy cows Av max

temp Av max humidity

Extreme temp

Associated humidity

Exposure level

Mortality Morbidity Yield* Fertility**

2020s 19.1 77.8 30 55 High Very Low

Very Low

↓ 5% ↓ 12%

2050s 20.4 75.3 31 53 High Very Low

Very Low

↓ 5.5% ↓ 12%

2080s 22.4 71.6 32 50 High Very Low

Very Low

↓ 6% ↓ 12%

*Adapted from Ravagnolo et al. (2000) **Igono et al. (1992) Table 9.2 Summary of climate change in spring in dairy regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on mature dairy cows Av wind sp Av rain Av temp Exposure

level Mortality Morbidity Yield

2020s 5.7 89.7 8.9 Med-High Low Low Reduced 2050s 6.4 87.8 9.7 Med-High Low Low Reduced 2080s 7.4 85.2 10.9 Med-High Low Low Reduced Table 9.1 shows that the warming summers are considered likely to have a negative impact on production and reproduction of dairy cattle in the UK, if no intervention (or system changes) takes place. Cook et al. (2007) showed that during times of heat stress dairy cows modify their behaviour, decreasing lying time and increasing drinking time. Table 9.2 shows that it is expected that dairy cows may not only suffer affects of heat stress during summer months but that during a spring grazing period they may also suffer cold stress. Igono et al. (1992) suggests that a cool period of less than 21oC will minimise the decline in milk yield. Dairy calves are generally housed all-year-round, but will be exposed to high temperatures even when housed. Low temperatures will not affect them as it is only when calves become wet that cold temperatures cause problems. Table 9.3, therefore, summarises the impact (as summarised by experts) of a changing summer climate on dairy calves.

67

Table 9.3 Summary of climate change in summer in dairy regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on dairy calves Av max

temp Av max humidity

Extreme temp

Associated humidity

Exposure level

Mortality Morbidity

2020s 19.1 77.8 30 55 Med ↑ 5-8%* Increased 2050s 20.4 75.3 31 53 Med ↑ 5-8%* Increased 2080s 22.4 71.6 32 50 Med ↑ 5-8%* Increased *Francos & Meyer (1983).

9.4 Impacts of climate change on beef cattle Beef cattle are normally housed in the winter and at pasture from spring to autumn. However, beef cattle are more likely to be out-wintered in the future, particularly if pasture availability improves (Chapter 9). Table 9.4 shows that during summer months, when beef cattle are likely to be at grass, the impacts on health traits are likely to be low as beef animals are better placed to deal with these temperatures. However, there are likely to be impacts on production traits with reduced live weight gain and therefore reduced market value and/or longer finishing periods. Table 9.4 Summary of climate change in summer in beef regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on beef cattle Av max

temp Av max humidity

Extreme temp

Exposure level

Mortality Morbidity Yield*

2020s 19.1 77.8 30 High Low Low Grow slower 2050s 20.4 75.3 31 High Low Low Grow slower 2080s 22.4 71.6 32 High Low Low Grow slower *Gaughan et al. (2008). Table 9.5 shows that if beef producers take advantage of opportunities to out-winter animals there may also be some negative impacts on production traits. However, it should be noted that out-wintering may result in reduced costs of keeping animals through the winter months and this yield loss may be offset. Table 9.5 Summary of climate change in winter in beef regions of England and Wales in the 2020s, 2050s and 2080s with predicted (from literature) impacts on out wintered beef cattle Av wind

speed Av rain Av temp Exposure

level Mortality Yield

2020s 5.8 128.6 7.2 High Low Grow slower 2050s 6.5 134.1 8.5 High Low Grow slower 2080s 7.6 142.0 10.3 High Low Grow slower

68

9.5 Impacts of climate change on sheep Currently sheep are predominantly distributed in northern and western parts of the UK. Sheep overlap with cattle in Wales and South West, and are more concentrated in Northern England and Scotland. In 2020 it is likely that the major impact of climate change on sheep welfare is that animals are likely to have to deal with increased frequency of extreme events at this point. It is important to note, however, that other factors/political issues such as CAP reform are currently having a bigger impact on sheep farming and sheep numbers than climatic variables. It is too early to judge the full impact of these yet, although they may have a greater effect on sheep welfare than climate will, especially in the short term. Chapter 4 shows that by 2080 temperature is likely to increase by 1.5 - 3.5 oC, it will be wetter in spring and winter, drier in summer and variable effects on wind speed. The new scenario temperatures are probably within the range that sheep can cope with by behavioural means - e.g. shade use, spatial behaviour and habitat use, as long as given the opportunity to express these behaviours. Rainfall could be important if it falls as critical times of the year - especially at lambing time. Also sheep find it hardest to cope with warm, wet weather, which is likely to lead to loss of condition and maybe increase in lamb mortality, plus increase in foot and leg problems. Extreme events are likely to be the most important: 1. Excessive rainfall, especially at critical times of the year - at lambing, immediately

after shearing, also preventing access to give supplementary feed/other management actions;

2. Drought in areas where sheep rely on natural water sources and plant moisture, may also lead to wildfires;

3. High winds at critical times especially lambing; 4. Sudden severe snowfall - animals cut off from food. Since nearly all sheep are extensively managed to some extent all sheep will be exposed to climate extremes and variability - warmer winters might lead to a reduction in housing over winter, although risks of extreme events might limit that. Ability of sheep to cope largely depends on their ability to express behavioural adaptations (e.g., using shade and shelter, exploiting different parts of the habitat) so hefted hill sheep should be able to cope well. Lowland/upland sheep might be more vulnerable in the short term as they are generally kept in more featureless terrain with less space to express behavioural adaptations. These animals are more likely to be housed, however, so may limit their exposure. Neonatal lambs are likely to be the most vulnerable group since even small increases in rainfall and wind speed at lambing time can increase mortality markedly. However, warmer summers/autumns, if they lead to an increase in the growing season, may improve maternal nutrition which can increase lamb survival, if weather conditions at lambing time are good. Very wet winters might affect maternal nutrition, by for example, limiting access to supply supplementary nutrition to hill sheep, which could reduce lamb survival in some areas. Neonatal lambs: Assuming a baseline of 15% pre-weaning lamb mortality (more than half this occurring within the first 24 hours of life) as best estimate of current mortality: extreme rainfall/wind speed at lambing could increase mortality to approximately 40% of all lambs born alive. Conversely, warmer winter and spring weather could reduce

mortality to nearer to 8% or less, due to both a prolonged growing season so better maternal nutrition, and fewer deaths from hypothermia if weather is good when lambs are born. However, warm, wet summer weather might increase mortality of older lambs due to infectious disease - more of an issue with housed animals than outdoor lambing. Adult sheep: Current baseline around 5-7% annual mortality - extreme rainfall leading to flooding or wildfires will have a catastrophic effect on survival if sheep are trapped and unable to escape but impossible to estimate losses. Evidence from Australia with drought conditions suggests this could be very high. In the absence of extreme events, an overall improvement in survival would be likely, unless disease conditions increase and are not controlled. If animals have the option to express behavioural adaptations then heat stress is unlikely to increase mortality markedly, potential issues in housed, late pregnant ewes, or unshorn ewes in featureless paddocks on some days of the year. Increased foot problems are likely if summers are warm and wet, so an increase in lameness may occur - current estimates are of around 10% sheep are lame annually. Mortality may not increase directly (although lameness can influence fertility and lamb survival, plus affect culling decisions), but an increase in suffering through lameness.

69

9.6 Impacts of climate change on pigs and poultry Around 855 million chickens are reared in the UK per year – over 90 per cent of the farm animals reared each year in the UK. There are approximately 5.2 million tonnes of poultry feed produced in the UK each year. As 60% of poultry feed is cereal this equates to 2.8 million tonnes of wheat incorporated into poultry feed each year or 65% of all feed wheat used in the UK. As such, the higher density of pig and poultry units tends to coincide with the areas producing cereals. Eggs are produced in caged (~63%), barn (~5%), free-range (~27%) or organic (~5%) systems. Meat birds are grown either in intensive systems (some 94%) or extensively (free-range and organic). But growth is in extensive free-range – for eggs 50% by 2012 and rising. In the UK, pigs are housed in both indoor and outdoor systems. Pigs in indoor systems are largely shielded from the effects of climate, as long as the limits of building design are not exceeded (McGlone & Pond 2002). Indoor pig farming takes place successfully in many regions of the world, including some (e.g. the midwest USA, eastern Europe) where seasonal temperature extremes are much greater than in the UK. Different categories of pigs have very different thermo neutral zones (Table 9.6, taken from McGlone & Pond 2002). Young piglets thrive when it is very warm, and are extremely vulnerable to becoming chilled in cold and/or wet and/or windy weather. Older pigs prefer cooler temperatures. Table 9.6 Acceptable temperature limits for productivity of pigs. Below the lower limit, animals will use feed energy to keep warm and feed efficiency will suffer. Above the upper limit feed intake will reduce and weight gain suffer (taken from McGlone & Pond 2002)

70

Since pigs and poultry are fed predominantly on processed cereals, the impacts of climate change elsewhere in the world (e.g. drought in Australia) and competition for land resources is likely to affect food crop yields and thus feed costs for UK pigs and poultry (Turnpenny et al. 2001). External factors such as these are likely to have a much larger effect on pig and poultry producers than the direct impact of climate change. Indoor pigs Hotter summers and extremely high temperatures will lead to heat stress in pigs. Heat stress results in reduced feed intake and growth in growing pigs. Wellock et al. (2003) estimated that the reduction in growth above a threshold of 25 degrees C was between 7 and 28 grams per degree C per day. Turnpenny et al. (2001) simulated an indoor pig farm comparing 1997 and 2050 scenarios, and estimated that a 20% increase in heat stress would occur despite a 10% increase in ventilation costs (increased use of existing fans). Changes in gross margin resulting from reduced growth rates due to this were estimated to be very small (Turnpenny et al. 2001), although an increased mortality risk is associated with high temperatures, this was not quantified. Effects on the breeding herd are probably more marked (see later for details on outdoor sows). Outdoor pigs Outdoor pig production works best on light, free-draining soils in low rainfall areas and is thus more popular in the east of the country, e.g. East Anglia, Yorkshire and Aberdeenshire. Primarily sows and pre-weaning piglets are kept outdoors. Growing pigs are moved indoors between weaning and finishing. Outdoor systems will, obviously, be more affected by climate change. During wetter winters there are likely to problems with the survival of neonates. Piglets, particularly young piglets are likely to suffer from chilling, thus leading to increased piglet mortality. Wetter, muddier ground could result in a greater risk of sow foot problems. There are also associated environmental problems with such ground conditions such as nitrate leaching. During hotter summers outdoor pigs are likely to suffer heat stress. Heat stress results in reduced reproductive output in sows (e.g., Jeon et al. 2006). Problems typically begin above the level of 25 deg. C where sows begin to pant (the evaporative critical temperature, Randolph et al. 2005). An experimental example of this was shown by Omtvedt et al. (1971) whereby litter size was lower by between 14-46% in hot conditions (37.8 deg C for 17hrs a day, 32.2 for 7h) versus cooler conditions. These results are extreme, but an extreme event would be likely to affect sows at these stages most dramatically. Pre-weaning piglet mortality can also be higher in hot temperatures. McGlone (1999) found that average mortality rose above 20% at around 25 deg C and kept rising at a rate of about 0.72 per cent per degree above that (e.g. reaching 27.2% at 35 deg C). In the UK, outdoor piglet mortality was at 18% in 2006, rather than the typical 15% (BPEX 2007). This has been put down to the unusually hot summer that year by some observers. There were anecdotal reports of sows farrowing in huts but then not returning to the huts because it was too hot, leaving piglets to starve. There were also anecdotal reports of sows farrowing in wallows, resulting in dead piglets. During the hot summer in late June/July 2006 on a typical 500 sow unit (White 2007), 6 sows (1.2%) died, and fertility suffered: there were 11% fewer litters and litter size was

71

9.7

also reduced by 0.9 piglets/litter. In this case, this single event resulted in the loss of 200 piglets at £30 per weaner pig cost this farm £6000. Milder winter temperatures might result in favourable impacts such that piglet mortality in outdoor systems may reduce due to reduced chilling of piglets. Berger et al. (1997) showed the variability of the percentage of liveborn piglet mortality across seasons, using French national data with the warmer months having the significantly lower piglet mortality than winter months (approximately 18 vs 24%). On the other hand, other researchers claim that the thermal microclimate within well-bedded farrowing huts is relatively constant despite extremely cold outside temperatures (Algers & Jensen 1990) suggesting that any benefit may be minimal. Poultry Temperature rises in the summer months are likely to impact on both growing chickens and laying hens. However, the problem is probably a bigger problem for chickens, who struggle to thermoregulate in the last 2 weeks of life (produce too much metabolic heat, more than they can dissipate). For layers, increased mortality from heat stress is already observed during mid 20 degrees and higher, In both cases, most sheds are not air conditioned, relying only on fans sucking in outdoor air to cool the shed. This not only poses a problem for bird welfare but also for system economics (bird losses). As with other livestock species, it is dealing with the extremes of temperatures that could cause the biggest problems, particularly extreme heat. Cold is less of a problem for birds, because they are able to heat selves up more easily than they can cool themselves down, but will cost producer (in terms of feed, which birds will eat more of, or in terms of gas/electricity to keep the temperature of the room up). For laying hens, studies have reported mortality rates during heat stress as 16% for brown hens (Franco-Jimenez et al., 2007). Heat stress also reduces egg production by 30% or more (Mashaly et al., 2004; this study also showed mortality increased from 5% to 32%). As stated earlier, the number of free-range poultry systems is increasing and expected to continue to do so. Increased rainfall at different times of the year is likely to adversely affect these systems, in particular, because increased rainfall will ruin the ground (cost for farmer). Laying hens are less likely to go out in heavy rainfall and therefore have the ability to range freely.

Impacts of climate change on livestock in transport. Animal transportation (by road, sea and air) constitutes one of the most important “acute” threats to animal welfare and productivity in commercial animal production. Good work over many weeks or months, in terms of animal housing and husbandry, can be undone in a matter of hours or days if transportation stress is excessive. In addition to the risks associated with animal handling, loading and unloading and the vibrations and accelerations experienced by animals in transit, the “on-board” thermal microenvironment represents a major source of transport stress and is the cause of increased mortalities, poor welfare, reduced production efficiency and product quality and decreased performance in livestock after completion of the journeys. Transportation of animals may take place over very short periods of minutes or hours but may also involve journeys of several days. It may be proposed that the thermal micro-environment within the transport container poses the greatest threat to the animals’ welfare and well being (Mitchell and Kettlewell 1998; Cockram and Mitchell

72

1999; Mitchell and Kettlewell 2003; Mitchell et al 2003; Mitchell 2006a and b). Adverse thermal conditions resulting in either heat or cold stress may lead to reduced welfare, overt tissue damage or injury and increases in mortality in transit (Mitchell et al 1997; Mitchell and Kettlewell 2002a and b; Mitchell and Kettlewell 2005; Mitchell 2006; Mitchell and Kettlewell 2008). Special attention should be paid to the incidence of increased thermal loads and heat stress as this is a recognized cause of transport stress in all climatic regions of the world. The thermal micro-environment in transport containers or vehicles may be complex and results from the interactions of several factors. These include the external climatic conditions, heat and water production of the animals, ventilation regimes, distribution and flow rates and any additional external sources of heat and/or moisture. The metabolic heat and water production of animals in transit is a major determinant of the "on-board" thermal environment and this can be of vital importance if external conditions and/or ventilation are likely to precipitate heat stress (Kettlewell et al 2001a and b). Thus, during transportation, the carriage of large numbers of closely packed animals will result in the addition of a great deal of heat and water vapour to the inside of the vehicle. In warm weather and increased temperature and humidity within the vehicle, the animals will respond by increased evaporative water loss through panting or sweating. The net effect is the creation of a hot, humid transport micro-environment in close proximity to the animals which imposes ever increasing demands upon the animals' thermoregulatory and other homeostatic systems (Kettlewell et al., 2000; Kettlewell et al 2001a and b). As external conditions become warmer or cooler then the risks of the transported animals being subject to heat stress or cold stress increases and this situation may be exacerbated by poor vehicle ventilation regimes. Fundamentally animal journeys by road are of two types:

(1) Animals transported to slaughter (2) Animal transported for relocation (breeding animals, sporting animals, animals

for fattening, animals involved in recreational activities)

Under current EU regulations journeys are also classified as (revisions to these journey limits have been proposed EU 2009):-

(3) Short – journeys of over 65 kms. and up to 8 hours duration (or up to 12 hours in the UK only

(4) Long – journeys of greater than 8 hours duration (which can only be undertaken on higher standard vehicles)

Climate change is likely impact upon all types of journey and the potential consequences must be considered in each category. The effects of climatic conditions and therefore climate change upon animals in transit differ from scenarios relating to other aspects of animal production as the animals, by definition are travelling. In addition to the animals experiencing a “modified transport environment” within the container or vehicle they may move through different weather conditions, even during short journeys. On long journeys, particularly export journeys for breeding stock (both intra-community trade and true exports to non-EU states) animals may be transported through different climatic zones. Road journeys from north to south in Europe may be subject to widely different weather conditions over the matter of a few days and very long journeys from west to east often involve exposure to large differences in thermal conditions due the “continental” weather conditions prevailing in Eastern Europe and beyond in Ukraine and Russia. All these areas are destinations for regular animal export journeys. These scenarios must be considered from a UK perspective as several species of livestock are routinely exported from the UK (e.g. breeding stock) to

73

9.8

parts of Europe and beyond and thus their welfare in transit and potential risks of reduced efficiency of production are important in relation to the animal breeders and producers and to the UK economy. When considering potential climatic scenarios affecting the welfare and production of animals in the UK many of the above factors must still be considered. Animals reared in the UK destined for slaughter will travel significant distances to the processing plants. In addition to routine slaughter journeys to preferred abattoirs there are other constraints which determine the nature of extended journeys. A good example of this is the transportation of “cull” sows from north eastern Scotland to south eastern England to a slaughterhouse designed and operated for this specific purpose. Such journeys take the full 12 hours allowed (the UK has a derogation enabling hauliers to travel up to 12 hours rather than hours on standard vehicles) and the pigs may experience widely varying thermal conditions depending upon the season. The transportation of spent laying hens is another issue as birds may travel large distances to the small number of slaughterhouses available for this procedure. As stated above the UK has significant trade with other European countries in relation to breeding stock and animals for further fattening and these animals may travel for several days across Europe and beyond. An important feature of Scottish animal production that impinges upon animal transportation is the rearing of livestock on outlying islands (e.g. Orkney, Shetland and the Hebrides) which requires their transportation by ferry from the place of origin to the mainland before their subsequent transportation to various destinations as indicated above. Thus the effects of potential climate change scenarios are complex and extend well beyond concerns for the influences upon the production conditions for each species whether outdoors or in environmentally controlled housing. Perhaps the climate change scenario of most concern in animal transport is a major alteration in the incidence of “extreme events” and increased frequency of extreme episodes. Thus very high temperature conditions or very low temperatures, even over a one day period, will cause problems for animal transportation. Extended periods of elevated temperatures (or very low temperatures) will constitute a threat to animal welfare (and indeed survival in transit) even on relatively short journeys to slaughter (up to 12 hours duration). For example the nature of poultry transport to slaughter (bird stocking densities, vehicle design and operation and tight slaughter scheduling) may be particularly vulnerable to unexpected high temperature or very low temperature conditions. At the moment hot summer (and spring and autumn) days are often associated with increased losses (mortalities or DOAs) in slaughter poultry accompanied by marked alterations in product quality. Extreme cold and wet days in winter often induce increases in mortalities and altered product quality. It is safe to assume that such incidents are also associated with a reduced welfare status of the transported animals.

Examples of thermal stress in transport The actual safe thermal envelopes for the road transportation of livestock are a matter of continuing debate and discussion in Europe. Current regulations (EC 1/2005 and WATO 2006) prescribe upper and lower limits for in transit temperature of 30° and 5°C respectively with a ± 5°C tolerance. Recommendations have been made for incorporation in to future legislation that will relate to each species and age of animal

74

separately and that include a decreased range of acceptable temperatures compared to those currently in the regulations and guidelines (EFSA 2004). The scientific literature relating to the safe thermal envelopes for the transportation of pigs, sheep, cattle, goats and horses has recently been reviewed (Mitchell et al 2009). The review concluded that the available scientific information indicates generally that the current thermal limits, particularly for high temperatures, require little if any revision. An important finding, however, was that acclimatisation of livestock to local thermal conditions might protect animals from the threat of heat stress at higher temperatures than those tolerated by animals reared in more temperate geographical zones. In relation to the current and proposed European legislation prescribing thermal limits for transportation of livestock a large study has been completed examining the thermal conditions inside transport vehicles on journeys in mainland Europe involving pigs, sheep, cattle, goats and horses (Fiore et al 2009). The study examined a total of 515 journeys involving movement of animals on long journeys (84% were >8 hours to 3 days duration) and some short journeys (16% - <8 hours). The important finding that under current environmental temperature conditions 20-25% of journeys exposed animals to temperatures beyond the limits prescribed in the current and proposed regulations in June, July and August and 10% in May and September. The problems were greatest for pigs and sheep and least for cattle. It is therefore clear that increases in mean ambient temperatures accompanied by an increased number of high temperature episodes or extreme events would exacerbate these existing problems. It is apparent that many journeys in Europe (including the UK) will take place outside the currently prescribed ranges. Elevated mean ambient temperatures and any increases in the frequency of episodes of elevated temperature (or very cold weather) resulting from climate change will increase the number of exposures of animal in transit to conditions in which welfare and productivity may be reduced. For example in broiler chickens it is known that at ambient temperatures greater that 24°C heat stress may occur and mortalities in transit will increase and product quality will decline (Mitchell and Kettlewell 1998). In young calves temperatures greater than 28°C and less than 5°C will cause transport stress and may compromise welfare (Mitchell unpublished results). As stated above, in addition to alterations in the number of potential heat stress and cold stress episodes increases in average temperatures as a consequence of climate change will also pose problems. Obviously animal production environments may be affected by these changes but the average thermal loads under which animals will be transported may also increase. This may mean that a larger number of journeys will be undertaken in conditions which are close to or above the recommended thermal envelopes in areas of the country and throughout Europe where previously heat stress in transit was confined to occasional episodes of hot weather. Thus, whilst in areas such as the northern UK normal journeys of short duration may not be detrimentally affected by increased average temperatures, longer journeys for both slaughter and export purposes, particularly to southern Europe or eastern states may be undertaken in climatic conditions more likely to induce heat stress. Finally in many areas the degree of adaptation or acclimatisation induced in resident livestock may be altered by climate change and this may have as yet little understood consequences for the effects of transport thermal micro-environments upon livestock Some of the potential problems (UK) described above are presented in Table 9.7

75

Table 9.7 Potential heat and cold stress problems affecting the UK Journey Type

Duration Causing heat stress Causing cold stress

Threshold Temps

Adaptations/ remedial actions

Slaughter (short)

</=8 (or 12 hours)

Increased average temperature Increased frequency of hot episodes

25-30ºC 25-30°C

Mechanically ventilated vehicles and containers Air conditioned vehicles

Slaughter (short)

</=8 (or 12 hours)

Increased frequency of cold episodes

0-5°C Heated vehicles Air conditioned vehicles

Slaughter (long)

Few in UK but common in Europe (upto 24 hours

Increased average temperature Increased frequency

of hot episodes

25-30ºC 25-30°C

Mechanically ventilated vehicles and containers

Air conditioned vehicles

Slaughter (long)

Few in UK but common in Europe (upto 24 hours

Decreased average temperature Increased frequency of cold episodes

0-5ºC <0ºC

Heated vehicles Air conditioned vehicles

Re-location and export (long)

>8 hours and up to 9 days

Increased average temperature Increased frequency of hot episodes

25-30°C 25-30°

Mechanically ventilated vehicles and containers Air conditioned vehicles Revisions of legislation Changes in maximum journey durations and travel times

Re-location and export (long)

>8 hours and up to 9 days

Decreased average temperature Increased frequency of cold episodes

0-5°C <0ºC

Heated vehicles Air conditioned vehicles Revisions of legislation Changes in maximum journey durations and travel times

The consequences of the above thermal challenges will include poor welfare, reduced product quality in slaughter animals and increased incidence of mortality in transit. Re-location journeys will apply to breeding stock, stock for further fattening and some intra-community slaughter trade using higher standard vehicles. It is clear that the objective in terms of thermal stress in transit is to ensure that the transport micro-environment is maintained in the thermal comfort zones for each species and age of animal. The thermal comfort zones may be defined in terms of the “thermo-neutral zone” for the animals and may be related to various indices of the overall thermal loads imposed upon the animals as described below. What is clear is that for the various climate change scenarios proposed that the risk of transportation of livestock under conditions that are beyond such prescribed limits is markedly increased.

76

An important feature of the transport environment other than the dry bulb temperature is the water vapour density. In general the effects of high ambient temperatures upon the animals’ well being and indeed survival are exacerbated by high absolute humidity. Thus it is critical that an integrated index of the “on-board” thermal environment during transportation is used to evaluate the risk of thermal stress. As described by Mitchell et al (2009) the use of temperature humidity index or THI, Enthalpy, apparent equivalent temperature or AET, wet bulb globe thermometer index or WBGT and other heat stress indices constitute useful approaches in this context (da Silva 2003). In order to assess the impact of climate change upon each of the livestock species of major importance it is essential that the effects of the alterations in average temperatures and the frequency of extreme events upon the total or integrated transport thermal micro-environment are considered and incorporated in to any predictive models. Humidity (or more strictly water vapour density) is critical in determining heat exchange between livestock and the environment and thus the extent of the heat stress experienced. This is particularly important in the transportation micro-environment where both temperature and water vapour content are modified by the animals and the nature of the transport container or vehicle may restrict behavioural thermoregulation and exploitation of convection regimes thus imposing effectively high thermal loads upon the animals than production based THIs might predict. This in turn reduces the applicability of the readily available thermal load indices directly in to predictions for the transport micro-environment including predictive models pertaining to differing climate change scenarios and possible impacts on the welfare of animals in transit.

9.9 Impacts of thermal challenges for livestock in transit The full range of impacts of climate change scenarios upon the various components of the animal transport process and the consequent welfare and production issues must be considered for all the livestock species of central significance. The major impacts are:-

• Increased mortality in transit or “dead on arrivals” or DOA • Reduced welfare status (fear, stress, injury, pathology) • Imposition of thermal stress (heat stress is the main concern) • Induction of extreme physiological stress involving tissue damage and overt

patho-physiology • Detrimental effects of physiological adaptive responses e.g. blood gas and acid

base disturbance associated with thermoregulatory responses • Reduced product quality in slaughter animals • Increased risk of dehydration In all livestock) and detrimental weight loss in

slaughter animals • Increased “losses” and waste due to the above impacts • Increased risk of transport associated pathology or disease e.g. shipping fever

in cattle and calves • Impaired immunological responses in relocated animals rendering them more

vulnerable to subsequent infection and disease • Impaired physiological function in relocated animals reducing post transport

performance and reproductive function Very little information (limited data – scientific, statistical or economic) is currently available upon the actual costs to the industry of the effects of transport mortalities, compromised welfare, product quality changes, and compliance with existing legislation. Indeed in the Defra Report quoted above (CC0361 (2008)) evaluation of

77

the economic effects of the impacts and adaptations that might be related to animal transportation were extremely limited due the lack of availability of “suitable data). Perhaps one method of examining the consequences of the various climate change scenarios is to address the effects upon mortality in transit (impact) for two intensively produced species that are known to be susceptible to heat stress and being intensively produced indoors are generally transported at legal but high stocking densities (for economic reasons) to the slaughterhouse or processing plant. These are broiler chickens and slaughter-weight pigs (100kg). Both species are produced throughout the UK with concentrations in some specific areas but the journey lengths or durations to slaughter vary considerably from <1 hour to up to 8 hours. The mortalities during these journeys have been extensively examined and surveyed and the association with journeys duration and journey conditions examined. It should be stressed that the literature indicates large differences between mortalities observed in different countries, at different times of year and even in different studies within the same country but generally the studies demonstrate the mortality is greater on longer journeys and greater when temperature increases. Warriss and Brown (1994) performed a survey of 2.9 million transported pigs in the UK over a 2 year period and reported an overall DOA of 0.061%. Losses of pigs in transit were greater in the summer months and described a relationship between mortality and temperature that was curvi-linear with a more serious detrimental effect of temperature above 15-17ºC. Warriss (1998) has recommended that trailer temperatures for pigs transported to slaughter should be maintained below 30ºC to remain within the thermo-neutral zone of the animals. Verecek et al (2006) examined losses in slaughter pigs in the Czech Republic over a 7 year period from 1997-2004. The study examined over 4 million pigs on journeys from 50 to over 300 km. distance in all seasons of the year. It was concluded that the highest losses occurred in summer in June, Jul and August when the temperature was highest and that this was the main factor in determining the reduction in welfare of the transported pigs and the mortalities observed. From the data provided by a study of 739 journeys to 5 different slaughterhouses Averos et al (2008) created a multilevel logistic model of the effects of factors affecting the mortality of pigs being transported in 5 different EU member states. The factors studies included average journey temperature in addition to journey duration. The mortality in transit increased significantly with temperature. Mortalities were higher in general and increased at temperatures above 14ºC. For a average journey temperature of 35ºC the predicted risk of mortality was 0.127% for a 30 minute journey and 0.133% of 24 hours. This compares to an average mortality of 0.11% across the whole study Dewey et al (2005) have examined transport losses in slaughter pigs during transport in Canada (Table 9.8). The journeys were of 200km or greater. Table 9.8 Mortality losses in slaughter pigs in transport over a range of temperatures (Dewey et al 2005

Dry bulb temperature (ºC) Number of deaths per 10,000 pig marketed

14 or less 15-19 20-22 23-25 26-28 29-31

14 24 43 42 59 76

78

It is apparent that losses increase with higher temperatures. It was asserted that a temperature of 27ºC but with a low RH (15%) would have the same impact as 23ºC and 75% RH. The mortality in those conditions would be 4 times higher than for pigs transported at 20-23ºC and low RH or 17-19ºC and high RH. What is clear from the published data is that whilst the current and proposed regulations relating to thermal conditions for livestock in transit may have been derived in order to protect the welfare of the animals there are indications that a temperatures lower than the prescribed thresholds mortalities in transit may increase. If the data presented in UKCP02 are employed and the low, medium and high emissions scenarios are considered in relation to mean temperature increases in northern and southern parts of the UK then for high and medium emissions the predicted mean temperature increases will have significant effects upon transport environments for 2020 onwards and for low emissions from 2050. It might be predicted that if the current transport regulations, vehicle designs and operation and features such as stocking densities and journey durations remained unaltered then very significant increases in transport mortalities of slaughter pigs would be inevitable. Based on the published work transport mortalities currently running in the UK at 0.03% might increase 10 fold or more. The corresponding effects upon animal welfare cannot currently be predicted but must be a major cause for concern. Similarly in broiler chickens journey duration and thermal load have marked effects upon mortality and DOA in addition to detrimental effects upon welfare (Mitchell and Kettlewell 1998; 2004; Mitchell 2006). The nature of the design of the broiler transport containers and the stocking densities employed creates a unique micro-environment in which heat stress may be precipitated well below current prescribed temperature limits (a fact recognised by the industry) and result in increased transport mortality. DOA rates of 0.1% increase when outside temperature is approximately 24ºC and may rise sharply as temperature increases further. In the UK transport at normal stocking densities on passively ventilated vehicles when ambient temperature exceeds 26-27ºC will result in extremely high mortalities. All the proposed increases in mean ambient temperature will result in summer conditions and high temperature events or episodes that would cause massive losses of broiler chickens in transit (days where ambient temperature >24º) if current vehicles and practices remained unaltered. Whilst these predictions cannot be reliably quantified at this time there are several reports (although not scientifically verified) of major increases in DOAs in the UK broiler industry of DOA rising to 10% on specific loads on single days due to an unpredicted high temperature episode. This magnitude of loss and the compromise of the welfare of the surviving birds would of course be totally unacceptable if the occurrence were to increase. It therefore must be stressed that whilst it is obvious that animal transport is one component of the animal production industry which is particularly vulnerable to the effects of climate change, specifically elevated mean ambient temperature and an increased incidence of potential heat stress episodes or extreme events there is little available data that allows accurate prediction of the effects upon the welfare of the animal, the losses to the industry and the costs of adaptations. It thus is essential that existing predictive models of livestock responses (Mitchell and Kettlewell 1998; Turnpenny et al 2000a and b; Fiahlo et al 2004; Mitchell 2006) to thermal loads and which allow assessment of acceptable ranges for thermal conditions in transit are integrated with climate change predictive models (e.g. similar to those employed by Parsons et al 2001; Turnpenny et al 2001) and economic models (e.g. St-Pierre et al 2003) to facilitate a much more comprehensive assessment of the climate

79

9.10

change impacts upon the welfare and production efficiency of livestock in the UK. Animal transport may constitute only a small temporal component of an animal production cycle but represents a period of very high risk to animal welfare and production efficiency. Transport is also a process which is extremely sensitive to the climate change effects and even small increases in mean temperatures and in the frequency of episodes of elevated temperature (extreme events) will impose excessive heat loads on animals in transit resulting in compromised welfare ,increased losses and mortalities and may economic losses. The accurate and reliable predictive modelling of such effects is vital for the provision of a sound scientific basis for and precise definition of sound strategies and adaptations for the various climate change scenarios.

Quantitative assessment of the impact of thermal challenges on livestock. As discussed earlier, many livestock species can have problems when they reach the lower or upper limits of their thermo neutral zone and suffer cold/heat stress as a consequence. This can adversely affect animal performance, health, and well-being when coping capabilities of the animals are exceeded. These are both detrimental to the animal itself as well as the overall system profitability. The aim of this section is to quantify the impact of moving out with the thermo neutral zone for livestock in the UK. Cold stress Climate predictions for Great Britain show that on average the climate will become warmer. However, there are many regional and temporal variations to this scenario. Animals are affected by cold weather at the lower limit of their thermo neutral zone. The ambient temperate at which animals begin to feel the cold is higher in wet and windy conditions. The following quantifies the impact for beef cattle of colder weather in relation to their lower critical temperature (LCT). The areas that had high densities of beef cattle were identified and predicted windspeed and air temperatures taken from UKCIP02. The NRC formulae were used to calculate the LCT of the types of animals that are likely to be at pasture during the spring months (March-May) (NRC, 2000). The LCT calculations take into account heat loss vs heat production (which is affected by the bodyweight, lactation or growing state of the animals, metabolisable energy intake, hide depth and coat depth), ambient windspeed and degree of wetness/muddiness of the coat. For beef cattle, youngstock, and both autumn- and spring-calving cows of large and small breeds were considered. The LCT’s were compared against the predicted average ambient air temperatures for each particular region. Windspeeds are predicted to be high for western counties, particularly those in the south-west of England. Beef cattle are likely to be exposed to the strong winds and increased rainfall that are predicted for the spring period (March-May). As temperatures rise and grass growth increases, it is likely that animals will be at pasture during this period. Table 9.9 shows the probability of experiencing cold stress (ambient temperature below LCT) in the classes of beef cattle likely to be at pasture during the spring. The models used to produce these indicators consider single animals in exposed conditions. In reality, when conditions are below the LCT, beef cattle are likely to lose weight as they use body reserves to produce heat and grazing is reduced as they often huddle with others or seek shelter.

80

Table 9.9 Table showing risk of cold stress during spring (March-May) during rain for categories of beef cattle likely to be grazing during these periods. Main beef-production areas are shown Animal class THI Threshold Production Reproduction Mortality Dairy cow 70 • DMI_loss

• Milk_loss • DO_loss • PDeath

Dairy heifers (0 to1) 77 • DMI_loss • Gain_loss

• PDeath

Dairy heifers (1 to2) 72 • DMI_loss • Gain_loss

• PDeath

Beef cows 75 • DMI_loss • DO_loss • PDeath Finishing beef 72 • DMI_loss

• Gain_loss • PDeath

Sows 74 • DMI_loss • DO_loss • ARate

• PDeath

Growing-finishing pigs 72 • DMI_loss • Gain_loss

• PDeath

Poultry broiler chickens 78 • DMI_loss • Gain_loss

• PDeath

Poultry layers 70 • DMI_loss • Egg_loss

• PDeath

Poultry turkeys 78 • DMI_loss • Gain_loss

• PDeath

Heat stress The study of St-Pierre et al. (2003) showed how the impact of heat stress in US livestock could be estimated, considering both biological and economic consequences. This methodology will be employed for estimating the impact of heat stress on production and fitness traits in UK livestock under the medium-high climate change scenario for 2020, 2050 and 2080 (UKCIP02). The base period, 1961-1990, was also modelled, however there was little or no impact of heat stress in livestock and therefore results are not presented. Also, there were little or no impacts of heat stress on livestock in Scotland over the time periods studied and therefore these results are omitted. The environmental conditions that induce heat stress can be calculated using the temperature humidity index (THI), which combines the ambient temperature and humidity. UKCIP02 monthly 50km2 weather data were taken for the base period (1961-1990), 2020s, 2050s and 2080s. From these data, the THI was calculated for each square in each time period. To account for the variation across a day (within a month), the maximum THI was calculated using the maximum monthly temperature and the minimum THI was calculated using the minimum monthly temperature. Each class of animal was assigned a THI threshold above which that class of animal began to suffer heat stress. A total of 10 classes of animals were defined, namely, dairy cows, dairy replacement heifers (0 to 1 yr and 1 to 2 yr), beef cows, finishing cattle, sows, market pigs, broilers, layers and turkeys. Additional variables were calculated to account for the extent and cumulative severity of heat stress across a day, namely the duration of heat stress (hours) the heat stress load (time summation in excess of threshold). For further details please refer to St-Pierre et al. (2003). Classes of animals that had little or no heat stress over the time periods will not be presented in the results. For each class of animal the biological response to heat stress was modelled based on the equations of St-Pierre et al. (2003), which were built based on an extensive

81

literature review. The traits modelled per class of animal are shown in Table 9.10. It is important to note that these models assume that system management does not change and that animal systems stay in the same region. For this study we are interested in the effects on production, reproduction and mortality, as these traits relate both to animal welfare as well as economic losses. For mortality, the model produces a value for the proportion of animals that die due to heat stress for each month. These values are then multiplied by the number of animals per year, to obtain a value for the deaths in each time period. The effects on milk yield and the days to conception in dairy cows are also analysed, as well as the loss in weight gain for finishing cattle and pigs, and the loss in eggs laid by poultry.

82

Table 9.10 Summary of the classes of animals, assumed THI threshold and the biological traits modelled in studying the impacts of heat stress on livestock Animal class THI

ThresholdProduction Reproduction Mortality

Dairy cow 70 • DMI_loss • Milk_loss

• DO_loss • PDeath

Dairy heifers (0 to1) 77 • DMI_loss • Gain_loss

• PDeath

Dairy heifers (1 to2) 72 • DMI_loss • Gain_loss

• PDeath

Beef cows 75 • DMI_loss • DO_loss • PDeath Finishing beef 72 • DMI_loss

• Gain_loss • PDeath

Sows 74 • DMI_loss • DO_loss • ARate

• PDeath

Growing-finishing pigs 72 • DMI_loss • Gain_loss

• PDeath

Poultry broiler chickens 78 • DMI_loss • Gain_loss

• PDeath

Poultry layers 70 • DMI_loss • Egg_loss

• PDeath

Poultry turkeys 78 • DMI_loss • Gain_loss

• PDeath

DMI_loss = reduction in dry matter intake from heat stress (kilogram per animal per day) Milk_loss = reduction in milk production from heat stress (kilogram per animal per day) Gain_loss = loss in body weight gain (kilogram per animal per day) Egg_loss = loss in egg production (kilogram per hen per day). (kilogram per animal per day). DO_loss = change in the average number of days open ARate = abortion rate PDeath = change in monthly death rate from heat stress Dairy Table 9.11 describes the impact of heat stress on dairy animals (cows) by region (further results given in Appendix 4). Generally, heat stress did not dramatically impact on biological performance until 2050 and was inevitability, greatest in 2080. The impacts on biological performance were greatest in the South and South East of England. There was no impact of heat stress on biological performance of young dairy animals (up to 1 year old). The impact on dairy heifers (1 to 2 years old) was small, with this class of animals experience heat stress for an average of 262 hours per year by 2080. The level and duration of heat stress in dairy heifers was predicted to reduce livestock gain by an average 180 g/animal/year and result in approximately 0.7 heifers in 1000 dying from heat stress by 2080 (see Appendix 4).

83

Table 9.11 Annual impacts of heat stress on the duration of heat stress (hours/cow/year) and selected production (Milk_loss: kgs of milk lost/cow/year), reproduction (DO_loss: increase in the number of days open/cow/year) and mortality (PDeath: number of deaths in 1000 cows due to heat stress) in dairy cows by region in 2050 and 2080

Region Year Duration Milk_loss DO_loss Pdeath EE 2050 186 1.3 0.42 1.25 2080 601 21.5 3.19 2.87 EM 2050 132 0.4 0.19 1.04 2080 516 15.7 2.58 2.54 NE 2050 0 0.0 0.00 0.00 2080 111 0.0 0.05 1.71 NW 2050 0 0.0 0.00 0.00 2080 170 1.2 0.37 1.45 SE 2050 350 4.0 1.13 1.77 2080 852 41.9 5.28 3.73 SW 2050 137 0.6 0.27 0.87 2080 495 15.4 2.33 2.59 WA 2050 52 0.2 0.09 0.37 2080 290 7.3 1.23 1.78 WM 2050 138 0.3 0.15 1.23 2080 497 14.8 2.48 2.46 YH 2050 0 0.0 0.00 0.00 2080 245 1.6 0.54 1.57

EE = East England; EM = East Midlands; NE = North East; NW = North West; SE = South East; SW = South West; WA = Wales; WM = West Midlands; YH = Yorkshire and Humberside Dairy cows have a higher level of metabolic activity than younger dairy animals, in that they are producing high levels of milk, maintaining pregnancy as well as maintaining immune function. This means that their thermo neutral zone is smaller that other categories of livestock. Overall, dairy cows are predicted to experience 111 hours of heat stress per year in 2050, which rises to 420 hours by 2080 (Table 9.11; duration of heat stress in 2080 ranges from 111 hours/year (North East) to 852 hours per year (South East)). Milk yield was predicted to reduced 13.2 kg/cow/year (0 – 42 kg) in 2080 due to the impact of heat stress. Reproductive performance of dairy cows was also predicted to get worse, with the numbers of days open increasing by 2 days/cow/year (0.05 – 5.28 days) by 2080. Reproductive problems are one of the main reasons for involuntary culling in the dairy herd and therefore, in regions where the impact of heat stress is large, the rate of involuntary culling may increase. Increasing the level of heat stress in dairy cattle over time is predicted to have an unfavourable effect of livestock survival, in that approximately 2.3 cows in 1000 are predicted to die due to heat stress in 2080. The financial costs of these impacts are calculated and presented in Chapter 13 on impact costs. The following section discusses impacts on the other animal categories. Beef Cattle The impact of heat stress on beef cows over the time periods studied are relatively small with only beef cows in East England and the South East experiencing a small amount of heat stress with an average of 55.5 hours/cow/year. This has little or no impact on the biological traits studied. The results for the impact of heat stress in beef finishers and beef cows for 2020 and 2050 are shown in Appendix 4, although only a

84

proportion of regions are predicted to result in a scenario where beef animals are suffering from heat stress. Table 9.12 shows the impact of heat stress in selected biological traits in finishing beef animals by region in 2080. Overall, beef animals in England and Wales are predicted to be experiencing a total of 240 hours of heat stress per year, nearly half the duration that dairy cows experience. The impact of this heat stress will affect production by, on average, 270 g of live weight gain/animal/year. Although small, this would be cumulative across the herd and in the South East, for example, is predicted to be 710 g. The impact of heat stress of the survival of beef animals is lower than seen for dairy cattle with a relatively consistent prediction across regions of 0.8 animals out of 1000 dying as a result of heat stress per annum. Table 9.12 Annual impacts of heat stress on the duration of heat stress and selected production (Gain_loss: reduction in kgs of live weight gain/animal/year) and mortality (PDeath) in finishing beef animals by region in 2080 (Further results in Appendix 4)

Region Duration Gain_loss PDeath EE 291 0.340 0.82 EM 260 0.231 0.87 NW 40 0.002 0.43 SE 429 0.715 0.91 SW 329 0.400 0.89 WA 276 0.247 0.88 WM 250 0.219 0.81 YH 48 0.004 0.43

Pigs and poultry Table 9.13 shows that selected classes of monogastric livestock will suffer heat stress, to some extent, by 2080. The duration of heat stress that animals will undergo in a year by 2080 is 92 hours for sows, 214 for growing pigs and 420 for laying hens. The difference in the duration of heat stress reflects the difference between these classes of animals in terms of their metabolic rate with layers and their systems of production being highly efficient with a high turnover of feed into product. The overall impacts of heat stress differ across livestock class as well. The impact of heat stress on production traits is greatest in the layers, followed by the growing/finishing pigs, with an average of reduction of 24 g of egg/bird/year in layers and 425 g reduction in live weight gain/animal/year in pigs. The impact of heat stress on survival varied to a lesser degree across species with 1, 0.7 and 1.2 animals out of 1000 dying as a result of heat stress in sows, growing pigs and layers respectively.

85

Table 9.13 Annual impacts of heat stress on the duration (Dur) of heat stress and selected production (Gain_loss; Egg_loss: reduction in kg eggs/animal/year), and mortality (PDeath) in pigs (sows and growers/finishers) and poultry (layers) by region in 2050 and 2080

Pigs - Sows Pigs – Growers/Finishers Poultry – Layers Reg Year Dur Pdeath Dur Gain_loss Pdeath Dur Egg_loss Pdeath EE 2050 0 0 8 0.002 0.06 186 0.005 0.62 2080 173 1.72 291 0.601 0.82 601 0.038 1.44EM 2050 0 0 0 0 0 132 0.003 0.52 2080 149 1.72 260 0.409 0.87 516 0.031 1.27NE 2050 0 0 0 0 0 0 0 0 2080 0 0 0 0 0 111 0.001 0.86NW 2050 0 0 0 0 0 0 0 0 2080 0 0 40 0.004 0.43 170 0.005 0.72SE 2050 0 0 45 0.009 0.43 350 0.014 0.89 2080 230 1.73 429 1.266 0.91 852 0.062 1.87SW 2050 0 0 0 0 0 137 0.003 0.44 2080 106 1.29 329 0.708 0.89 495 0.028 1.30WA 2050 0 0 0 0 0 52 0.001 0.19 2080 102 1.72 276 0.438 0.88 290 0.015 0.89WM 2050 0 0 0 0 0 138 0.002 0.62 2080 67 0.86 250 0.387 0.81 497 0.030 1.23YH 2050 0 0 0 0 0 0 0 0 2080 0 0 48 0.008 0.43 245 0.007 0.79

9.11 Summary There are a number of challenges that a changing climate will bring to livestock systems, particularly related to thermal challenges. This thermal challenge will range from cold stress, which will be exacerbated by wetter and windier winter months to heat stress, exacerbated by warmer summer months. These thermal challenges will affect livestock systems, impacting on livestock production, health and welfare and may lead to increased mortality rates due to thermal challenges year round. Not only will thermal challenges impact on livestock systems but also on livestock in transport, when external weather conditions can be compounded in transport containers when full of animals. Also, some of the impacts of climate change on animal production and functionality (e.g., health and welfare) may be further exacerbated by adaptations farmers make in other areas of the farming system, such as taking advantage of a longer growing season and keeping animals outdoors for longer periods, thus increasing livestock exposure to prevailing weather conditions. It is important that livestock systems, producers and transporters are aware of these risks and account for them when planning for the future. The results presented here in quantifying the impact of heat stress in the future highlight a number of factors. First, there is considerable regional variation. The most affected region for all animal types is the South East, which is projected to see the greatest climate change impacts. Scotland is not expected to experience any losses due to mortality from heat stress in any of the time periods. Impacts start occurring predominantly from the East Midlands south.

86

Many livestock sectors will not be affected by heat stress until the 2080s, however some, notably dairy and poultry, begin to see some losses as early as 2020, and in these two sectors only Scotland is unaffected by the 2080s. Sheep have not been included in the analysis of the impacts of heat stress as models/data on UK breeds of sheep were not available. It is likely that the extensive nature and locations of many sheep systems in the UK will be subjected to limited period that would induce heat stress and provide opportunities for animals to dissipate the effects by both behavioural modification (e.g., panting) or utilise natural (or artificial) environmental shading. For sheep, however, the lower threshold of the thermo neutral zone may mean that they are more vulnerable to cold stress (not shown). This chapter has generally focussed on gradual warming conditions with quantification of the future climate change on heat stress in livestock, assuming UKCIP02. However, the impact of extreme events rather than gradual warming may have a larger impact on livestock systems and provide a more critical “shock” to the system. UKCIP02 does not allow the modelling of such an extreme event, however it has been suggested in UKCIP02 and UKCP09 that extreme events are likely to increase in frequency. Table 9.14 shows the impact of heat stress due to gradual warming in the month of August compared to impact of heat stress due to gradual warming with a 5 day heat wave in the month of August in the time periods 2020s and 2080s. A heat wave was defined as a period of 5 days (or more) when the average daily temperature was 3oC higher than the average for that time period. UKCIP02 daily values were adjusted to represent this change and the subsequent effect on duration of heat stress and production, reproduction and deaths in diary cattle were modelled. Table 9.14 Impacts of heat stress in the month of August with and without and heat wave on dairy cows by region in 2020 and 2080 Production (Milk_loss: kgs of milk lost/cow in August), reproduction (DO_loss: increase in the number of days open/cow/) and mortality (PDeath: number of deaths in 10000 cows due to heat stress)

2020 2080

Region Milk

kg lost DOpen days

Death /10,000

Milk kg lost

DOpen days

Death /10,000

EastEng 0.0 1.4 0.00 0.22 0.00 1.51 12.8 24.4 1.68 2.13 9.53 9.82 EastMid 0.0 0.9 0.00 0.17 0.00 1.48 9.5 19.6 1.41 1.93 9.36 9.70 N_Irel 0.0 0.0 0.00 0.00 0.00 0.00 0.0 0.2 0.00 0.07 0.00 1.42 NEast 0.0 0.0 0.00 0.00 0.00 0.35 0.0 1.5 0.03 0.24 8.58 8.70 NWest 0.0 0.1 0.00 0.02 0.00 0.70 0.8 3.7 0.22 0.53 8.69 8.86 Scot 0.0 0.0 0.00 0.00 0.00 0.00 0.0 0.4 0.00 0.09 0.00 1.43 SEast 0.0 2.5 0.00 0.32 1.71 3.00 23.9 38.7 2.56 2.80 10.06 10.22 SWest 0.0 0.8 0.00 0.14 0.00 1.46 9.4 19.0 1.28 1.75 9.29 9.59 Wales 0.0 0.3 0.00 0.05 0.00 0.65 4.5 9.2 0.68 1.03 8.95 9.16 WestMid 0.0 0.9 0.00 0.16 0.00 1.47 9.3 19.2 1.39 1.92 9.35 9.69 Yorks 0.0 0.1 0.00 0.03 0.00 0.88 1.1 5.0 0.34 0.70 8.75 8.96 On the whole, the addition of the extreme event, in this case a heat wave, resulted in a greater level of losses compared to the gradual warming scenario only during the month of August. The addition of the heat wave scenario increased the metabolic stress of dairy cows in all time periods, with nearly all regions experiencing some period of heat stress by 2020 (national average of 23.4 hours) compared to the gradual warming scenario where only cows in the (national average of 1 hour) South East

87

experienced heat stress (results not shown). This heat stress impacted on production and fitness of the animal which can be seen in the numbers of lost due to heat stress increasing 10 fold in the 2020s compared to the gradual warming scenario. In the gradual warming model neither Scotland nor Northern Ireland experienced any heat stress in any of the time periods. However, adding a heat wave meant that by 2050 (not shown) and 2080 both these regions would experience losses due to heat stress if a heat wave occurred. This single heat wave modelled was at the lower limit of a heat wave definition in terms of duration and increase in daily temperatures. However, heat waves could last for longer periods, have higher dairy temperatures and may occur more than once a year thus compounding the impacts on animal performance.

88

10.1

10.2

10 Impacts of climate change of disease risk to livestock

Introduction Many pathogens and parasites of livestock have lifecycle stages, vectors or intermediate hosts that are sensitive to climate and thus it might be expected that climate change would have fundamental and far reaching impacts on their epidemiology. This section considers the likely impacts of climate change on animal health. There is much uncertainty about the likely climate-induced impacts of disease and it is impossible to undertake a full integrated valuation of the likely climate-incidence of diseases. This section does scope the categories of disease risks but the draws on illustrative examples to provide a preliminary health impact and likely adaptation needs.

Background Many livestock parasites and pathogens are transmitted via the environment and are thus exposed to and affected by climate. Exposure to the sun (UV radiation) affects the survival of bacterial pathogens (e.g. bovine tuberculosis & e-coli O157) in the environment. Similarly, the rates of development of gastrointestinal nematode parasites (from eggs in faeces to infective stage larvae on pasture) are temperature and humidity regulated and often highly sensitive to desiccation. There is the potential for an increase in many livestock pest and disease problems due to less ‘winter kill’ and longer disease seasons as they can persist in the environment for greater proportions of the year. Both increases in the number of generations of parasites and treatments for parasitism, can increase the risk of development of resistance to drugs and pesticides, further exacerbating the direct effects of climate change. However, despite these well known effects of climate on parasite and pathogen survival in the environment, to-date there is no agreement between leading researchers on whether, at the global scale, the amount of disease will change (Lafferty, 2009a; 2009b). A recent series of discussion papers highlighted the varying opinions of researchers on the effects of climate change on disease risk (See Lafferty 2009a and subsequent forum in Ecology, 2009, 90, No4, Introduced by Wilson 2009). Whilst there was much disagreement on how climate change will affect the amount of disease, Wilson, (2009) summarised the areas where there was agreement:

(1) climate change is altering the geographical distribution and incidence of (at least some) infectious diseases and will continue to do so,

(2) detecting a climate signal in disease-range changes is likely to be difficult because of the influences of other confounding factors, such as changes in land use, socioeconomic factors, vector control strategies, and health care practices,

(3) better data sets and modelling approaches are required to be able to make robust predictions of the impacts of climate change on disease dynamics,

(4) whether or not specific infectious diseases expand or contract their geographical ranges will depend not only on extrinsic factors (e.g. climate change), but also intrinsic factors (such as immunity, phenotypic plasticity, and evolution).

89

10.3

It is with this background of accepted unknowns and uncertainty that we consider the potential impacts of climate change on livestock disease risk in the UK. What we do know from past experience is that episodes of certain livestock diseases can lead to heavy economic impacts within and beyond the sector. As an adaptation strategy, the limited evidence suggests that disease surveillance can deliver high social rates of return and is therefore a prudent capacity investment.

Qualitative predictions of future disease risk to the UK Expert opinion Given the above mentioned complexity and current difficulties in quantifying the future distributions and prevalences of diseases threatening the UK, literature review and expert opinion is preferred to a systematic downscaling of existing climate data. A combination of information can be used to identify potential high risk parasites and pathogens (for example see Gale et al. 2008). Whilst there are a number of publications that discuss future disease threats to the UK, few if any include lists of high risk livestock diseases. Appendix 5 lists the potential disease threats to UK livestock based on SAC veterinary expertise. The appendix provides lists of viruses, vector borne disease and parasites sensitive to the environment and that were considered to be a potential future risk to the UK livestock industries. For completeness, Appendix 5 includes a list of plant toxicoses that have the potential to change in incidence in the UK. For example under wetter and warmer conditions grain to be used for pig feed may be more susceptible to moulds and thus contamination with Mycotoxins that impact on production but also food safety (For example. Fusarium graminearum is predicted to become a greater threat in the UK). The lists are a starting point as they include the livestock parasites and pathogens that may be influenced by climate change. Ongoing research aims to develop these lists further in consultation with wider expert opinion to add evidenced argument to rank the parasites and pathogens in relation to their relative risk to the UK. Clearly the information in the appendix could be the starting point of an extensive ex ante assessment of economic damages in the event of an outbreak or spread of specific diseases. Equally, an adaptation strategy can be mapped out for the eventually of any single or disease outbreak combination. For illustrative purposes, Appendix 5 details relevant ex ante information for Blue Tongue virus. Assessment of notifiable diseases To compliment the expert opinion above, we considered the potential future risks to the UK from each of the diseases classified as notifiable by the World Organisation for Animal health (OIE). Threats are deemed as those OIE notifiable diseases that are already present in the UK or are increasing in range toward the UK (see Appendix 5). This much shorter list of parasites and pathogens highlights those that are already in the UK and may be affected by changes in livestock distribution and management (e.g. Mycobacteria, see below) and those that are close to and spreading toward the UK (e.g. Blue Tongue). For non-vector borne pathogens out with the UK, climate change may actually have little effect on whether or not the pathogen enters the UK as there are often effective barriers to introduction (e.g. animal movement regulations and testing, and sanitation). This is especially true for pathogens that have previously been eradicated from the UK

90

10.4

as it has already been demonstrated that they are not limited by the UK’s climate. In contrast, it is difficult to control the majority of insect vectors whose range and prevalence is largely dictated by climate and that reproduce and adapt rapidly, are highly mobile and are both habitat and host generalists. These qualitative approaches pick out the known imminent threats to UK livestock e.g. Blue tongue, as well as potential future threats, as climate change in the UK creates habitats suitable for parasites and pathogens and their vectors that are currently restricted to southern Europe and further afield. These lists can be used to provide target parasites and pathogens for increased surveillance within and outside the UK. Whilst quantitative approaches can provide more detail on the likelihood and timescales of parasite and pathogen risk to the UK, they are often reliant on data. Quantitative approaches would benefit greatly in terms of improved predictive power from long term data sets on the high risk parasites and pathogens identified from these qualitative studies (see below). Furthermore, as it may not be possible to track the movement and spread of all the parasites and pathogens that represent a potential future threat to the UK, any improvements in UK capability in identifying initial outbreaks of exotic diseases are likely to provide disproportionately greater benefits in terms of control and economic loss.

Quantitative approaches to predicting disease risk A technique often used to consider the effect of climate change on biological systems is climate matching, also known as climate envelope modelling. In a disease context, this technique considers the epidemiology of parasites and pathogens in countries with current climates similar to those predicted for the UK under climate change scenarios. All else being equal, the UK might then be expected to experience the same levels of parasite/pathogen burdens and the resultant productive losses as the reference country (climate envelope). This approach is highly dependent on data and the confidence that can be attributed to the predictions must be assessed in relation to reliability and detail of the data that in itself determines the analytical approach. A basic climate matching analysis to determine the risk of Haemonchus contortus to the UK sheep industry suggests that this highly pathogenic nematode parasite may increase in incidence and intensity to levels comparable with levels in New Zealand (see Appendix 5). Gastrointestinal parasites are perhaps the most pervasive challenge to the sheep industry. Mixed nematode parasite infection (including Haemonchus contortus) in sheep in New Zealand was estimated to reduce lamb growth by 60% and fleece production by 30%. Currently nematode parasites in the UK are controlled largely through the use of anthelmitic drugs. However, parasites develop resistance to the drugs and the Moredun Research Institute estimated that anthelmintic resistance costs the UK sheep industry £65 million per annum. The recent release of a new anthelmintic, Zolvix, will in the short term help to control nematodes, including emerging species such as H. contortus. However, parasites are expected to develop resistance to this new drug, and the rate of development of resistance is a function of the level of use. Research is needed to maximise the effective lifespan of new antiparasitic drugs (i.e. alternative parasite control options to limit the use of the drugs) as climate change driven changes in the UK parasite communities are likely to add a further stressor to the system. The developers of Zolyix, which is not yet released in

91

the UK, already apparently have an expected timeline for the development of resistance. This timeline is overtaking drug development. However, not all parasite, pathogen and pest problems in the UK might be expected to increase. The sheep parasite Nematodirus battus requires a period of cold weather to complete its lifecycle. As such, all else being equal, warming scenarios are predicted to be associated with a reduction in risk to UK sheep populations. The limitation of the climate envelope modelling approach is that it does not include intrinsic factors such as the ability of parasites and pathogens to adapt to new climates and or new host communities. For example, Nematodirus battus may be able to adapt to the relatively gradual change in climate and thus persist in the UK (van Dijk & Morgan 2008). At present N. battus is one of the most pathogenic parasites of sheep. The worst case scenario for the UK livestock industries is that the current pathogenic parasites and pathogens (e.g. N. battus in sheep) adapt to the changing climate and persist whilst at the same time exotic and emerging parasites and pathogens (e.g. H. contortus in sheep) increase in incidence and intensity of infection. Effects of climate change on parasite and pathogen exposure rates of livestock in grazing systems

Livestock in agricultural grazing systems will come into contact with faeces whilst they are grazing. Both macroparasites (e.g. parasitic helminths) and microparasites (e.g. bacterial pathogens) are found in host animal faeces and thus can be transmitted to herbivores via grazing. Pathogen transmission from faecal-contaminated vegetation can occur via two main routes, ingestion of faecal-contaminated vegetation, and investigation of contaminated vegetation. Thus, the behavioural contact pattern of grazing herbivores with faeces in the environment plays an important role in the risk of parasite/pathogen transmission via the faecal-oral route. However, herbivores generally avoid grazing near faeces (Forbes and Hodgson 1985; Benham and Broom, 1991; Hutchings and Harris, 1997; Bao et al., 1998; Hutchings et al., 1998), and will modify their grazing behaviour to become more selective when forced to graze faecal-contaminated forage (Hutchings et al, 1998) or pasture spread with slurry (Pain et al, 1974; Broom et al, 1975; Pain and Broom, 1978; Swain et al, 2008). This selective grazing behaviour affects the sward structure of the grazing environment, creating a heterogeneous landscape of gaps (short, non-contaminated, grazed patches) and tussocks (tall, faeces-contaminated, avoided patches) (Hutchings et al., 2001a,b). Additionally, the faecal deposits contain nutrients that leach into the surrounding area causing tussocks to have increased nutrient content relative to gap swards (Haynes and Williams, 1993). Thus the mosaic represents a nutrition versus parasitism trade-off in that the faeces-contaminated tussocks are localised concentrations of nutritional resources, providing herbivores with up to 32% increased forage intake rate and 41% increased nitrogen content. However tussocks also contain up to 17 times greater numbers of nematode parasites (Hutchings et al., 2007). Livestock must make grazing decisions based on the costs and benefits of this trade-off, which are affected by both animal factors (e.g. physiological state) and environmental factors (e.g. forage availability) (Hutchings et al., 1999). Thus, in agricultural grazing systems it is the interplay of these factors that will determine the livestock parasite exposure and infection risk.

Different grazing management systems are used to provide livestock with a supply of herbage while effectively utilising the productivity of the pasture (Holmes, 1989). For example, continuous set stock grazing and rotational grazing. In a continuous set

92

stock grazing system livestock have access to an area of pasture for extended periods of time and grass growth is about equal to animal intake. In contrast, rotational grazing involves dividing the pasture area up into a number of similar-sized paddocks and the livestock are moved in a regular sequence between the paddocks. Herbivore grazing behaviour in relation to faeces and thus parasites and pathogens, has been shown to be affected by the grazing environment (e.g. nutritional environment) (Smith et al., 2006). In terms of livestock grazing management the overall amount of cattle contact with faecal-contaminated patches is similar in both set stocking and rotational grazing scenarios, suggesting no difference in the risk of infection between the different grazing systems (Smith et al. 2009). However, the timing and absolute amounts of peak contact varies greatly indicating that different grazing management systems expose livestock to risks of different types of parasites at different times of the grazing season. Intensive rotational systems with small pasture blocks (especially the first grazing period) maximise livestock contact with fresh faeces, and thus exposure to microparasites (e.g. bacterial pathogens). Livestock re-entering pasture blocks in rotational systems and set stocked livestock have the highest contact with old faeces and thus have a greater risk of macroparasite transmission (gastrointestinal nematodes) (Smith et al. 2009). This highlights how livestock management affects the highly dynamic interaction between livestock and distributions of parasites in the environment and thus the levels of livestock exposure to parasites and pathogens via the faecal-oral route. Climate change is likely to affect the spatial and temporal distribution of forage resource (see section 7) and parasites in the environment. The knock on consequences of predicted changes in forage availability and farm management practices in response to climate change on parasite exposure rates are discussed below. However, before we extrapolate from knowledge of the patterns of parasite and pathogen exposure in agricultural grazing systems, it is necessary to establish whether climate change is likely to create non-linear effects on exposure rates and thus disease risk. For example, at the landscape scale, preferred wetter grazing habitats can become highly contaminated with faeces and thus nematode parasites that survive for longer periods in the wet conditions potentially resulting in disproportionately greater parasite exposure. (van der Wal et al. 2000). Changes in habitat type and distribution driven by climate change may thus have disproportionately greater effects on the levels of contact between hosts and environmental distributions of parasites and pathogens and thus exposure and risk. Fasciolosis (or Fascioliasis) (Liver fluke) is a major parasitic disease of livestock with more than 700 million production animals at risk of infection worldwide. It leads to substantial annual losses to livestock and food industries, worth around € 2.5 billion with additional penalties to their health and welfare (Spithill et al, 1999). In temperate areas it is commonly caused by the liver fluke Fasciola hepatica. Adult parasites reside in the bile ducts in the liver, and they cause the chronic disease, which leads to anaemia, reduced body condition and poor fleece quality in sheep. Acute form of the infection is also common in sheep; it is caused by immature parasites migrating in the liver of the host, which is often responsible for sudden deaths (Mitchell, 2007). There has been an increase in the incidence of the disease in sheep populations in the UK since 1999, and a suggestion that the disease is spreading to previously unaffected regions (Mitchell, 2002; Nadis Health Bulletin, 2007). Furthermore, the number of outbreaks of fasciolosis has increased dramatically in Scotland in recent years, and reached unprecedented levels in 2002/3. Figure 10.1 demonstrates that liver fluke outbreaks in sheep during that period, accounted for almost 30% of all diagnosable submissions to SAC's Veterinary Services. This increase in disease incidence is believed to be associated with changes in climate, notably an increase in temperature

and rainfall. These changes potentially predispose to fasciolosis, by influencing the intermediate host of the parasite Lymnaea truncatula by extending and prolonging the period and distribution available for the development of the intermediate stages of the parasite. It is noted that the high levels of disease incidence reported in 2003 (Figure 10.1), coincided with temperatures exceeding average values in ten out of twelve months, and rainfall was above average in nine out of twelve months during 2002. Adult liver fluke produce eggs, which are excreted with the faeces and subsequently hatch when the environmental temperature exceeds 10° C (typically between Mid May- September). The larvae invade the intermediate host and develop to cercariae. Typically, the complete life-cycle of the liver fluke in the intermediate host lasts between 2-3 months. Similarly, the duration from the time of infection to the maturation of parasites in the definite host, is around 3 months. Consequently, chronic disease is usually seen in sheep flocks between winter and spring.

Sheep Fluke Outbreaks as %age of Diagnosable Submissions

0%

5%

10%

15%

20%

25%

30%

35%

January February March April May June July August September October November December

2003200420052006

Figure 10.1 Sheep fluke outbreaks in Scotland during 2003-2006 from the SAC veterinary records

Appendix 5 details a process based modelling study of the effects of the amount and distribution of wet habitats in agricultural grazing systems on the exposure of grazing livestock to parasites and pathogens via the faecal oral route and to liver fluke. The model outputs show the increased exposure to both parasites and pathogens transmitted via the faecal-oral route and liver fluke with increased amounts of wet habitat. These models do not include the improved survival of parasites and pathogens in wet environments as a result of reduced rate of desiccation, and so these effects are likely to be conservative. However, the potential for non-linear effects of wet habitat distribution on parasite exposure rates seen at the landscape scale are not realised at the agricultural field scale. This is likely due to the high rates of movement of the livestock enabling them to encounter the vast majority of the available patch choice options within a field on a daily basis (Marion et al. 2008). Here we have not

93

94

10.5

considered the high level of temporal variation in weather patterns (including extreme events such as droughts) and thus habitat suitability for environmental stages of the parasites nor have we considered the epidemiological impacts at the population level (of hosts and parasites), both of which would improve our understanding of the effects of climate change on livestock disease risk in grazing systems of production.

Changing management in response to a changing climate As climate changes, farm management practices are also likely to change as farmers adapt to the new conditions. Livestock management drives the contact processes between hosts (livestock and wildlife) and parasites whether through rates of population mixing (i.e. risk of direct transmission) or rates of contact with environmental distributions of parasites and pathogens i.e. risk of indirect transmission). Climate affects on direct transmission: Predicting the changes in livestock distributions in the UK driven by climate change was outside the remit of this project. However, the environmental impact of future climate scenarios on the suitability of he UK for livestock production e.g. the ability to grow forage for livestock (see section 7) is likely to alter the business opportunities associated with food production and some movement is envisaged. For example, a decrease of 50% in river flow in the summer and autumn months is projected, reaching an 80% decrease in some areas. In contrast the river flow in the winter months could increase by 15% (The Environment Agency, 2008). Consequently increased flooding in the winter and drought in the summer is predicted (Githeko et al., 2000). Additionally, a rise in sea levels of up to 80cm by 2100 is also predicted, with greatest rises expected in the south of England (UKCIP, 2002). Parasite prevalence and diversity is closely correlated with stocking density (Altizer et al., 2003; Cote et al. 1995; Gillespie et al., 2005). This is due, in part, to the impact of host density and ensuing levels of social contact on parasite transmission; with the frequency of intraspecific contact between infected and uninfected individuals increasing as group size increases. As well as increasing contact with conspecifics, diminishing resources (e.g. water in summer) may result in the crowding of livestock with infective wildlife, increasing interspecies transmission potential. As the climate envelopes of wildlife shift, the levels of contact between them and livestock will again be affected. Livestock disease epidemiology is highly sensitive to host community composition and particularly population mixing (e.g. livestock moment through trade). Any change in these key epidemiological variables is likely to have knock-on consequences affecting the risk profile of parasites and pathogens transmitted via direct contact. Climate effects on indirect transmission: Exposure to parasites and pathogens via the faecal oral route is highly sensitive to grazing management practices (see above). The increases in forage availability predicted in section 7 may be utilised through changes in grazing management e.g. stocking density and as described above any change in grazing management practices will have associated disease risk costs and benefits. For example, extended growing seasons may result in extended livestock grazing seasons and thus increased exposure to environmental distributions of parasites an pathogens (including wildlife). However, such a change would result in less time in housed conditions where livestock are maintained on conserved forage. Conserved forage e.g. farm stored grains are often contaminated with wildlife excreta (e.g. rodent and bird faeces) that represent a risk of infection (Daniels et al. 2001). Livestock have limited behavioural ability to avoid ingestion of wildlife faecal contamination in farm

95

10.6

stored feed (Daniels et al. 2002). As such any reduction in feeding livestock farm stored feed stuffs would reduce the risk via this route of transmission. The above highlights the potential for any change in livestock husbandry to be associated with both increased and decreased disease risk. Livestock diseases are not evenly distributed throughout the UK and can be distributed in highly localised regions. Whilst climate change is likely to alter the distribution of disease risk it is unlikely to remove this heterogeneous spatial distribution. When faced with multiple disease threats each livestock holding should consider their often unique set of disease risks. For example, if the threat of vector borne disease increases in a region of the UK, farmers may bring animals indoors at times of high risk. However, the benefits of this action must be offset by its potential costs e.g. enforced confinement could increase the transmission of contact or airborne pathogens and stress associated high stocking density (especially in environmental extremes). As our understanding of the effects of climate change on disease risk increases, knowledge exchange and information flow to the livestock industries will form an important component of disease control.

Conclusions Given the damaging potential in some animal diseases, the adaptation imperative is to determine a clear strategy on what diseases to prevent rather than cure. For some endemic diseases this distinction is moot and the emphasis is on how current controls are exacerbated by changing climatic conditions. For non endemic diseases, a distinction between anticipatory (ex ante) and reactive (ex post) responses needs to be informed by an assessment of which diseases are worth stopping - i.e some disease risks will be exacerbated by climate change, but their prevention may be economically not worthwhile. While surveillance is likely to be the strategy in both cases, recent experience suggests that ex post responses can be very costly involving containment and wider economic and social disruption to the industry and in the countryside). There is therefore a premium on improving anticipatory capacity as a climate change adaptation. Given the current uncertainties in predicting which and when exotic pathogens will reach the UK the costs of such outbreaks are best predicted from previous experience. For example the UK has recently experienced two main threats to the livestock industry from highly infectious diseases that may be used to set bounds on the costs of control. Where disease eradication is required (for trade or health reasons) foot and mouth disease may be used as an estimate of the costs of eradication. Where eradication is not feasible, the vaccination campaign used to control bluetongue may be used as an example. For many infectious diseases a key bottleneck in predicting the future risk to the UK driven by climate change is a lack of long term basic epidemiological data. Modelling the effects of climate change on biological systems irrespective of whether it is climate matching models or process based models requires good data for model building and testing. Without data there is limited to no ability to test the predictive power of mathematical models. Given that we are at an early stage in quantifying the impacts of climate change on the livestock industries, we also need to consider early warning systems for disease outbreaks in the UK. For example the use of fast throughput technologies e.g. arrays to test large numbers of animal samples for multiple pathogens.

96

Statistical modelling (e.g. climate envelope modelling) offers the greatest potential for immediate return in terms of predicting changes in the disease risk posed to UK livestock by climate change. However, to improve our ability to predict changes in the amounts and distributions of infection and thus the associated economic losses to the industry, we must not disregard the complex interaction of intrinsic and extrinsic factors that drive disease transmission in livestock production systems. Process based simulation modelling (e.g. Agent-Based modelling) offers potential insights in to this complexity and ultimately when developed in association with data it offers the potential for improved predictive ability. Such models can be developed to incorporate the explicit role of human agency (i.e. farmer behaviour) in disease control and spread. In this regard we recognise private and public good costs associated with disease outbreaks and prevalence. Private human behaviours (i.e. management choices) are implicated in the external costs of livestock disease transmission. Increased surveillance as an adaptation does not obviate strategy of changing behavioural incentives through potential cost-sharing arrangements. Climate change adds a further rationale for behavioural change and it is important that its inevitability is emphasised as part of the cost sharing policy message.

97

11 Estimating the cost of impacts Selected climate change impacts affecting livestock production in the UK have been described and quantified in previous sections. In this section we draw together some of the quantitative evidence on impact costs, with a view to determining the relative significance of likely damages in the sector. The Climate Change Act (2008) sets out the government’s responsibilities and a commitment to undertake a periodic climate change risk assessment (CCRA), part of which entails an attempt to map out and value impacts as a prerequisite to deciding on adaptation spending priorities. In a scoping review for the CCRA, Watkiss (2009) identified several gaps in the integration of impacts and value information in agriculture. Relative to crops, livestock was notable for the lack of quantitative information on impacts and their costs. This exercise is therefore a first attempt to integrate impact and valuation information as a basis of deciding whether livestock impacts are significant and where these impacts fit in (agricultural) sector and national priorities. Assessments of impact costs usually assume that no private adaptation will take place (the ‘dumb farmer’ approach). This necessarily results in an over-estimation of the costs. However it is important to understand this value, which is essentially an upper-bound of possible impact costs, in order to highlight the consequences of not adapting to changes. However in reality, farmers will adapt to some extent, and the costs are unlikely to reach the magnitude suggested in most estimates of impacts. We have selected impacts which have an identifiable cost over a time horizon to 2080. We have not covered all cost categories. For example, we exclude the impacts of extreme events like flooding. Therefore it is not possible to provide an estimate of the total cost of climate change to the livestock sector in the UK. In addition, some of the impacts are likely to be mutually exclusive. At this point we do not account for interaction of impacts and therefore the cost categories are not strictly additive. The analysis distinguishes between private and public costs. Costs may be private (financial) or public (social). Many of the costs (or benefits) identified in this report are private; i.e. they accrue to the individual farmer in the form of change in production. Some of the costs have an external cost, such as mortality arising from heat stress, disease spread or the increased emissions of pollutants to air and water. Private costs are typically valued using market prices to determine losses to producers. We typically assume that these costs are internalised and that there is an incentive to undertake an appropriate adaptation to reduce these costs. Social costs include external impacts beyond the original impact. For example, disease transmission from one farm to another leads to a private cost being incurred beyond the original source of the contagion. Again, such social costs can be quantified with reference to market prices of the damage (i.e. a sick animal). Other social costs are much more problematic (though not impossible) to quantify because they cannot be related to a market price. Animal welfare is an example of a non-market impact that nevertheless has a social value. Only part of this impact will ever show up in actual behaviours such as purchases of welfare-friendly goods, but this does not obviate the existence of preferences for avoidance of poor welfare states. If consumers are notionally willing to pay to avoid such impacts, they can be quantified. Indeed, a range of existing

98

e rojects

private, inefficient animals are typically ore polluting in terms of waste generation.

1.1 Costs from each impact category

ection and magnitude. Table 11.1 also dicates whether the cost is private or social.

willingness to pay (or stated preference) estimates for welfare are available for crude benefits transfer if necessary. Another form of convenient value is facilitated by the shadow price of carbon, which is essentially a damage cost estimate for the external costs of an extra tonne of greenhouse gas emissions expressed in pounds per tonne of carbon equivalents (Price et al 2007). The valuation of other external costs to air and land are not as straightforward. For example, eutrophication and acidification impacts are often highly site specific and hindered by the absence of reliable data on the value of environmental costs at the point of impact. A similar caveat applies to the valuation of health impacts from particulates. In these cases, an approximation of impact values can be made with reference to existing willingness to pay or cost of illness data9. However, the application of benefits transfer to these impact categories is beyond the scope of thesp There are obvious synergies between the ways in which private costs and adaptations can affect external costs and, in addition to the absolute value of damage costs this has a bearing on which adaptation s should be prioritised. As an example disease surveillance is likely to deliver high benefit cost ratios. For a given risk of incursion and spread, several diseases can lead to high damage costs relative to the cost of surveillance expenditures that can reduce risks. The same expenditures can also be successful for detecting other endemic disease problems that affect animal efficiency. While these productivity issues may well be m

1 Table 11.1 summarises the specific impact costs. Only a selection of impacts have been costed as (a) a market does not exist for many of the impacts so estimating a cost would be complex and beyond the scope of this report and (b) as discussed previously, we are not trying to estimate total costs to the livestock sector, we are more interested in providing a picture of the likely dirin

9 As an example, there is some potential to link diffuse pollution impacts and their change

through adaptation to data on national water quality developed as part of the Water Framework Directive implementation process. These data show diffuse water quality from agricutre as a significant cause of quality standard failures, which are in turn valued using non market estimates

99

Table 11.1 Impacts costed from each chapter Chapter Impact Cost Private/social Impacts on grazing systems

Grass:silage forage ratio

Total cost of forage to 2080; discounted to present value

Private

Pollution/waste Net farm income % change from base

Private

Eutrophication Mitigation or clean up or willingness to pay

Social

Change in greenhouse emissions

Shadow damage costs

Social

Acidification Mitigation or cost of illness or willingness to pay

Social

Welfare (heat stress) Mortality Cost of mortality due to heat stress to 2080; discounted to present value

Private and social

Production loss Cost of production loss due to heat stress to 2080; discounted to present value

Private

Dairy days open Cost of a change in dairy cows days open10 due to heat stress to 2080; discounted to present value

Private

Disease/health Animal morbidity/& loss of productivity mortality

Increased surveillance and treatments costs Vaccination

Private and social

11.2

Impacts on grazing systems The section on grazing systems identified and quantified several potential impacts of a changing climate on grazing systems. For the purposes of this report we have taken the changes to the ratio of grass to other forage feed (mostly silage), and estimated the financial impact of this out to 2080. Other effects identified in the section, such as the time of first cut, do not have an obvious financial impact, and other effects such as changes to the stocking rate are complex in that they will be affected by many other factors, particularly policy and land availability.

10 Days available for conception

100

We therefore compare the difference between the ratio of grass to silage in 2008, with the projected changes in 2020, 2050, and 2080. For each of the three sectors looked at (dairy, beef and sheep), a representative system identified in SAC Farm Management Handbook (2007) is used to determine current volumes of grass and silage per animal, and current prices. The changes in total costs (grass + silage) per cow for the dairy sector over the four points in time, by region are shown in Table 11.2. Costs are not discounted at this point. Costs per animal in the beef and sheep sectors are presented in Appendix 6. Table 11.2 Total forage cost per cow over four points in time under climate change £/cow 2008 2020 2050 2080Scotland 94.61 92.80 93.32 92.80North East 94.09 92.80 93.32 93.32North West 94.61 92.80 92.80 91.25Yorkshire and Humberside 93.32 92.80 93.32 93.32East Midlands 93.84 93.32 92.80 91.25West Midlands 93.32 92.80 91.25 90.99East of England 93.06 92.28 92.80 92.28Wales 94.09 92.28 92.80 92.28South West 91.25 91.25 91.25 90.99South East 92.80 91.77 91.25 90.99 The table illustrates that forage costs in most region decline over this time period, as a result of the increased grazing availability (grazing is cheaper than silage). The changes are not particularly large however, ranging from no change to a reduction in cost per cow of 3.55% by 2080 in the North West. The impact per cow is then multiplied by the numbers of animals in each region, to estimate the financial impact per region. Although projections of animal numbers exist out to 2020 (BAU3) (Defra SFF0601)11, these are not broken down by region. Therefore numbers from the June Census (Defra) for 2007 are used and these are assumed to remain constant over all time periods (in the absence of climate change). While this assumption may be unrealistic, any projections of animal numbers beyond a few years are also likely to be uncertain. We assume a linear trend between the four points in time (which is necessary in order to obtain a cost for each year). These costs are then discounted to their present value using a discount rate of 3.5% for the first 30 years and 3% for the remaining 42 years.

11 Baseline Projections for Agriculture and implications for emissions to air and water

101

11.2.1 Dairy The discounted total grazing and forage costs to the dairy sector under the climate change scenario is shown in Table 11.3. The costs in each year are cumulative. To enable comparison between a scenario in which no climate change occurred, the cumulative discounted costs assuming the current ratio between grass and silage continued out to 2080 are also presented (as a total, not by region). The comparison between these two totals indicates that a changing climate may actually lead to a saving in forage costs to the dairy sector of £42m by 2080. This is however less than a one percent change from base situation. Additionally, these savings may be offset by the impact of extreme events that may adversely affect grazing systems. Table 11.3 Discounted cumulative regional dairy grazing and forage costs under climate change £million 2008 2020 2050 2080Scotland 18.58 196.36 430.76 534.28North East 1.92 20.30 44.54 55.25North West 33.76 356.83 782.78 971.38Yorkshire and Humberside 11.49 122.24 268.86 333.10East Midlands 10.14 107.61 236.65 293.78West Midlands 20.40 217.57 478.66 592.70East of England 2.57 27.39 60.04 74.30Wales 26.21 277.00 607.63 753.66South West 50.20 535.34 1179.44 1463.41South East 8.94 94.83 207.94 257.66Total 184.22 1955.47 4297.31 5329.52 Total - no CC 184.22 1964.37 4327.80 5371.15 Difference 0.00 8.90 30.48 41.64 % difference 0 0.45 0.70 0.78

11.2.2 Beef The discounted total grazing and forage costs to the beef sector under the climate change scenario is shown in Table 11.4. Again, the cumulative discounted costs assuming the current ratio between grass and silage continued out to 2080 are also presented (as a total, not by region). The comparison between these two totals indicates that a changing climate may actually lead to a saving in forage costs to the beef sector of £5m by 2080, a saving of less than one percent in comparison to current grazing:silage ratios.

102

Table 11.4 Discounted cumulative regional beef grazing and forage costs under climate change £million 2008 2020 2050 2080 Scotland 30.14 322.00 715.81 891.49North East 4.91 51.08 110.94 137.41North West 12.49 133.49 295.79 367.72Yorkshire and Humberside 5.32 56.75 124.15 153.50East Midlands 4.72 49.94 109.63 136.00West Midlands 5.60 59.49 130.32 161.03East of England 2.56 27.18 59.27 73.33Wales 13.15 138.34 302.62 375.06South West 12.19 130.01 286.45 355.65South East 4.59 48.98 107.61 133.13 Total 95.67 1017.27 2242.59 2784.32No CC 95.67 1020.17 2247.58 2789.44 difference 0.00 2.90 4.99 5.11 % change 0.00 0.28 0.22 0.18

11.2.3 Sheep The discounted total grazing and forage costs to the sheep sector under the climate change scenario is shown in Table 11.5. The costs in each year are cumulative. Again, the cumulative discounted costs assuming the current ratio between grass and silage continued out to 2080 are also presented (as a total, not by region). The comparison between these two totals indicates that a changing climate may actually lead to a saving in forage costs to the sheep sector of £90m by 2080, which is a saving of almost three percent in comparison to current grazing: silage ratios.

103

Table 11.5 Discounted cumulative regional sheep grazing and forage costs under climate change £million 2008 2020 2050 2080Scotland 33.71 355.30 771.70 946.24North East 5.48 55.96 119.28 146.82North West 8.62 91.30 197.49 242.31Yorkshire and Humberside 5.71 60.86 133.58 165.20East Midlands 2.56 27.57 60.95 75.68West Midlands 6.19 65.09 142.10 176.09East of England 0.94 9.90 21.53 26.65Wales 34.81 371.20 811.45 1001.22South West 8.89 92.77 201.57 249.25South East 3.48 36.86 80.19 98.99 Total 110.40 1166.82 2539.83 3128.45Total – no CC 110.40 1177.20 2593.56 3218.82 Difference 0.00 10.38 53.73 90.37 % difference 0.00 0.88 2.07 2.81 Clearly the saving in forage costs resulting from a changing climate is greatest in the sheep sector, however a three percent saving over 70 years is not particularly significant. However, it does highlight that the impacts of climate change, particularly in the UK, are not all likely to be negative.

11.3 Impacts from environmental pollution losses from grass-based systems Section 8 describes the methodology and presents results from simulating the impact of climate change on environmental pollution losses from grass-based systems. This impact has the potential to affect mitigation efforts from the agricultural sector, through the change in the GHGs methane and nitrous oxide projected under climatic changes. As discussed in section 8, the change in GHG emissions varies between the two gases and by animal type. Both greenhouse gases and eutrophication are social costs and are currently an externality of agricultural production. If agriculture were included in an emissions trading scheme then the change in emissions becomes a private cost (i.e. a permit price) to be borne by the producer. Alternatively the social cost of GHG emissions can be calculated using the social price of carbon (SPC). Either way, section 8 suggested that on balance a reduction in GHG can be expected. There was a suggestion that eutrophication potential could worsen, though we do not undertake an extensive valuation of these effects due to data limitations in linking the location of increased loading and receiving waters.

104

Section 8 also calculates the change in net farm income as a result of these changes to grassland systems. This is presented again in Table 11.6, which shows a mixed picture of gains and losses. Table 11.6 Projections of net farm income in dairy systems EE EM NE NW SC SE SW WA WM YH Scenario Net farm income (£) current 51139 48875 49376 49147 48506 52111 52709 48443 49733 491632020 3% 1% 2% 2% 1% 2% -1% 1% 5% 3% 2050 3% 4% 3% 2% 1% 1% 2% 5% 5% 5% 2080 4% 7% -5% 6% 6% 2% 2% 8% 6% 7%

11.4 Heat stress As detailed in section 9, heat stress can have a negative effect on a range of traits that affect the economics of the system, from losses in dry matter intake and milk yield, as well as affecting reproduction and ultimately, mortality. The model described in Section 9 (St Pierre et al., 2003) produces information on the impacts of heat stress on a number of biological traits for dairy cows, beef cattle, pigs and poultry, for the time periods 2020, 2050 and 2080 (and base period), at the UKCIP 50km resolution grid for each animal species by grid square. The RDP regions are overlaid on the grid and a representative grid square is used for each region (Appendix 6). Animal numbers are again used from the Defra June Survey. As in the grazing model, this model produces values for 2020, 2050 and 2080 as points in time. In order to calculate a present value of this impact, effects in the intermediate years are required. These are calculated assuming a linear path between the modelled estimates. Prices are discounted using a rate of 3.5% over 30 years and 3% beyond 30 years. Commodity prices used to calculate costs are documented in Appendix 6 Results are presented by animal type below. In all cases the increase is above a business-as-usual (BAU) scenario, which assumes no climate change and zero impacts for these categories. Costs are presented in £ million unless otherwise stated. For some animal categories, there were no impacts projected, and these figures are omitted. Note that rows (regions) have been removed from a table where there are no impacts across the three time periods As discussed earlier, mortality from heat stress represents a social cost as well as merely a private financial cost. That is, our financial losses estimated here should ideally be augmented by some understanding of public preferences (or willingness to pay) for the avoidance of adverse animal welfare outcomes. We do not derive these figures here.

105

11.4.1 Dairy cows Table 11.7 presents the present value (i.e. discounted sum of costs up to that year) of the estimated increase in dairy cow deaths by region, for the three time periods of interest. The largest costs are seen in the South West of the country. The South West experiences some of the largest temperature increases (but not as large as the South East) combined with having large numbers of dairy cows. While the South East experiences higher changes in climate, there are less dairy cows in that region so the economic damage is less. The South East is the only region to experience any losses by 2020. By 2080 all regions are expected to bear some losses, with the exception of Scotland. Dairy heifers are also vulnerable to heat stress. Table 11.7 presents the projected present value for mortality from heat stress for the three time periods. The South East is the only region to experience any losses in 2050, but by 2080 only the northern regions are exempt from impacts. The South West is expected to experience the largest economic losses, partly due to the large numbers of dairy cows in that region, combined with relatively large increases in temperature. Table 11.7 Present value of mortality, reproductive and production losses due to heat stress in dairy animals (cows and heifers)

Region Year PDeath - cows

(£’000s) Milk_loss(£’000s)

DO_loss(£’000s)

Pdeath – heifers (£’000s)

EE 2050 300 0 1.19 0 2080 760 40 5.62 10 EM 2050 990 10 2.09 0 2080 253 110 15.33 40 NE 2050 0 0 0 0 2080 130 0 0.48 0 NW 2050 0 0 0 0 2080 195 20 5.75 50 SE 2050 194 50 11.10 40 2080 407 310 38.90 90 SW 2050 422 50 14.84 0 2080 1,175 560 78.25 160 WA 2050 910 10 2.65 0 2080 325 130 18.93 140 WM 2050 237 10 3.40 0 2080 561 200 28.77 60 YH 2050 0 0 0 0 2080 730 10 2.91 20 Total 2050 10,730 130 35.27 40 (£’000s) 2080 30,780 1,390 194.50 570

Heat stress can have an effect on the milk yield of dairy cows. The model produces an estimate of the loss in milk yield per cow under the climate change scenario for the three time periods. This loss per cow was multiplied by the number of cows in each region to obtain the total loss in milk yield per region. The average price of milk per litre for the previous five years (2003 – 2007) was used to calculate the value of this loss, and discounted to the present value. The present value of the loss in milk yield per regions in shown in Table 11.7. The South West again suffers the greatest losses,

106

a combination of relatively large climatic changes together with a large number of dairy cows. None of the regions are expected to experience any losses in milk yield by the 2020s, and the northern regions of the UK are not affected in any of the scenarios. Heat stress can also affect reproduction. One indicator of reproduction is extra days to conception. Table 11.7 shows the extra days for cows to conceive over the different time horizons (DO). While the days to conception are not affected until 2050, by 2080 the number of days increases to almost six extra days in the South East. This clearly has economic implications. Esselmont et al (2001) estimate that for cows with typical yield the net cost after all factors are considered is £1.73 per day of delay to conception. This cost rises depending on the length of delay but for the purposes of this report £1.73/cow/day is the value that will be used. Table 11.7 shows the present value of the increase in days to conception (note that values are in £ rather than £million).

11.4.2 Beef cows Older beef cows are less affected by heat stress than dairy cows. Table 11.11 shows the present value of mortality of cows over two years as a result of heat stress. Appendix 6 shows the regional breakdown of this, where only the East of England and the South East of the country is affected. The relative resilience of beef cows to heat stress is, in part due to the lower metabolic load that this class of animal has compared to, say, a dairy cow as they are just growing. Beef finishing cattle are more affected by heat stress and suffer greater losses from mortality than the older cattle. This is due both to the more intensive production systems involved in finishing cattle, and the energy involved in the growing and finishing phase of the animal’s development. Table 11.7 presents the damages from heat stress for beef finishing cattle. Heat stress also affects the bodyweight gain of finishing cattle. This means that an animal will either take longer to reach a target weight, or will be sold at a lower weight for a lower price. The economic cost of the loss of bodyweight gain is shown in Table 11.8. Once again the impacts do not begin to show until 2080 but by then from the Midlands south, all regions are affected. The total (discounted) cost is only around £311k however, which is small in comparison to the losses arising from mortality. Note that costs are in £ not £million. Table 11.8 Present value of beef finishing loss in body weight gain

(£) 2020 2050 2080North West 0 0 1,274Yorkshire and Humberside 0 0 1,829East Midlands 0 0 83,112West Midlands 0 0 103,735East of England 0 337 57,224Wales 0 0 378,387South West 0 0 432,000South East 0 3,920 235,729Total 0 4,256 311,079

107

11.4.3 Pigs and Poultry Pigs are not expected to suffer losses from mortality due to heat stress until the 2080s. The East of England is expected to experience the largest economic losses, largely due to the number of pigs kept in that region. Table 11.11 summarises the projected present value of the damages to sows while the regional breakdown is shown in Appendix 6. Grower/Finisher pigs are more affected by heat stress than sows, possibly because of the nature of the production systems and their metabolic turn-over. Table 11.11 also presents the present value of mortality losses due to heat stress. The regional breakdown in Appendix 6 shows losses in the East and South East by 2050, but other regions not expecting any impacts until 2080. The North East and Scotland are not expected to experience any losses over these time periods. Like beef finishing cattle, grower finisher pigs are also likely to experience a loss in body weight gain as a result of heat stress. Table 11.9 shows the costs associated with this loss for the three time periods. Again it is only the East and the South East that are expected to experience any costs in 2050. By 2080 however, the total discounted cost over the country is over £3 million, which is relatively serious, and more than the cost from mortality. Table 11.9 Present value of body weight gain loss in grower/finishers (£million) 2020 2050 2080Yorkshire and Humberside 0 0.00 0.02East Midlands 0 0.00 0.41West Midlands 0 0.00 0.22East of England 0 0.02 1.57Wales 0 0.00 0.08South West 0 0.00 0.82South East 0 0.03 0.80Total 0 0.04 3.93 Table 11.11 illustrates the present value of mortality in laying hens as a result of heat stress for the three time periods. The regional disaggregation is shown in Appendix 6 and illustrates that the South East begins to experience losses by 2020, and by 2050 most regions south of the Midlands are experiencing heat related mortality losses. By 2080 the only region not to experience costs resulting from heat related mortality is Scotland. Heat stress can also result in a reduction in the amount of eggs laid. Table 11.10 presents the present value of the loss in egg production in the three time periods. While most regions experience a loss by the 2080s, the value of this loss is relatively small (in total only around £140k in 2080). It is possible that a loss of this magnitude would simply be accepted by the producers, rather than investing in adaptation. Values are presented in £ rather than £million.

108

Table 11.10 Present value of egg loss (£) 2020 2050 2080North East 0 0 23North West 0 0 1257Yorkshire and Humberside 0 0 1944East Midlands 0 4604 30650West Midlands 0 1649 12216East of England 0 3449 15714Wales 0 448 2998South West 0 5305 26740South East 55 14612 50259Total 55 30068 141800

11.5 Total losses from mortality Regional costs arising from mortality are presented for the other animal categories in Appendix 6. Table 11.11 presents the total present values from all animal categories resulting from animal mortality. This illustrates that by 2020 the costs to the industry are only around £200k, which is likely to be able to be absorbed by the industry. By 2050 however, the costs have reached over £11m, and by 2080 the costs of mortality alone have risen to over £34m. This is a more serious cost to the agricultural sector, and in addition does not include losses other than mortality, which are discussed earlier. The dairy sector is expected to experience the greatest losses, a combination of the location of much of present dairy farming being in areas projected to experience greater levels of change, together with the value of each dairy cow, which is significant. This indicates that the dairy sector is vulnerable to gradual warming. In addition, mortality represents a social cost beyond the purely financial cost because of the implications for animal welfare. Table 11.11 National level present value of costs arising from mortality from major animal categories (£million) 2020 2050 2080Dairy - cows 0.19 10.73 30.78Dairy - heifers 0.00 0.04 0.57Beef – over two years 0.00 0.00 0.24Beef - finishing 0.00 0.12 1.52Pigs - sows 0.00 0.00 0.11Pigs - grower/finishers 0.00 0.03 0.27Poultry - layers 0.01 0.21 0.54Total 0.20 11.13 34.03 Costs as a whole begin to increase by the 2050s and by the 2080s become significant (over £34 million). It should be noted that biological modelling of heat stress, and follow-on costs of heat stress, refer to the gradual warming. However, an extreme event, such as prolonged heat wave in a given year, is likely to have a more dramatic effect on biological traits described and therefore costs to the industry would increase dramatically for that given year. There are likely to be other costs incurred from other impacts such as other extreme events (droughts and flooding) and increases in disease incidence. On the other hand there may be benefits resulting from a longer growing season and milder winters, as discussed previously in section 7.

109

11.6

The choice of discount rate will also affect the damage costs. The rates used here are recommended by the Treasury for public sector decisions, however they are relatively high compared to the discount rate used in some climate change assessments (notably the Stern Review). A lower discount rate would result in higher present damage costs. These are the costs that would be incurred without any adaptation/intervention (the ‘dumb farmer’ assumption), such as a change in the distribution of livestock systems across the country and/or the adoption of management changes to help circumvent heat stress. However, farmers are likely to take action to avoid these impacts, and so the cost is likely to be much lower (see Figure 5.1 in section 5). A final point is that over the time horizon considered here, there are many possible changes that are likely to have far greater economic impacts on the livestock sector than climate change. Changes in agricultural policy, international trading regimes, social and political factors and consumer preferences are some of many possible influences on agriculture.

Losses from transport Section 9 outlined challenges in calculating the effect climate change will have on transport in each animal sector i.e. for each species of livestock. Very little information (limited data – scientific, statistical or economic) is currently available to determine the actual costs to the industry of the effects of transport mortalities, compromised welfare, product quality changes, and compliance with existing legislation. In the Defra (CC0361 (2008)) evaluation of the economic effects of the impacts and adaptations that might be related to animal transportation were extremely limited due the lack of availability of “suitable data). An increase in mortality during transport of 10% as estimated in section 11 could impose serious costs on the livestock sector.

11.7 Conclusion It should be emphasised the even under the worst case climate scenarios, the impacts affecting the UK are unlikely to be so significant that agriculture will be severely disturbed. The projections, even for the 2080s, suggest the climate of the UK may become analogous to current climates in major agricultural producing areas in Europe. A greater climate-related impact relevant to the UK, unfortunately beyond the scope of this report, is likely to be the effect on world markets of impacts of climate change in other major agricultural producing areas in the world.

110

12.1

12 Identification of adaptation options and attitudes Based on the relatively benign impacts projected for the UK, farmers are likely to adapt autonomously to the changes. The role for government is likely to be in building the adaptive capacity, for example to improve predictive modelling capability to understand impacts through time as well as to provide farmers with adequate information regarding impacts and adaptation options. Intervention may also be required to integrate climate change with existing policy, e.g. animal housing and transport regulations, surveillance programmes for livestock disease, and wider CAP objectives. As previously noted “adaptation” can have different objectives and may also be variously interpreted by different stakeholders. For some it may be to maintain or preserve existing practices, breeds, land, market share. For others it may be to accept that changes will occur presenting opportunities as well as threats. It is important to understand how these varying interpretations and expectations are actually aligned, such that policy complements rather than displaces autonomous action. Accordingly, this section identifies likely adaptation options for specific climate change impacts, and indicates whether these are management, technical, or infrastructural measures. These options are possible in theory, but there may be particular reasons why they may not be feasible for certain areas or production types in the UK, or there may be barriers to their uptake. The project used two survey methods to narrow the possible options down to a shortlist based on stakeholder workshops held as part of this project.

Possible adaptation options Section 9 detailed the potential impacts of a changing climate on livestock, focussing on welfare of the animals, and to a lesser extent animal productivity. The following section highlights some of the farming system and transport adaptations that could be put in place to help farmers reduce these impacts.

12.1.1 Potential adaptation to climate change in dairy systems There are a number of options that exist to help minimise the effect of thermal stress on dairy cattle, including altering the environment the cow lives in. In the US, farmers reduce the effect of summer temperatures on lactating dairy cows by housing them throughout the summer. They may also install fans or evaporative cooling systems (spray fine mist onto cows) in the sheds. In addition to warmer summers, extreme weather events may mean that cows require availability of ad hoc housing, with the necessary support (e.g., availability to concentrates, adequate cooling/air flow) in place to deal with that housing period. The costs associated with adaptation for dairy cattle and their systems will be mainly due the housing requirements. Many UK dairy farmers already house animals for some period, but in the future these periods may need to increase to avoid detrimental effects of heat and cold stress. Also, buildings may need to be modified to ensure that sufficient cooling is in place to deal with warmer summers, particularly heat wave risks. This will require capital investment for some dairy farmers in building development/maintenance as well as cooling equipment installation (and has

111

implications for mitigation efforts). Also, during periods of housing animals will require concentrate feed and silage and extra bedding.

12.1.2 Potential adaptation to climate change in beef systems As with dairy systems, in very hot weather beef producers may decide to house animals to help keep them cool. As beef animals have a higher upper critical temperature and are better able to cope with warmer conditions it may not be necessary to house animals for as long as a lactating dairy cow. Also, additional mechanical air flow aids may not be necessary, once the building is sufficiently well designed to take account of air flow. During the winter months there may be periods where out-wintering may not be appropriate. Animals may well have to be housed during such periods. Costs associated with these types of interventions are similar to those for dairy systems.

12.1.3 Potential adaptation to climate change in sheep systems Generally sheep in the UK are extensively managed and potential adaptations to climate change may be limited by this unless management scenarios change. A list of potential adaptations for sheep systems include: • Better provision of shade and shelter to allow behavioural adaptation • Reactive housing of vulnerable stock if extreme weather forecast at certain times of

year. Also improved access to feed and check stock if wet weather/flooding risk increases and current access cannot be maintained.

• Shearing sheep earlier to avoid/limit heat stress in warmer summers. • Disease control (Section 10), especially lameness - increase lameness in sectors,

e.g. hills, where traditionally lameness is very low • Breed substitution to limit losses and/or to take advantage of possibility of improved

productivity.

12.1.4 Potential adaptation to climate change in pig and poultry systems For indoor pigs and poultry, the main climate change challenge is coping with warmer summers. Temperature spikes may present the main largest challenge for animals and therefore welfare, particularly when systems are inadequate and animals are less able to cope with sudden changes. Producers could respond by investing in new ventilation and cooling systems (e.g. water sprays) to be fitted to existing buildings (Turnpenny et al. 2001), or when replacing buildings, to employ designs more similar to those used in other countries. Particularly for indoor livestock, milder winters could have a favourable effect in that there is less need to heat buildings, particularly in farrowing buildings, and therefore reduce winter costs. This, however, may not offset the additional costs associated with mechanical intervention in the summer months. Another potential adaptation is breeding indoor (and outdoor) animals to be able to deal with heat stress. For example, white hens have a lower mortality (4-8%) during periods of heat stress that brown hens. For outdoor pigs, piglet mortality due to chilling in colder and/or wetter months could be reduced with producers ensuring that plenty of dry straw is provided, and that the farrowing ark provides adequate shelter from rain. Outdoor pigs may also suffer due to

112

increased summer temperatures. Producers can respond to this challenge by: providing wallows, shade (or ideally shaded wallows), water sprays, appropriate insulation and (passive) ventilation of farrowing huts. A more radical option would be to use farrowing huts designed for hot weather, such as those designed for use in Italy (Barbari & Ferrari 2001). When ground is wetter in outdoor pig production it may be necessary to reduce stocking density to minimise the health implications for the animals but also to offset the environmental damage that may occur.

12.1.5 Potential adaptation to climate change for livestock transportation In the face of climate change induced alterations in mean temperature and the frequency of extreme events the consequence of potential thermal stress due to the corresponding vehicle micro-environments will require a wide range of adaptations or strategies for the amelioration of the effects upon animal welfare in transit and production efficiency. These adaptations may be implemented by breeders, producers, hauliers, processors and policy makers and legislators. In Table 9.7 a number of possible adaptations were indicated relating primarily to engineering solutions and modifications to legislation that might limit exposure times to elevated or reduced temperatures and/or that might improve ventilation of vehicles and containers thus reducing the risk of heat stress when either average temperature is increased the frequency of episodes of elevated temperature is increased. It is possible to minimise the effects of external climatic conditions by improvements in vehicle design and operation by utilizing engineering solutions to match “on-board” environmental conditions to the animals’ biological requirements. The outputs of models based upon these principles can provide sound scientific basis for improved vehicle design and operation but will also inform relevant legislation and codes of practice aimed at optimizing both animal welfare and productivity in relation to transportation of livestock on journeys of both long and short duration (Mitchell and Kettlewell 2008). The effectiveness and efficiency of the various strategies might be considered in relation to the potential impacts of the differing climate change scenarios on the various livestock species and journey types that may be deemed most vulnerable. Livestock may be transported at different ages and this is a factor that determines their vulnerability to thermal challenge in transit. The species that may merit special consideration in this context include pigs, poultry, cattle and sheep. Adult pigs, adult sheep, adult cattle (all for slaughter or for long distance export) and broiler chickens and turkeys at slaughter weight may be particularly susceptible to heat stress in transit whereas young animals (e.g. weaner pigs, piglets, calves, lambs and day old chicks) may be more affected by cold exposure and cold stress. Of course, in the case of extreme events or episodes animals of all ages are detrimentally affected by both high and low temperatures. The potential adaptations to livestock transportation to reduce the impacts of climate change for animals in transit that are most likely to be implemented are: • Improved vehicle design and operation

o Mechanical ventilation, showering (cool water), misting systems or air conditioning and continuous provision of cool drinking water (implemented by producers and hauliers in a pre-emptive and voluntary response)

• Changes in loading or stocking density or space allowances per animal by producers//hauliers/drivers

113

o Carriage of fewer animals per vehicle to reduce on-board heat and moisture production – major financial implications

• Re-scheduling of journeys – more night time transportation – requires rescheduling of processing plant operation for slaughter animals

• Reductions in the transport of live animals (particularly for slaughter) – introduction of more local slaughtering and the increased transport of carcases. Changes in the locations of breeding hubs and increasing the number of such centres to provide breeding animals that are transported shorter distances

• Changes in EU/UK legislation (welfare) relating to transportation of animals o Changes in regulations applying to vehicles design, operation, ventilation,

climate control and provision of cool drinking water o Changes in prescribed space allowances for each species, physiological

status and age of animal to allow for additional heat loads (see above) o Changes in legal travel times for animals, total journey durations and

journey breaks and stops in control posts and lairages (for each species physiological status and age of animal) – it is well established that the duration of exposure to excessive thermal loads has important effects upon mortality, welfare and productivity

o Changes in legislation relating to vehicle design and operation in particular ventilation regimes and control and on-board cooling systems

• Relocation of geographical sites of production and slaughter where the risk of thermal stress is less

• Genetic/Genomic selection of breeds and strain of livestock or use of existing breeds that are more thermo-tolerant and more resistant to the effects of heat stress (as may occur on animal production grounds alone). There is already evidence that slower growing lines of broiler chickens may exhibit some of these characteristics and may be more resistant to transport stress.

It is of interest that in a recent review report (Defra Project CC0361 (2008) – Changes in Agricultural Management Under Extreme Events – Likelihood of Effects and Opportunities Nationally) the authors found very little scientific literature or other reference sources that might inform strategy or policy relating to possible adaptations for implementation in animal transportation. The transportation of animals at night was recommended as was the use of vehicles with mechanical ventilation with a general call to reduce all animal stocking densities (production) during heat stress. Decreases in the stocking density in transit for both broiler chickens and lambs to reduce the risk of losses in transit were identified as possible strategies. However, the literature sources quoted for each of these adaptations were chiefly MAFF, Defra and NFU guidelines and advice documents. Table 12.1, adapted from AEA Energy and Environment (2007), summarises climate change impacts affecting agriculture, the consequences of these for agricultural production, and generic measures in response to the impacts. Adaptations are also categorised according to whether they are technical (T), management (M), infrastructural (I), or equipment (E) based.

114

Table 12.1 Climate change impacts and adaptation options for the livestock sector Climate change impact: Deterioration of conditions for livestock production (due to increased heat stress; new pests and diseases; change of optimal crop areas; wetter winters) Consequences for agricultural production

Adaptation option

Category

Changes in livestock health and productivity

Decline in number of native bred livestock and introduction of more heat tolerant breeds

T

Move herds from waterlogged fields M

Increase shelter for animals, including from heat

I

Change breeding and shearing patterns for sheep production

M

Supplemental feeding M Loss in forage quantity and quality

Balance of grazing and cutting M

Changes in grazing regime M Increased use of legumes M Change of seed mixture M Changing time of operations M Adjust stocking density M Change in livestock health and productivity, animal welfare issues resulting from heat stress

Improve ventilation in buildings I

Addition of cooling pads, fans systems, water sprays/misters to building and/or outdoor areas (e.g., collecting yards)

I

Bring animals indoors M Plant trees for shade T Ensure adequate access to water

(indoors and outdoors) to aid thermoregulation

T/I

Adjust timing of transport M Assess building and transport

regulations to accommodate new temperatures

I

Breeding for heat tolerance (against heat stress and/or breed substitution

M/T

Adjust breeding season to minimise exposure of vulnerable young animals

M

Adjust diet to ensure sufficient dealing with the hot weather (e.g.,

M

115

energy requirements are being met if heat is reducing total feed intake, overfeeding protein should be avoided as energy required for excretion, certain minerals required)

Manage health and disease implications of hot weather (e.g., fly strike, acidosis increases during heat stress)

M

Change in livestock health and productivity, animal welfare issues resulting from cold stress (wetter/windier winter/spring)

Improve shelter I/M

House animals M Timing of shearing M Adjust breeding season to minimise

exposure of vulnerable young animals

M

Assess building and transport regulations to accommodate new temperatures

I

Breed for “hardiness” and/or breed substitution

M/T

Climate change impact: Increased risk of drought and water scarcity (due to decreased annual or seasonal precipitation; increase in the frequency of extreme conditions) Consequences for agricultural production

Adaptation option Category

Shift pasture from drought-sensitive areas

M Conflicts among users

Set clear water use priorities M Increase water use efficiency M

Increase rainfall collection capacity M / I Improve field drainage and

absorption capacity I

Reduce run-off through contoured hedgerows and buffers

M

Introduce more drought-tolerant crops

M

Woodland planting M Use of precision agriculture

techniques T/I

Water management practices M

Water charging/tradeable permits M

Reduced water supply

Insurance I/M Climate change impact: Increased risk of agricultural pests, diseases and weeds (due to increased water logging; increased average temperature)

116

Consequences for agricultural production

Adaptation option Category

Use of pest-resistant varieties M Use of thermostats and rapid-

cooling to reduce pest and disease infestation

E

Develop sustainable integrated pesticides strategy

M

Use of natural predators M Vaccination M/ T

Pest populations and distributions increase

Monitoring of pests and disease patterns to prevent damages

M

Develop sustainable integrated pesticides strategy

M Pollution by increased use of pesticides

Advisory service for farmers M Climate change impact: Increased risk of floods (due to increase in extreme events frequency; loss of soil water retention capacity) Consequences for agricultural production

Adaptation option Category

Develop contingency plans M Create/restore wetlands M Enhance flood plain management M

Increased expenditure in emergency and remediation actions

Hard defences I Increase rainfall interception

capacity M/ I Flash flood frequency and

intensity increase Reduce grazing pressures to protect

against soil erosion from flash flooding

M

Increase rainfall interception capacity/soil management

M

Contour ploughing M Increase drainage I Addition of organic matter into clay

soils M

Flooding and storm damage increase

Insurance for farm infrastructure M Animal mortality and welfare issues

Provide adequate shelter I/M

117

Climate change impact: Pasture area changes due to change in optimal farming conditions (due to changes in monthly precipitation distribution; increased temperatures in critical periods; increased erosion; loss of soil water retention capacity) Consequences for agricultural production

Adaptation option Category

Livelihood diversification M Strengthen local adaptive capacity

to reduce sensitivity M

Conversion of ambient storage to refrigerated stores

E / T

Irrigation I Changing cultivation practices M Additional weed/pest control M Movement of crops to more

favourable areas M

Optimal conditions altered resulting in increased risk to rural incomes

Increase in crop breeding investment

T

Climate change resilient species T Loss of indigenous species Insurance M Extensification: enhance carbon

management and zero tillage – I get the zero tillage bit but what exactly is implied by ‘enhance carbon management’ as part of extensification?

M Soil deterioration due to land use changes

Precision agriculture M Intensify research efforts and

training T Land abandonment due to

very large changes in optimal conditions Livelihood diversification M Climate change impact: Increased irrigation requirements (due to increased average and extreme temperatures; increase of drought and heat stress frequency; decreased precipitation) Consequences for agricultural production

Adaptation option Category

Technical improvements in irrigation equipment and ability to collect rainwater

E/T/I

Trickle irrigation E / I Irrigation during the night M Separation of clean and dirty water E/ I

Water availability decrease Water shortage in irrigated areas

Installation of small-scale water reservoirs on farmland

I

Climate change impact: Sea-level rise Consequences for agricultural production

Adaptation option Category

Hard defences I Alternative drainage systems I Set-aside of land for buffer zones M Alternative crops M

Sea level intrusion in coastal agricultural areas and salination of water supply

Livelihood diversification M

118

12.2 Refined adaptation options Understanding the position of those faced with the prospect of adapting is important for identifying barriers and prioritising action. It is also important because assumptions need to be made regarding the likely future level of private adaptation for impact and public adaptation costings. At present the impact calculations assume farmers undertake no adaptation. This dumb farmer assumption is as unrealistic as assuming they will be ‘clairvoyant’ and adapt efficiently to all impacts. We are currently at the point of assuming an arbitrary level of adaptation; but understanding the likelihood of uptake of certain actions will make the assumption less arbitrary and more informed.

12.2.1 Workshop discussions Two industry workshops were undertaken in 2008 to explore attitudes and views on climate change scenarios and potential adaptation responses. The workshops used an open-ended discussion format to explore key issues on impacts, adaptation, behaviours and perceived responsibilities for planning. These discussions were wide-ranging with recognition of the potential impacts that could result from extreme climate scenarios. However, it proved difficult to concentrate attention on the formation of concrete adaptation strategies that were simply not regarded as being a priority in business plans. In essence, business time horizons are more short-term than the perceived time horizon for significant impacts. At the very least, producers felt they needed better information on potential damages relevant to them to allow them to consider the relevance of climate impacts. To the extent that any adaptation could be motivated at this point, then action would only be likely where adaptation action was coincidental with management options that improved productivity or that might deliver on mitigation obligations. Indeed there was much more recognition of the immediacy of mitigation obligations and how these could impinge on business costs. To some extent there was a view that this agenda distracted from the complications of having to navigate the intricacies of uncertain climate projections. In terms of responsibility there was an understandable tendency to suggest that longer term planning was a public (rather than private) responsibility with some emphasis put on the encouragement of improved breeding for improved resilience among species. Additional observations suggested that the age profile of producers is significant in attitudes and behaviour on climate change. This was possibly related to the experience of how market support policy had traditionally served as an industry buffer from adverse global conditions. The findings from these exercises were revealing to the extent to which they confirmed an a priori expectation that private enterprises are reluctant to contemplate investment plans without better information or downscaled climate forecasts. This indicates that future resources might focus on developing a capacity to translate national scenarios into storylines that are relevant to these groups. In methodological terms, the open-ended nature of discussions was perhaps not best suited to the identification of more definite (albeit hypothetical) adaptation options and

119

decisions. However, break-out groups were useful for providing a range of position statements on adaptation that could subsequently be a basis for further empirical survey analysis.

12.2.2 Further survey research on potential adaptation choices To further our understanding of adaptation attitudes, the project undertook two further exercises. The first was a survey exercise to elicit views using participants in a SAC annual Animal Health and Welfare conference (2008). The second exercise used the statements from the initial expert workshops in a Q survey that was later mailed to the same SAC conference survey participants who had indicated a willingness to participate in a further on-line survey. The initial survey was administered to fifty-two conference participants from a range of agriculture and veterinary related bodies. The majority were from within research (40%), with a smaller number being involved in commercial businesses (including vets and farmers) (20%) or public bodies (6%). Participants were presented with a questionnaire (see Appendix 7) comprising a mix of eight open-and closed ended questions eliciting opinions on the most likely adaptations to be applied now and by 2020, in response to a number of climate change-related conditions. 2020 was used a representative future year as it is far enough in the future to begin experiencing some impacts, but also not so far in the future that it becomes unimaginable. The survey considered four issues: heat stress, wetter winters, changing seasonality and changes in disease trends. A detailed description of the questions and responses is presented in Appendix 7. The three most commonly identified strategies for each of the four climate impacts in 2008 and 2020 are summarised in Table 12.2. This summary shows that the vast majority of these response categories fall under the adaptation category “management”. A smaller number fall within the “technical” category. Infrastructural related adaptations include water management, housing and drainage. None of the adaptations fall into the “equipment” category as mentioned in Table 12.1 (although it could be argued that some of the infrastructure categories could also be equipment).

120

Table 12.2 Three response categories with greatest number of comments, for each question Question Response category 1 Response category

2 Response category 3

What adaptations may be made now (2008) in response to heat stress?

More housing/shelter Ventilation Access to water / water management

What adaptations may be made by 2020 in response to heat stress?

Animal breeding Better housing Access to water

What adaptations may be made now (2008) in response to wetter winters?

More housing/ shelter Improve drainage Disease monitoring/ control

What adaptations may be made by 2020 in response to wetter winters?

Animal breeds and breeding

Water management / flood defences / flood management

Housing / shelter

What adaptations may be made now (2008) in response to changing seasonality?

NB: Four response categories with equal number of comments Animals out earlier / outwintering / in later Changes to cropping systems Other farm management changes Research

What adaptations may be made by 2020 in response to changing seasonality?

Changes to cropping systems

Research Changes to lambing / calving

What adaptations may be made now (2008) in response to changes in disease trends (arising due to climate change)?

More monitoring/ surveillance

Research Vaccination

What adaptations may be made by 2020 in response to changes in disease trends (arising due to climate change)?

Animal breeding Vaccination NB: Two response categories with equal number of comments Surveillance Changes in farm management

These adaptations span private and public roles. Some adaptation categories are likely to involve only land managers (e.g. relocation of stock), others will involve government (legislation), wider industry participation (biosecurity) and public and private research institutions (animal breeding). The roles and responsibilities of actors regarding adaptation actions are discussed further in Section 13. Likewise the categories include a mix of technical and managerial adaptations, and would need to be addressed at different scales.

121

12.2.3 Q-methodology A second survey methodology (Q methodology) was developed on the basis of the initial expert workshop break-out group exercise. A detailed description of the methodology is included in Appendix 7. Here we focus on the key results from this survey. In this application comments were collected from participants at a workshop held at SAC Edinburgh in 2008, in a series of facilitated break-out groups. Within these groups participants were asked to consider questions such as “What are your overall opinions about adaptation to climate change within the livestock industry?”; “What do you think are likely to be the main problems in implementing adaptations to climate change, within the livestock industry?”; “What strengths does the livestock industry have that will aid adaptation?”; “What additional information (if any) do you want/need with regard to climate change, the livestock industry and potential adaptations?”; “What adaptation strategies do you think are most likely to be implemented first (and when would this be?) (across the livestock industry or within the livestock sector with which you are most familiar?) Why?”; and, “What adaptation strategies do you think are least likely to be implemented within the livestock industry? Why? “. The three breakout groups yielded a total of 129 usable statements. The aim was to select a smaller number of these statements for use in a ranking exercise that forms the central part of the Q methodology. In all cases the statements needed to be something that a respondent could agree or disagree with – hence had to be an opinion statement, rather than a statement of fact. Statements were categorised under the headings: • Economics • Role of government • Information • Farm level • Industry level This categorisation enabled duplications and problematic statements to be removed, leaving a collection of 32 statements. The aim of this selection process was to ensure that, as far as possible, the diversity of opinion expressed in the break-out groups was maintained in the final collection of statements. A final group of 24 statements were identified, and then ranked by stakeholders against a seven point scale from agree to disagree. This process is known as Q sorting, and the ranking process follows a forced distribution of opinion that can be the basis for identifying more specific sub groups. Appendix 7 describes the methodological approach and data analysis in detail. The results from this study suggest four potential adaptation scenarios. The study reveals that there are stakeholders who support each position. Other views may exist but if they do, they represent additional positions not alternative ones. The four scenarios are outlined below: Scenario one: Financial support for farmers with freedom over decision-making The answer to climate change adaptation problems lies with livestock farmers who know best how to manage their own enterprises, but additional financial support should be provided as this is a broader social problem.

122

12.3

Scenario two: Regulation The answer to climate change adaptation problems lies with regulation and definitely not GM technology. Scenario three: Education/information provision The answer to climate change adaptation problems lies with education of and information provision for operators within the livestock industry Scenario four: Technology The answer to climate change adaptation problems lies with technology, specifically GM. These four scenarios illustrate that stakeholders hold quite differing views on the best way to adapt to climate change. From a broader policy perspective, they need not be mutually exclusive. Scenario 1 very strongly advocates freedom over decision-making, with financial support where necessary. While Scenario 2 advocates regulation used as a safety-net to ensure certain standards and safeguards are met beyond the autonomy provided by scenario 1. In reality a compromise between scenario 1 and 2 may be the most appropriate approach to ensuring effective adaptation without frustrating and perhaps disengaging farmers who may be adapting anyway. Education provision can go hand in hand with both Scenario 1 and 2, as can technology.

Conclusions This section has provided a clearer indication of attitudes among a limited sample of livestock industry, in addition to a summary of possible adaptations that may occur. The findings from each survey provide several insights about the attitudes towards adaptation from within the livestock industry. The first survey provides a useful shortlist of which adaptations are prioritised now and in 2020; the Q survey reveals four distinct positions as regards likely and preferred scenarios of climate change adaptation within the livestock industry. These results imply an expectation of a combination of private and public responses . The wider implications these adaptation options have with regard to uncertainty in impacts, their effects on mitigation activities and the role of the public and private sectors are discussed in section 13.

123

13 Assessment of key public responses The evidence gathered here suggests that autonomous adaptation will be gradual and that private agents are adept at assessing the economic viability of interventions aimed at maintaining their own productivity. From the available evidence, we anticipate that the external costs of this process will be small and that the management of these impacts is in any case likely to be overtaken by other policies that are specifically focussed on these endpoints. For example, increased emissions from livestock will be addressed by more aggressive mitigation incentives at the farm scale and an evolving trend towards the use of carbon benchmarking (including nitrogen) or even an instrument based on a carbon price. Similarly, diffuse pollution will in the interim be targeted by provisions under the Water Framework Directive, Nitrate Vulnerable Zones and a range of related guidance on management of waste. These changes should also be set against a trend of falling livestock numbers. What then is the public role for the sector and how to appraise efficient intervention? The appraisal of adaptation responses pre-supposes some indication of the magnitude of impacts and when responses might take place. Yet the residual impacts of heat stress and disease incursion are highly uncertain. Only the former can be potentially modelled using downscaled data accompanying extreme event information included in UKCIP09 and the impacts are arguably a shared responsibility. A straightforward adaptation response would involve a periodic review of regulatory standards for housing and transport. These standards can also be informed by a review of the impacts of conditions in countries with projected climate conditions. In the case of disease, while the data are uncertain, we can arguably base a worst case scenario on recent experience with epidemics. The total costs arising from the last major episode of Foot and Mouth outbreak in 2001 have been put at around £9 billion, with at least £3 billion in direct costs to the public sector and about £5 billion in costs to tourism and the rural economy. Out of the total costs incurred during the outbreak, compensation payments to farmers for the slaughter of their animals and welfare reasons were placed at £1.34 billion. These figures clearly dwarf any other costs we have been able to identify in this study. The Foot and Mouth episode raised several useful lessons; specifically, this and other recent crises have led to a clear government message about the need for a sharing of risk in disease management. The desire to share risk and responsibility is set out in the government's Animal Health and Welfare Strategy (Defra 2004), and is restated in Partners for Success (Defra 2005). In essence, if government acts as an underwriter of losses there is little incentive for appropriate behavioural change that can minimise disease risks. While government cannot monitor risky behaviours it can transmit a clear message in relation to the eventual sharing of costs in the event of future episodes. This message on ultimate liability for damages is what can lead to modified behaviour and the learning from this experience can be extended to encompass the increased risks related to climate change. In essence the cost sharing message inherent in Partners for Success simply needs to be reinforced by the message that climate related risk is part of the same joint-responsibility storyline. Thus the adaptation is actually already happening albeit in a way that is not entirely focussed on the part of the risk due to climate.

124

13.1

13.2

Only in the case of extreme weather events is there likely to be a significant deviation from the gradual adaptation pathway and there is some merit in exploring the expected damages from the risk scenarios

Placing livestock in context Where do these findings position livestock in the context of a notional national risk assessment? Downing and Watkiss (2005) advocate a sequential approach to prioritising adaptation actions, with actions becoming “harder” as the evidence base becomes clearer.

1. Building adaptive capacity. This involves research, awareness raising, testing adaptation actions based on current levels of knowledge. Ensuring that policies that facilitate adaptation do not create perverse incentives or trade-offs with other goals, such as mitigation, is also important.

2. Alter existing plans to manage climate risks and take advantage of new opportunities. This step would focus on urgent and high priority actions, as well as win-win solutions which reduce vulnerability to current climate variability. Ideally working within existing frameworks to minimise costs and disruption.

3. Implement adaptation actions. If possible, delay action in areas where levels of future benefit are difficult to ascertain. As the evidence evolves, other options can be considered, particularly the more “hard” technological developments.

Arguably actions for the livestock sector currently fit predominantly into category 1, with elements of category 2 also being relevant.

Analysis of shortlisted adaptation options In this section we take the options identified in Table 12.2 and assess them in terms of their robustness under uncertainty, based on the framework proposed by Hallegatte (2009) (discussed in section 6.2), their likely interaction with mitigation efforts, and whether they are likely to be undertaken by private actors, i.e. the farmers. The adaptations are again distinguished between those that were identified as being necessary now, and those that should be put in place in 2020. Table 13.1 summarises the assessment and the individual categories are discussed in more detail in the following subsections.

125

Table 13.1 Adaptations assessed as to their robustness against uncertainty, their effect on mitigation, and who would be adapting (++ = this adaptation always has the characteristic of the column it is in; + = in some cases the adaptation has this characteristic, but not all) Adaptation option

Reversible/ flexible

No-regret

Safety margins

Soft strategy

Reduced decision time horizons

Public/ Private

Effect on mitiga- tion

Heat Stress - now More housing/shelter

+ +* ++ Private +ve/-ve

Ventilation + +* ++ Private -ve Water management

+ ++ ++ + ++ Public/ Private

Heat Stress – 2020

Animal breeding ++ + Public +ve /-ve Better housing + +* ++ Private +ve/-ve Access to water + ++ + + Public/

Private

Wetter winters – now

More housing/ shelter

+ +* ++ Private +ve/-ve

Improve drainage + +* + Private +ve

Disease monitoring/control

++ ++ ++ ++ ++ Public

Wetter winters – 2020

Animal breeding ++ + Public +ve /-ve

Water and flood management/ defences

+ +* ++ Public/ Private

Housing/shelter + +* ++ Private +ve/-ve

Changing seasonality - now

Animals out earlier/ outwintering/ in later

++ ++ ++ ++ Private +ve /-ve

Changes to cropping systems

++ ++ ++ ++ Private

Farm management changes

++ ++ ++ ++ ++ Private +ve /-ve

Research + + + + + Public

Changing seasonality – 2020

Changes to cropping systems

++ ++ ++ ++ Private

Research + + + + + Public

Changes to ++ ++ ++ ++ Private

126

lambing/calving

Changing disease trends – now

Monitoring/ Surveillance

++ ++ ++ ++ ++ Public

Research + ++ ++ ++ + Public

Vaccination ++ ++ + + Private

Changing disease trends – 2020

Animal breeding ++ + Public +ve /-ve

Vaccination ++ ++ + + Private/ Public

Surveillance ++ ++ ++ ++ ++ Public/Private

Changes in farm management

++ ++ ++ ++ ++ Private +ve /-ve

*Some adaptations will make agriculture more resilient to current climate variability. Options which have a (+*) in the No-regrets column assume that the action will increase resilience to the current climate.

13.2.1 Uncertainty Because of the uncertain nature of the impacts of climate change adaptation decisions need to be robust against a variety of future climate outcomes. Hallegatte (2009) proposes a framework for assessing robustness under uncertainty (discussed in more detail in section 6.2). Many of the suggested adaptations relevant for the livestock sector appear relatively robust against uncertainty, using Hallegatte’s criteria. Few major infrastructural irreversible adaptations are applicable to agriculture, with the exception of some related to water storage and irrigation. While some adaptations will entail an initial cost, such as improving shelter or ventilation, these are likely to be no-regrets in the sense that they will provide protection from current climate variability. Many of the adaptation actions are changes in management systems, which will not entail an initial cost and are reversible and flexible; for example adjusting the timing of lambing and calving in response to changing seasonality.

13.2.2 Public/Private Some adaptations are less clearly distinguishable as public or private. Water management and access to water is classified as both public and private, because while individual farmers can improve their water management and water storage capacity, abstraction demands can become a regional issue, particularly if there are competing demands for water. Other adaptations may also be adopted by private or public actors, such as vaccination and surveillance, which may be adopted on an individual farm level, or in the event of an outbreak there may be a regulation requiring vaccination. Research into breeding goals and vaccine development are both public and private.

127

Many of the adaptations listed in Table 13.1 occur at a private level. However there may be barriers to the implementation of adaptations that require an initial investment. A range of instruments might be deployed to overcome barriers. In agriculture, possible policy instruments may include insurance instruments, microfinance, and R&D incentives (Fankhauser et al 2008). The insurance sector (risk sharing) is likely to become more relevant to future adaptation decisions, whether through traditional indemnity-based insurance, or through other options that may be more suitable for climate based insurance, such as index-based schemes, weather derivatives or catastrophe bonds. For more detail on these schemes refer to Barnett and Mahul (2007), Fankhauser et al (2008), Mills (2008). Ideally insurance can create incentives for adaptation and reducing risk by sending market signals about the climate risk and encouraging risk-reducing behaviour through discounted premiums. However in reality this may not occur exactly in this way, because of uncertainty about actual climate impacts, budget constraints and structural, social and cultural barriers which prevent individuals and businesses from adapting, particularly if relocation would be the most appropriate adaptation. In addition, as climate risks increase, insurance costs will also increase and may prove to be too costly for some actors, leaving them highly vulnerable to climate change. In these cases public intervention may be necessary to facilitate the sharing of climate risks between the insurance sector and the state. Fankhauser et al (2008) discuss the role of environmental pricing, particularly in water markets, in encouraging and promoting adaptation to climate change. More generally, the appropriate pricing of natural resources can in fact improve the resilience of ecosystems and enable them to cope better with climate change. The identification and protection of ecosystem services such as watershed protection through appropriate agricultural management and/or forest cover, can provide protection against flooding and erosion, as well as regulating the water table and minimising water pollution. It is important therefore to think laterally about the valuation of ecosystem goods and services; in this case, the role of natural assets as buffers to climate impacts12. Public-private partnership is also an area that could contribute usefully to facilitating adaptation. As well as the financial benefits, public sector involvement sends a clear signal to private industry and individuals that the public sector takes adaptation seriously and is committed to it. Barriers to adaptation identified in some sectors include uncertainty regarding future policy commitment to adaptation, therefore if the public sector is engaged in adaptation activities themselves this may remove some of these barriers. Examples of public private partnership exist in other sectors, such as health, education and research and development. In agriculture, the most relevant public private partnership is likely to be occur in R&D, where the development of technology may facilitate adaptation. Examples already exist in the work of crop genetic improvement networks and proposed foundation of an animal genetic equivalent. In summary, there are significant challenges in promoting adaptation to climate change through policy intervention in agriculture. While economic efficiency (i.e. cost-benefit analysis) provides a rational basis for adaption planning, it is important to recognise a number of complicating factors that limit adaptation responses relative to mitigation

12 We note that the National Ecosystem Assessment for the UK is currently under preparation

and encompass the linkages between ecosystem integrity and benefits to related economic sectors.

128

13.3

action. The first is that while mitigation is likely to be more of a mandatory requirement with more immediate actions, adaptation responses are continual processes requiring constant refinement as damage scenarios become more certain and/or impacts become increasingly apparent. The costs of on-going adaptation and the residual impacts lead to complications in identifying the costs and benefit of adaptation, and in determining the distributional impacts of future adaptation. A second complication is in terms of coordinating how private adaptation responses can be reconciled with desirable public good outcomes. Little is known about how the promotion of private adaptation will impact on public goods, or how these impacts can be minimised through cooperative adaptation planning. This leads to a final observation on the respective private and public good roles. There is clearly a public interest role in the conservation of public goods, and in the facilitation of private resilience. But in the absence of more definitive impact scenarios, that role is largely limited to information provision and investment in research to understand how coordinated action can work. There is currently a limited evidence base on comprehensive adaptation measures, particularly in livestock systems and their costs. Part of the public good role should be to develop inventories of adaptation measures and reconcile these with mitigation requirements.

Interaction with mitigation It has already been noted that many adaptation outcomes are likely to be ancillary benefits of existing environmental policies and regulations that are in place to safeguard environmental goods and services. The extent of these policy synergies warrants further analysis, as does the important interaction with more immediate greenhouse gas mitigation actions. Some adaptations may have unintended consequences on efforts to reduce GHG emissions (mitigation). This is known in the climate change literature as maladaptation. The actions presented in Table 13.1 are screened for whether they would lead to a reduction or increase in emissions. Generally, “pushing” a livestock system in terms of minimising wastage, therefore increasing efficiency and reducing emissions, may have a negative impact on that system’s ability to adapt to a changing climate and may therefore require additional adaptation measures, some which may have a GHG emissions associated with them (e.g., mechanical ventilation). It is important that each mitigation and adaptation tool is screened and their interactions studied to ensure that mitigation and adaptation measures are compatible rather than in conflict. The most obvious example is in terms of increase emissions from cooled or ventilated buildings and transportation. Housing animals for longer will affect the nitrous oxide emissions from manure (being collected centrally rather than from grazing on the fields) but the direction is uncertain. New housing would require energy in the building process and also possibly in the ongoing running of the building, and if mechanical ventilation was included this would almost certainly result in an increase in energy use. Unless the energy was sourced from renewable sources, this would lead to an increase in emissions. Drainage is considered to be positive for mitigation, due to the alteration of the soil structure which will reduce nitrous oxide emissions.

129

Animal breeding is a tool that could help both with mitigation of GHG emissions (e.g., breeding for animal and system efficiency) as well as adaptation (e.g., breeding for thermo tolerance and immunity traits). Animals that can cope with higher temperatures are generally smaller so greater numbers would be required to maintain the same output as larger animals, resulting in a greater level of emissions. Larger, more efficient, animals generally produce fewer emissions per unit of product, but may be less resilient to extremes in temperature or other impacts of climate change. Therefore animal breeding could also incorporate traits of lower methane emissions, whilst balancing breeding objectives to consider adaptation important traits. These interactions and trade-offs highlight the importance of a coherent approach to responding to climate change, and the importance of considering adaptation and mitigation responses together.

130

14 Conclusions and recommendations This report has scoped the requirement for public intervention to facilitate climate change adaptation in the UK livestock sector. A secondary objective has been to determine the basis for identifying economically efficient adaptation if intervention is required. As a man-made adjunct to natural ecosystems, agriculture is potentially the sector of the UK economy most exposed to climate change. The sector may possess a potential adaptive capacity, but in addition to uncertain warming scenarios, inherent biological complexities complicate our understanding of what planned adaptations will be necessary, their timing and effectiveness, and how adaptation responsibilities can be clearly divided between private and public sectors. A key question is whether the extent and variability of predicted changes will motivate accelerated private action that has unanticipated consequences for the rest of society. The focus here is in terms of impacts to public goods that are valued by wider society. The broad conclusion we reach at this point is that while there is a need to adapt, the extent of required adaptation is largely within the capacity of the livestock industry and can be motivated by information to provoke an attitudinal shift and increased awareness that climate change presents certain risks that need to be factored into farm and production planning. This shift does not currently represent a radical change from the gradual autonomous adaptation that characterises the adjustments different parts of the industry have had to undertake in response to market liberalisation over the last two decades. Using current data on grazing potential and pollution modelling, adaptation responses are unlikely to bring about significant impacts in terms of air and water pollution. But the potential for extreme events, heat waves, flooding, and disease are highlighted as key vulnerabilities and significant research gaps that will require revision as part of the Climate Change Risk Assessment process. In relation to the disease impacts, an important policy adaptation has already been noted in terms of the intention to promote cost sharing as a means to alter behaviours. But even with this policy shift, the UK is at risk of introductions of exotic diseases some of which are highly pathogenic and contagious with significant damage potential. With current research we are forced to extrapolate further than the available data warrants when making these predictions and there is a particular need to focus on improving these scenarios. This involves: 1. Data on trends seen in the field for model fitting 2. Quantification of mechanisms of effects of CC on disease epidemiology/ecology to

improve predictive power of models 3. Early warning surveillance systems of approaching threats (i.e. outside the UK) and

outbreaks in the UK. In relation to impacts on livestock production and welfare, the changing climate is likely to be an additional stressor for livestock that needs to be managed. These results show that the gradual change in climate is likely to impact on production output as well as functional fitness of livestock, in some cases leading to increased livestock mortality. However it is extreme events that could have the largest impact on livestock systems and therefore provide the greatest challenge for livestock managers. Also, the

131

interaction of a changing climate with adopted mitigation tools needs to be monitored to ensure that both mitigation and adaptation measures are compatible and sustainable into the future. Continued development of information resources and tools will help farmers improve the resilience of their systems. This could include: 1. Monitoring the thermo tolerance of livestock in UK farming systems and transport

and advising on adaptations, both private (e.g., addition of shelter breaks, housing/grazing patterns) and public (e.g., updating transport and housing guidelines and regulation).

2. Research and development of new and novel tools (e.g., livestock and plant breeding) that help farmers adapt to climate challenges in a cost effective and sustainable way that have no or limited environmental impact in their own right.

3. Understanding the interaction between mitigation tools, a changing climate and adaptation tools.

There is much that the private sector can do to adapt, and in the UK, producers have both the capacity and the knowledge to adapt to projected temperature ranges. However producers do need information on both the potential for worst case scenarios and what their responsibilities are in the event that these arise. The public role therefore lies in building adaptive capacity, i.e. basically letting farmers operate with improved information on risk and in setting the appropriate regulatory and policy framework that facilitates flexibility to change while setting environmental constraints on the directions of change. The adaptation literature offers a number of principles for decisions about both private and public adaptation of relevance to the livestock sector. Broadly these indicate the need to foster an adaptive capacity in the industry and to identify no-regrets investments, and for adaptation options to be flexible (e.g. reversible) to cope with potential worst-case scenarios, or changing conditions (e.g. increasing mitigation requirements as well as a change in projected climate). These principles place more emphasis on so-called “soft” strategies; those that do not involve engineering or infrastructure, but involve institutional or financial tools, which generally have greater flexibility or reversibility than hard adaptation options. Soft adaptive strategies are already inherent in government objectives for fostering resilience and adaptive capacity in agriculture using better information and R&D policy. Alongside these interventions is a clear message on understanding and acceptance that there will be impacts and corresponding losses for which the sector should take responsibility. As noted, a range of existing policy shifts and regulations (e.g. milk quota reform, Nitrates Directive, and those related to housing and transport) can be construed as setting environmental constraints on the direction of adaptive responses. Policies relating to water demand and the conservation of ecosystem goods and services can also be viewed as adding a buffer to the agricultural sector. Adaptation needs to be considered in parallel to a more proximate mitigation objective that is also likely to shape the way livestock are produced in the UK. While we have noted that some adaptations can hinder mitigation, the imperative here is for early mitigation plans to be adaptable to better impact data. In general, there is a need for research on how alternative mitigation strategies may be developed to simultaneously maintain adaptive capacity in farm systems while reducing emissions. The conclusions we reach here do not obviate the need for more detailed social cost-benefit analysis to judge the return to public interventions for specific impacts (e.g.

132

diffuse pollution and disease). However, uncertainty and a lack of data complicate the monetary valuation of impacts which is the ambition under the periodic climate change risk assessment (CCRA).

133

15 References Adger W.N., Agrawala S., Mirza M.M.Q., Conde C., O'Brien K., Pulhin J. et al. (2007).

Assessment of adaptation practices, options, constraints and capacity. In M.L.Parry, O. F. Canziani, J. P. Palutifkof, P. J. van der Linden, & C. Hanson (Eds.) Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change., (pp. 717-743). Cambridge, UK: Cambridge University Press.

Adger W.N., Lorenzoni I., O’Brien K.L (2009). Adapting to Climate Change: Thresholds, Values, Governance. Cambridge University Press, Cambridge.

AEA (2007). Adaptation to climate change in the agricultural sector. Report to European Commission Directorate-General for Agriculture and Rural Development (Rep. No. AGRI-2006-G4-05). AEA Energy and Environment.

Agrawala S. and Fankhauser S. (Eds.) (2008) Economic Aspects of Adaptation to Climate Change: Costs, Benefits and Policy Instruments. OECD Publishing. Available online at: http://www.oecdbookshop.org/oecd/display.asp?sf1=identifiers&st1=9789264046030

Algers B. and Jensen P. (1990). Thermal Microclimate in Winter Farrowing Nests of Free-Ranging Domestic Pigs. Livestock Production Science 25, 177-181.

Altizer S., Nunn C.L., Thrall P.H., Gittleman J.L., Antonovics J., Cunningham A.A., Dobson A.P., Ezenwa V., Jones K.E., Pedersen A.B., Poss M. and Pulliam J.R.C. (2003). Social organisation and parasite risk in mammals: integrating theory and empirical studies. Annual Review of Ecological Evolutionary Systems. 34, 517-47.

Anon. (2006). The Welfare of Animals (Transport) (England) Order 2006, Statutory Instrument No. 3260. Her Majesty’s Stationery Office, London, 12

Averós, X., Knowles, T.G., Brown, S.N., Warriss, P.D., Gosàlvez, L.F. (2008). Factors affecting the mortality of pigs being transported to slaughter. Vet. Rec. 163, 386-390.

Barbari B. and Ferrari P. (2001). Evaluation of thermal characteristics of different types of farrowing huts for outdoor pig production in hot climatic areas. Campinas, SP, Brazil. Agribuilding 2001 3-9-2001.

Barnett B.J. and Mahul O. (2007) Weather index insurance for agriculture and rural areas in lower income countries. American Journal of Agricultural Economics 89, 1241 – 1247.

Bastiman B. and Altman J.F.B. (1985). Silage losses on commercial farms. Research and Development in Agriculture 2, 19-25.

Benham P.F. and Broom D.M. (1991). Responses of dairy cows to badger urine and faeces on pasture with reference to bovine tuberculosis transmission. British Veterinary Journal 147, 517-532.

134

Berger F., Dagorn J., LeDenmat M., Quillien J.P., Vaudelet J.C.and Signoret J.P. (1997). Perinatal losses in outdoor pig breeding. A survey of factors influencing piglet mortality. Annales de Zootechnie 46, 321-329

Berkhout F. (2005). Rationales for adaptation in EU Climate Policy. Climate Policy 5, 377-391.

Betts, R.A., Boucher, O., Collins, M., Cox, P.M., Falloon, P.D., Gedney, N., Hemming, D.L., Huntingford, C., Jones, C.D., Sexton, D.M.H. and Webb, M.J. (2007). Projected increase in continental runoff due to plant responses to increasing carbon dioxide. Nature 448, 1037-U1035.

Blantern P. (1991). Factors affecting nitrogen transformations in grazed grassland soils with specific reference to the effects of artificial land drainage and N fertiliser. PhD Thesis. University of Exeter, UK.

BPEX 2007. Pig Yearbook. MLC economics

Brown L., Scholefield D., Jewkes E.C., Lockyer D.R. and del Prado A. (2005). NGAUGE: a decision support system to optimise N fertilisation of British grassland for economic and/or environmental goals. Agriculture, Ecosystems and Environment 109, 20-39.

Byrne K.A., Kiely G. and Leahy P. (2007). Carbon sequestration determined using farm scale carbon balance and eddy covariance. Agriculture, Ecosystems and Environment 121, 357-364.

Casey J.W. and Holden N.M. (2005). Analysis of greenhouse gas emissions from the average Irish milk production system. Agricultural Systems 86, 97-114.

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.

Chadwick D.R., Pain B.F. and Brookman S.K.E. (2000). Nitrous oxide and methane emissions following application of animal manures to grassland. Journal of Environmental Quality 29, 277-287.

Chadwick D., Marshall A., del Prado A., Scholefield D., Weller R., Hyslop J., Bell J. and Morgan C. (2007). Forage Brassica for Out-Wintering Stock. MDC/Defra Project 06/09/B Report. http://randd.defra.gov.uk/Document.aspx?DocumentID=2456

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.

Ciais P., Reichstein M., Viovy N., Granier A., Ogee J., Ailard, V. et al. (2005). Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature, 437, 549-533.

Cockram, M.S., Mitchell, M.A. (1999). Role of research in the formulation of ‘rules’ to protect the welfare of farm animals during road transportation. In: Russel, A.J., Morgan, C.A., Savory, C.J., Appleby, M.C., Lawrence, T.L., (Eds.), Farm animal welfare – who writes the rules? British Society of Animal Production, Edinburgh,

135

Occasional Publication of the British Society of Animal Production, No. 23, pp. 43-64.

Cook N.B., Mentink R. L., Bennett, T. B. and Burgi K. (2007). The Effect of Heat Stress and Lameness on Time Budgets of Lactating Dairy Cows. Journal of Dairy Science 90,1674–1682

COPA COGECA. (2003). Assessment of the impact of the heat wave and drought of the summer 2003 on agriculture and forestry Brussels.

Da Silva, R.G. (2003). Weather and climate and animal production. In: Lacetera, N., Bernabucci, U., Khalifa, H.H., Ronchi, B., Nardone, A., (Eds.), Interactions between climate and animal production. EAAP Technical series, 7, (chapter 11), 81-90.

Defra (2004) Animal Health and Welfare Strategy for Great Britain. Defra, London.

Defra (2005). Partners for success A farm regulation and charging strategy. http://www.defra.gov.uk/farm/policy/regulation/charge/pdf/farm-regulation-strategy.pdf

Defra (2008) Agriculture in the UK 2008 – Tables and Charts. https://statistics.defra.gov.uk/esg/publications/auk/2008/excel.asp

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 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, 1-17.

Del Prado A., Brown L. and Scholefield D. (2004). Use of the NGAUGE DSS for analysis of options for compliance with NVZ legislation on UK dairy farms. Journal of Agricultural Science. 142: 246�247.

Del Prado A., Merino P., Estavillo J.M., Pinto M. and González-Murua C. (2006a). Nitrous and nitric oxide emissions from different N sources and under a range of soil water contents. Nutrient Cycling in Agroecosystems 74, 229-243.

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. (2006b). A modelling framework to identify new integrated dairy production systems. EGF: 21st General Meeting on Sustainable grassland productivity, Badajoz, Spain, 3�6 April 2006.

Del Prado A., Scholefield D. and Chadwick D. (2009). New integrated dairy production systems: specification, practical feasibility and ways of implementation. DEFRA Final Report IS0214.

Del Prado A., Chadwick D., Cardenas L., Misselbrook T., Scholefield D. and Merino P. (under review). Exploring systems responses to mitigation of GHG in dairy farms. Submitted to Agriculture, Ecosystems and Environment.

136

Dendoncker N., Van Wesemael B., Rounsevell M.D.A., Roelandt C. and Lettens S. (2004). Belgium's CO2 mitigation potential under improved cropland management. Agriculture Ecosystems and Environment 103, 101-116.

Dewey, C., Haley, C., Widowski, T. and Friendship, R. (2005) Lowering in transit losses. Advances in Pork Production, 16, 51-54.

Doorenbos J. and Pruitt W.O. (1977). Crop Water Requirements – Irrigation and Drainage Paper 24. FAO, Rome.

Downing T., Watkiss P. (2005) Adaptation: Policy Perspectives. Presentation at the launch of the European Climate Change Programme (ECCP) II launch. Brussels, 24th October, 2005.

Dragosits U., Chadwick D.R., del Prado A., Scholefield D., Mills J.A.N., Crompton L.A. and Newbold C.J. (2008). Implications of farm-scale methane mitigation measures for UK methane emission. Proceedings of the SAC and SEPA Biennial Conference, Edinburgh, 26-27 March 2008. ISBN 1 85482 867 3. p.168-174.

Easterling W.E., Aggarwal P., Batima P., Branker K.M., Erda L., Howden A. et al. (2007). Food, fibre and forest products Cambridge, UK: Cambridge University Press.

Easterling W.E., Crosson P.R., Rosenberg N.J., McKenney M.S., Katz L.A. and Lemon, K M. (1993). Agricultural impacts of and responses to climate change in the Missouri-Iowa-Nebraska-Kansas (MINK) region. Climatic Change 24, 23-61.

EFSA Journal (2004) 122, 1-25. European Food Safety Authority. “The welfare of animals during transport” Scientific Report of the Scientific Panel on Animal Health and Welfare on a request from the Commission related to the welfare of animals during transport. (Question N° EFSA-Q-2003-094). Adopted on 30th March 2004

Esslemont R.J., Kossaibati M.A., Allcock J. (2001) Economics of fertility in dairy cows. Paper presented at Recording and Evaluation of Fertility Traits in UK Dairy Cattle Workshop. Held at SAC Edinburgh, 19 – 20 November

European Commission, (EC)., 1998. Council Regulation (EC) No 411/98 of 16 February 1998 on additional animal protection standards applicable to road vehicles used for the carriage of livestock on journeys exceeding eight hours. Off. J, L052, 21/02/1998, pp. 8-11

European Commission EC 1/2005; European Union., 2005. Council Regulation (EC) No. 1/2005 of 22 December 2004 on the protection of animals during transport and related operations and amending Directives 64/432/EEC and 93/119/EC and Regulation (EC) No 1255/97. Off. J, L3, 05/01/2005, pp. 1-44.

European Commission (2009). Adapting to climate change: the challenge for European agriculture and rural areas Brussels: European Commission.

Fankhauser S., Agrawala S., Hanrahan D., Pope G., Skees J., Stephens C., Yasmin S. (2008). Economic and policy instruments to promote adaptation. In Agrawala,

137

S., Fankhauser, S (eds) Economic Aspects of Adaptation to Climate Change: Costs, Benefits and Policy Instruments. OECD, Paris.

FAO (Food and Agriculture Organization of the United Nations) (2008). FAOSTAT. 2008. Available: http://faostat.fao.org/

Fialho, F.B., Bucklin, R.A., Zazueta, F.S. and Myers, R.O. (2004). Theoretical model of heat balance in pigs. Animal Science, 79, 121-134

Fiore, F., Natale, J., Hofherr, J., Mainetti, S. and Ruotolo, E. (2009) Study on temperatures during animal transport. Final Report in Framework of the Administrative Arrangement between DG SANCO and JRC No 30775-2007-11 AICO ISP; January 2009, JRC Scientific and Technical Reports, pp 45

Firestone M.K. and Davidson E.A. (1989). Microbial basis of NO and N2O production and consumption in soil. In: Andreae M.O. and Schimel D.S. (eds.), Exchange of trace gases between terrestrial ecosystems and the atmosphere: report of the Dahlem workshop. Berlin, Wiley and Sons, New York.

Forbes T.D.A. and Hodgson J. (1985). The reaction of grazing sheep and cattle to the presence of dung from the same or other species. Grass Forage Science 40, 177-182.

Franco-Jimenez, D.J., Scheideler S.E., Kittok R.J., Brown-Brandl, T.M. Robeson L.R., Taira H. and Beck M.M. (2007). Differential Effects of Heat Stress in Three Strains of Laying Hens. Journal of Applied Poultry Research 16, 628–634

Francos, G., and Meyer E. (1983). Observations on some environmental factors connected with fertility in heat stressed cows. Theriogenology. 19, 625-634.

Gale P., Adkin A., Drew T. and Wooldridge M. (2008). Predicting the impact of climate change on livestock disease in Great Britain. Veterinary Record 161, 214-215.

Gaughan, J.B., Mader T.L. and Holt S.M. (2008). Cooling and feeding strategies to reduce heat load of grain-fed beef cattle in intensive housing. Livestock Science 113, 226-233

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.

Gillespie T.R., Greiner E.C. and Chapman C.A., (2005). Gastrointestinal parasites of the Colobus monkeys of Uganda. Journal of Parasitology. 91; 569-573.

Githeko A.K., Lindsay S.W., Confalonieri U.E and Patz J.A. (2000). Climate change and vector-borne diseases: a regional analysis. Bulletin of the World Health Organisation. 78, 1136 – 1147.

Glendining M.J., Dailey A.G., Williams A.G., van Evert F.K., Goulding K.W.T. and Whitmore A.P. (2009). Is it possible to increase the sustainability of arable and ruminant agriculture by reducing inputs? Agricultural Systems 99, 117-125.

Hahn G.L.; Mader T.I. and Eigenberg R.A. (2003). Perspective on development of thermal indices for animal studies and management. In: Interactions between

138

climate and animal production. EAAP Technical Series No.7, pp.31-44. Wageningen Academic Publ., Wageningen, The Netherlands.

Hallegatte, S (2009) Strategies to adapt to an uncertain climate. Global Environmental Change, 19, 240 – 247

Haynes R.J. and Williams P.H. (1993). Nutrient cycling and soil fertility in the grazed pasture ecosystem. Advances in Agronomy 49, 119-199.

Holmes W., (1989). Grazing management. In: Holmes, W. (Ed.), Grass: Its production and utilization, Published for The British Grassland Society by Blackwell Sceintific Publications, Oxford, pp. 130-171.

Hossell J.E., Temple M.L., Finlay I., Gay A., Oakley J., Symmonds W. et al. (2002). Identifying and costing agricultural responses under climate change scenarios (ICARUS) Final Report to Defra for project CC0357. http://randd.defra.gov.uk/Default.aspx?Menu=Menu&Module=More&Location=None&Completed=1&ProjectID=12008

Hughes G., Wilkinson M., Boothby D., Clarke J., Perkins S., Davies M. et al. (2008). Changes to agricultural management under extreme events - likelihood of effects and opportunities nationally (CHAMELEON) Project Report to Defra.

Hulme M., Jenkins G.J., Lu X., Turnpenny J.R., Mitchell T D., Jones R.G. et al. (2002). Climate Change Scenarios for the United Kingdom: The UKCIP02 Scientific Report Norwich: Tyndall Centre for Climate Change Research, School of Environmental Sciences, University of East Anglia, Norwich, UK.

Hunt A., Boyd R., Taylor T., Kovats R.S., Lachowyz K., Watkiss, P. et al. (2006). Report on the costs of the hot summer of 2003 Bath: Metroeconomic Ltd.

Hutchings M.R. and Harris S. (1997). The effects of farm management practices on cattle grazing behaviour and the potential for transmission of bovine tuberculosis from badgers to cattle. The Veterinary Journal 153, 149-162

Hutchings M.R. and Harris, S. (1999). Quantifying the risks of TB infection to cattle posed by badger excreta. Epidemiology and Infection 122, 167-174.

Hutchings M.R., Kyriazakis I., Anderson D.H., Gordon I.J. and Coop R.L. (1998). Behavioural strategies used by parasitised and non-parasitised sheep to avoid ingestion of gastrointestinal nematodes. Animal Science 67, 97-106.

Hutchings, M.R., Gordon, I.J., Kyriazakis, I. & Jackson, F. (2001a). Herbivore avoidance of faeces-contaminated patches leads to a trade-off between intake rate of forage and parasitism in subsequent foraging decisions. Animal Behaviour 62, 955-964.

Hutchings M.R., Kyriazakis I. and Gordon I.J. (2001b). Herbivore physiological state affects foraging trade-off decisions between nutrient intake and parasite avoidance. Ecology 82, 1138-1150.

Hutchings M.R., Knowler K.J., McAnulty R. and McEwan J.C. (2007). Genetically resistant sheep avoid parasites to a greater extent than do susceptible sheep. Proceedings of the Royal Society, Series B, 274, 1839-1844.

139

Igono M.O., Bjotvedt G. and Sanford-Crane H.T. (1992). Environmental profile and critical temperature effects on milk production of Holstein cows in desert climate, International Journal of Biometeorology 36, 77-87.

IPCC (2001). Climate Change 2001. The IPCC Third Assessment Report. Cambridge: Cambridge University Press.

IPCC (2007). IPCC Working Group Two Fourth Assessment Report. Cambridge: Cambridge University Press.

Jeon,J.H., Yeon,S.C., Choi,Y.H., Min,W., Kim,S., Kim,P.J. and Chang,H.H. (2006). Effects of chilled drinking water on the performance of lactating sows and their litters during high ambient temperatures under farm conditions. Livestock Science 105, 86-93.

Kettlewell, P.J., Hoxey, R.P., Mitchell, M.A., 2000. Heat produced by broiler chickens in a commercial transport vehicle. J. Agric. Engin. Res. 75, 315-326.

Kettlewell, P.J., Mitchell, M.A. (2001a). Comfortable ride: concept 2000 provides climate control during poultry transport. Res. Engin. Technol. Sustain. World 8, 13-14.

Kettlewell, P.J., Mitchell, M.A. (2001b). Mechanical ventilation: improving the welfare of boiler chickens in transit. J. Roy. Agric. Soc. 162, 175-184.

Kilsby C.G., Jones P.D., Burton A., Ford A.C., Fowler H.J., Harpham C., James P., Smith A. and Wilby, R.L. (2007) A daily weather generator for use in climate change studies. Environmental Modelling & Software 22,1705-1719.

Lafferty K.D. (2009a). The ecology of climate change and infectious diseases. Ecology

90, 888-900. Lafferty K.D. (2009b). Calling for an ecological approach to studying climate change

and infectious diseases. Ecology 90, 932-933. Macduff J.H. and White R.E. (1985). Net mineralization and nitrification rates in a clay

soil measured and predicted in permanent grassland from soil temperature and moisture content. Plant Soil 86, 151-172.

MAFF. (2000). Fertiliser recommendations for agricultural and horticultural crops (RB209), 7/Ed. The Stationery Office, London.

Marion G., Smith L., Swain D.L., Davidson R.S. and Hutchings, M.R. (2008). Agent-based modelling of foraging behaviour: the impact of spatial heterogeneity on disease risks from faeces in grazing-systems. Journal of Agricultural Science 146, 507-520.

Mashaly M. M., Hendricks, III, G. L., Kalama M. A., Gehad A. E., Abbas A. O. and

Patterson P. H. (2004). Effect of Heat Stress on Production Parameters and Immune Responses of Commercial Laying Hens. Poultry Science 83, 889–894.

Maw S.J., Johnson C.L., Lewis A.C. and McQuaid J.B. (2002). A note on the emission of nitrogen oxides from silage in opened bunker silos. Environmental Monitoring Assessment 74, 209-215.

140

Mayne C.S. and Gordon F.J. (1986). The effect of harvesting system on nutrient losses during silage making. 2. In-silo losses. Grass and Forage Science 41, 341-351.

McGlone J.J. (1999) Managing heat stress in the outdoor pig breeding herd. Presented at a symposium on outdoor pig production in Brazil, September, 1999.

McGlone J.J. and Pond W.G. (2002). Pig Production: Biological Principles and Applications. 1st edn. Clifton Park, New York: Delmar Cengage Learning.

Mendelsohn, R. (2000). Efficient adaptation to climate change. Climatic Change 45, 583-600.

Mendelsohn R., Morrison W., Schlesinger M. E. and Andronova, N. G. (2000). Country-specific market impacts of climate change. Climatic Change 45, 553-569.

Mendelsohn R., Nordhaus W. D. and Shaw, D. (1994). The impact of global warming on agriculture: A Ricardian analysis. American Economic Review 84, 753-771.

Metroeconomica (2004). Costing the impacts of climate change in the UK: Overview of guidelines. UKCIP Technical Report Oxford: UKCIP.

Metroeconomica (2006) Climate Change Impacts and Adaptation: cross Regional Research Programme - Quantify the Costs of Impacts and Adaptation. Available online at: http://randd.defra.gov.uk.

Mills E. (2008). Synergisms between climate change mitigation and adaptation: an insurance perspective. Mitigation and Adaptation Strategies for Global Change, 12, 809 – 842.

Misselbrook T.H., Sutton M.A. and Scholefield D. (2004). A simple process-based model for estimating ammonia emissions from agricultural land after fertilizer applications. Soil Use and Management 20, 365-372.

Misselbrook T.H., Van der Weerden T.J., Pain B.F., Jarvis S.C., Chambers B.J., Smith K.A., Phillips V.R. and Demmers T.G.M. (2000). Ammonia emission factors for UK agriculture. Atmospheric Environment 34, 871-880.

Mitchel G (2007) Liver Fluke. In Diseases of sheep, editor ID Aitken, pp 195-205 Blackwell Science, Edinburgh Mitchell, M.A., Carlisle, A.J., Hunter, R.R., Kettlewell, P.J., 1997. Welfare of broilers during transportation: cold stress in winter – causes and solutions. In: Koene, P., Blokhuis, H.J., (Eds.), Proc. 5th European Symposium on poultry welfare, 7-10 June, Wageningen. The World’s Poultry Science Association, Carfax Publishing Co., Abingdon, pp. 49-52.

Mitchell, M.A., Kettlewell, P.J. (1998). Physiological stress and welfare of broiler chickens in transit: solutions not problems! Poult. Sci. 77, 1803-1814.

Mitchell, M.A., Kettlewell, P.J., Hunter, R.R., Carlisle, A.J. (2001). Physiological stress response modeling – application to the broiler transport thermal environment. In: Stowell, R.R., Bucklin, R., Bottcher, R.W., (Eds.), Proc. 6th International livestock environment symposium, 21-23 May, Louisville, Kentucky. American Society of Agricultural Engineers, St Joseph, Missouri, pp. 550-555.

141

Mitchell, M.A., Kettlewell, P.J. (2002a). New poultry harvesting and transportation systems to improve bird welfare and product quality (Nuevos sistemas de recogida y transporte de pollos para la mejora del bienestar de las aves y la calidad del producto). An invited review lecture. II Congreso Internacional de producción y sanidad animal, XXXIX Simposium de WPSA, 4-8 November, Barcelona. WPSA, Barcelona, pp. 63-73.

Mitchell, M.A., Kettlewell, P.J.(2002b). Poultry transport – importance and control of the vehicle microenvironment. An invited review lecture. Proc. 11th European Poultry Conference, 6-10 September, Bremen. Archiv Geflugelkunde (special issue), pp. 1-13 Available on CD-Rom.

Mitchell, M.A., Kettlewell, P.J. (2003). Recent advances in understanding the livestock transport thermal environment: ventilation strategies – the way forward? Proc. Animal Air Transport Association (AATA), 29th International Conference, 6-9 April, Washington, DC. AATA, Houston, pp. 1-18.

Mitchell, M.A., Kettlewell, P.J., Hunter, R.R. (2003). Reduction of heat stress in broiler chickens exposed to high ambient temperatures by means of convective cooling. Poult. Sci. 82 (Suppl.), 20 pp (Abstract No. 83).

Mitchell, M.A., Kettlewell P.J.(2004). Transport and handling. In: Weeks, C.A. Butterworth, A., (Eds.), Measuring and auditing broiler welfare. CAB International, Wallingford, pp. 145-160.

Mitchell, M.A., Kettlewell, P.J. (2005). Minimización del stress en el transporte de ganado porcino. Plenary lecture. Proc. 3rd congreso Mundial de Jamón sobre ciencia, tecnología y comercialización, 17-20 May 2005, Teruel. Aragon Vivo, Teruel, pp. 1-11.

Mitchell, M.A. (2006). Using physiological models to define environmental control strategies. In: Gous, R.M., Morris, T.R., Fisher, C., (Eds.), Mechanistic modeling in pig and poultry production. CABI International, Wallingford, UK pp. 209-228.

Mitchell, M.A., Kettlewell, P.J., 2008. Engineering and design of vehicles for long distance road transport of livestock (ruminants, pigs and poultry). Vet. Ital. 44, 201-213.

Mitchell, M.A., Goddard, P., Kettlewell, P., van der Weerd, H., Morgan, M. and Marlin, D. (2009) DEFRA PROJECT AW0939. A review to appraise the evidence for acceptable temperature envelopes for pigs, sheep, cattle, goats and horses during transport.

Molle, G. et al. (2006) Low input animal production based on forage legumes for grazing systems (LEGGRAZE). Final Report for the EU (Contract No. QLK5-CT-2001-02328).

Muriel P., Downing T., Hulme M., Harrington R., Lawalor D., Wurr D., Atkinson C.J., Cockshull K. E., Taylor D.R., Richards A.J., Parsons D.J., Hillerton J.E., Parry M.L., Jarvis S.C., Weatherhead K. and Jenkins, G. (2000) Climate Change and Agriculture in the United Kingdom. Brochure prepared for the UK Ministry of Agriculture, Fisheries and Food.

142

Nadis Health Bulletin (2007). Bulletin on liver fluke control in sheep. Downloaded from http://www.nadis.org.uk/

Omtvedt I.T., Nelson,R.E., Edwards,R.L., Stephens,D.F. and Turman,E.J. (1971). Influence of Heat Stress During Early, Mid and Late Pregnancy of Gilts. Journal of Animal Science 32, 312-&.

Orson J. H. (1996). A review of the impact of the hot, dry summer of 1995 on the main agricultural and horticultural enterprises in England and Wales ADAS.

Pain B.F. and Broom D.M. (1978). The effects of injected and surface spread slurry on the intake and grazing behaviour of dairy cows. Animal production 26, 75-83.

Pain B.F., Leaver J.D. and Broom D.M. (1974). Effects of cow slurry on the herbage production, intake by cattle and grazing behaviour. Journal of British Grassland Society 29, 85-91.

Parry M. L., Rosenzweig C., and Livermore M. (2005). Climate change, global food supply and risk of hunger. Philosophical Transactions Royal Society: Biological Sciences 360, 2125-2138.

Parry M. L., Rosenzweig C., Iglesias A., Livermore M. and Fischer G. (2004). Effects of climate change on global food production under SRES emissions and socio-economic scenarios. Global Environmental Change 14, 53-67.

Parsons D.J., Armstrong A.C., Turnpenny J.R., Matthews A.M., Cooper, K. and Clark, J.A. (2001). Integrated models of livestock systems for climate change studies. 1. Grazing systems. Global Change Biology 7, 93-112.

Paul E.A. and Clark F.E. (1996). Soil microbiology and biochemistry. 2nd Edition. Academic Press, New York, 272 pp.

Peyraud J.L., Verite R. and Delaby L. (1995). Nitrogen excretion by dairy cows: effect of the diet and of the level of production. Fourrages 142, 131-144.

Porter (2005) Review of the outputs of the Defra programme of research on the impacts of climate change on UK agriculture from 1999 to 2005. CC0380

Parry M. L., Rosenzweig C., Iglesias A., Livermore M. and Fischer G. (2004). Effects of climate change on global food production under SRES emissions and socio-economic scenarios. Global Environmental Change 14, 53-67.

Price R., Thornton S. and Nelson S. (2007). The social cost of Carbon and the shadow price of Carbon: What they are, and how to use them in economic appraisal in the UK. Economics Group, Defra.

Randolph C.E., O'Gorman A.J., Potter R.A., Jones P.H. and Miller B.G. (2005). Effects of insulation on the temperature within farrowing huts and the weaning weights of piglets reared on a commercial outdoor pig unit. Veterinary Record 157, 800-805.

Ravagnolo O., Misztal I. and Hoogenboom G. (2000). Genetic Component of Heat Stress in Dairy Cattle, Development of Heat Index Function. Journal of Dairy Science. 83, 2120-2125.

143

Reilly J. N., Baethgen W. and Chege F. E. (1996). Agriculture in a changing climate: impacts and adaptations. In R.T.Watson, M. C. Zinyowera, & R. H. Moss (Eds.), Climate Change 1995: Impacts, Adaptations and Mitigation of Climate Change (pp. 427-467). Cambridge: Cambridge University Press.

Renwick A.R. and Reader M.A. (2004) Business as Usual Projections of Agricultural Outputs. Centre for Rural Economics Research: Report to the Environment Agency, July 2004

Rodda H.J.E., Scholefield D., Webb B.W. and Walling D.E. 1995. Management model for predicting nitrate leaching from grassland catchments in the United-Kingdom .1. Model development. Hydrolog. Sci. J. 40, 433- 451.

Rosenzweig C. and Parry M. L. (1994). Potential impact of climate change on world food supply. Nature 367, 133-138.

Rosenzweig C. and Tubiello F. N. (2007). Metrics for assessing the economic benefits of climate change policies in agriculture. (Rep. No. ENV/EPOC/GSP(2006)12 /FINAL). Paris: OECD.

SAC Farm Management Handbook (2007) The Farm Management Handbook 2007/08. 28th Edition. SAC, Edinburgh

Schils R.L.M., Verhagen A., Aarts H.F.M. and Sebek L.B.J. (2005). A farm level approach to define successful mitigation strategies for GHG emissions from ruminant livestock systems. Nutrient Cycling in Agroecosystems 71, 163-175.

Schils R.L.M., Olesen J.E., del Prado A. and Soussana J.F. (2007). A review of farm level modelling approaches for mitigating greenhouse gas emissions from ruminant livestock systems. Livestock Science 112, 240-251.

Scholefield D. (2004). Cost curve of nitrate mitigation options DEFRA Final Report NT2511.

Scholefield D., Lockyer D.R., Whitehead D.C. and Tyson K.C. (1991). A model to predict transformations and losses of nitrogen in UK pastures grazed by beef-cattle. Plant and Soil 1322, 165-177.

Shepherd A., Cardenas L., Wang J. and del Prado A. (2009). Simulation and measurement of nitrous oxide emission with varying soil moisture. N Workshop 2009.

Shepherd A., McGinn S.M. and Wyseure G.C.L. (2002). Simulation of the effect of water shortage on the yields of winter wheat in North-East England. Ecological Modelling 147: 41-52.

Smit B., McNabb D. and Smithers J. (1996). Agricultural adaptation to climatic variation. Climatic Change 33, 7-29.

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, 77-87.

144

Smith L.A., White P.C.L., Marion G. and Hutchings M. R. (2009). Livestock grazing behavior and inter- versus intra-specific disease risk via the fecal-oral route. Behavioral Ecology 20, 426-432

Smith L.A., White P.C.L and Hutchings M.R. (2006). Effect of the nutritional environment and reproductive investment on herbivore-parasite interactions. Behavioral Ecology 17, 591-596.

Spithill T.W., Smooker P.M., Sexton J.L., Bozas E., Morrison C.A., Creaney J.and Parsons J.C. (1999) Development of vaccines against Fasciola hepatica In Fasciolosis, editor JP Dalton, pp 377-410, CABI Publishing, Oxon, UK

Starr, J.R. (1980). Weather, Climate, and Animal Performance. WMO Technical Note No. 190. WMO Publication No. 684. WMO, Geneva, Switzerland.

Steinfeld H., Gerber P., Wassenaar T., Castel V., Rosales M. and de Haan C. (2006). Livestock’s Long Shadow: Environmental Issues and Options. Rome: Food and Agriculture Organization of the United Nations; 2006.

St-Pierre N.R., Cobanov B. and Schnitkey G. (2003). Economic Losses from Heat Stress by US Livestock Industries. Journal of Dairy Science 86:(E. Suppl.):E52–E77

Subak, S. (1997). Agriculture. In J.P.Palutikof, S. Subak, and M. D. Agnew (Eds.), Economic impacts of the hot summer and unusually warm year of 1995, Norwich, UK: University of East Anglia.

Swain, D.L., Friend, M.A., Mayes, R.W., Wilson, L.A. and Hutchings, M.R. (2008). Combining an active transponder system with sprayed n-alkanes to quantify investigative and ingestive grazing behaviour of dairy cattle in pastures treated with slurry. Applied Animal Behaviour Science 108, 211-222.

The Environment Agency, (2008). Climate change and river flows in the 2050s. Science Summary SC070079/SS1. www.environment-agency.gov.uk

Theobald M.R., Dragosits U., Place C.J., Smith J.U., Brown L., Scholefield D., del Prado A., Webb J., Whitehead P.G., Angus A., Hodge I.D., Fowler D. and Sutton M.A. (2004). Modelling nitrogen fluxes at the landscape scale. Water Air and Soil Pollution. : Focus 4, 135-142.

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

Turnpenny, J.R., McArthur, A.J., Clark, J.A. and Wathes, C.M. (2000a) Thermal balance of livestock 1. A parsimonious model. Agricultural and Forest Meteorology 101, 15-27.

Turnpenny, J.R., Wathes, C.M., Clark, J.A. and McArthur, A.J. (2000b) Thermal balance of livestock 2. Applications of a parsimonious model. Agricultural and Forest Meteorology 101, 29-52.

Turnpenny, J.R., Parsons, D.J., Armstrong, A.C., Clark, J.A., Cooper,K. and Matthews, A.M. (2001). Integrated models of livestock systems for climate change studies. 2. Intensive systems. Global Change Biology 7, 163-170.

145

UKCIP (2002). Climate change scenarios for the UK: the UKCIP02 scientific report.. Downloaded from:http://www.ukcip.org.uk http://www.ukcip.org.uk (Access date: 10/02/2009)

UNFCCC (2007). Investment and financial flows to address climate change. UNFCCC Report.

Van der Wal R., Irvine J., Stien A., Shepherd N. and Albon S.D. (2000). Faecal avoidance and the risk of infection by nematodes in a natural population of reindeer. Oecologia 124, 19-25.

Van Dijk, J. & Morgan, E.R. (2008). The influence of temperature on the development, hatching and survival of Nematodirus battus larvae. Parasitology 135, 269-283.

Van Soest P.J. (1994). Nutritional ecology of the ruminant. Second Edition. Cornell University Press, Ithaca.

Vecerek, V., Malena, M., Malena Jr, M., Volsarova, E., Chloupek, P., 2006. The impact of the transport distance and season on losses of fattened pigs during transport to the slaughterhouse in the Czech Republic in the period from 1997 to 2004. Vet. Med. 51, 21-28.

Warriss, P.D., 1998. The welfare of slaughter pigs during transport. Anim. Welfare. 7, 365-381.

Warriss, P.D., Brown, S.N., 1994. A survey of mortality in slaughter pigs during transport and lairage. Vet. Rec. 134, 513-515.

Warwick HRI (2008) Vulnerability of UK agriculture to extreme events. Project Report to Defra (AC0301)

Watkiss, P (2009). Literature Review. Scoping Study for a National Climate Change

Risk Assessment and Cost-Benefit Analysis, Metroeconomic; AEA; Paul Watkiss Associates. Published by Defra, 2009.

Watkiss, P. Hunt, H and Horrocks, L (2009). Final Report for the Scoping Study for a

National Climate Change Risk Assessment and Adaptation Economic Analysis, Defra Contract number GA0208. Metroeconomica, AEA group, and Paul Watkiss Associates. Published by Defra, 2009.

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., Anthony S.G., Brown L., Lyons-Visser H., Ross C., Cottrill B., Johnson P. and Scholefield D. (2005). The impact of increasing the length of the cattle grazing season on emissions of ammonia and nitrous oxide and on nitrate leaching in England and Wales. Agriculture Ecosystems and Environment105, 307-321.

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.

146

Wellock,I.J., Emmans,G.C. and Kyriazakis,I. (2003.) Modelling the effects of thermal environment and dietary composition on pig performance: model testing and evaluation. Animal Science 77, 267-276.

Wheaton, E., Kulshreshtha, S. and Wittrock, V. (2007). Canadian droughts of 2001 and 2002: Climatology, impacts and adaptation. Volumes I and II Saskatoon, Saskatchewan: Saskatchewan Research Council.

White,M. (2007) NADIS Pig Health – May 2007 Sunburn and Heatstroke/Heatstress. BPEX Knowledge Transfer.

Wiersma, F. (1990). Temperature-humidity index table for dairy producers to estimate heat stress for dairy cows. Dept. Agric. Eng., Univ. Arizona, Tucson.

Willows, R. and Connell, R. (2003) Climate adaptation: Risk, uncertainty and decision-making. UKCIP Technical Report. UKCIP, Oxford

Wilson K. (2009). Climate change and the spread of infectious ideas. Ecology 90, 901-902.

Wreford, A., Adger, N. and Hulme, M (2009) Adaptation in agriculture: historic effects of extreme events on UK agriculture and an assessment of the economics of adaptation. ADAM Project Report to EU Commission

Yamulki, S., Jarvis, S.C., and Owen, P. (1999). Methane emission and uptake from soils as influenced by excreta deposition from grazing animals. Journal of Environmental Quality 28, 676-682.