Skip to main content

AN Integration of Mitigation and Adaptation options for sustainable Livestock production under climate CHANGE

Final Report Summary - ANIMALCHANGE (AN Integration of Mitigation and Adaptation options for sustainable Livestock production under climate CHANGE)

Executive Summary:
The AnimalChange project has provided for the first time a vision of the future of the livestock sector under climate change and has provided direct support to establish policies, while reaching out to stakeholders by organizing symposia, training of scientists, technicians and policy makers.
Global and regional livestock sector scenarios. Shared Socio-economic Pathways (SSPs) were developed for key drivers of the livestock sector, including technology, diets, wastes and losses. Downscaled climate models based on Representative Carbon Pathways (RCPs) were used to project climate change impacts in combination with SSPs based on a cluster of models. Under middle of the road (SSP2) scenario, livestock supply would expand at the expense of forests and natural land, especially with climate change. Under SSP3, livestock productivity would be lower and poverty would reduce access to animal proteins in regions like Sub-Saharan Africa. Under SSP1, however, per capita animal product consumption would better converge across regions, while the sector impacts on deforestation would be moderated.
Reducing uncertainties on emissions and impacts. In a lifecycle analysis, the total uncertainty associated to livestock GHG emissions was estimated at 20% for Europe and ca. 30% for Latin America and Africa. Sources of uncertainty and their regional distribution were quantified leading to recommendations for GHG inventories improvements. Highly significant grassland soil carbon sequestration was evidenced in Europe, Brazil and South-Africa with variability across sites and treatments largely controlled by grazing/cutting management. In sub-tropical study regions, experimental results point to potential overestimates by IPCC Tier 1 methodology of enteric methane emissions at grazing and of N2O emissions from fertilized soils. The difference between precipitation and potential evapotranspiration was the key driver of annual productivity in grassland manipulation experiments across project regions. In a large scale experiment, elevated CO2 compensated for the negative impacts of an extreme heat and drought on grassland productivity. Calibrated models show increased risks of pasture failure by the end of the century in Central Brazil and Southern Africa under RCP4.5 and RCP8.5 while the frequency of forage deficit years increases in Europe especially under RCP8.5 indicating increased drought vulnerability. Nevertheless, on average, grassland production is less affected by climate change than feed crop yields. Hence, climate change could increase the share of herbage based meat and milk production systems. Excess heat impacts on pig and sow production have been assessed showing the role of body size, diets and building environement. A statistical modelling framework demonstrated species specific changes in the distribution of tick-borne diseases in Europe under climate change.
Mitigation options. Significant benefits of increased legume usage in pastures and variable benefits of high sugar grasses have been shown for mitigating soil N2O and enteric methane emissions, respectively. Dietary additives often had additive effects in their ability to mitigate enteric CH4, while their delivery system and cost was found to restrict their potential use.
Adaptation options. Grassland adaptation was assessed showing increased resilience to heat and drought of forage grass populations from Mediterranean compared to temperate origins and of mixed grass-legume swards compared to monocultures. Novel methods to reduce risks from an increased pressure of enteric parasites have been developed for tropical grazing systems.
Integrating adaptation and mitigation (A&M) options. A qualitative overview of A&M options and their possible synergies and trade-offs has been delivered, as well as process based estimates of the technical potential of A&M options. Moreover, the robustness of these options under various climate change scenarios was tested, including a case-study on animal mobility. A new farm-scale model has been developed and calibrated for a range of European, South American and African farms providing a Tier 3 methodology for assessing flows and losses of C and N. For each farm a list of preferred A&M options has been developed and assessed with the FarmAC model for 5 tropical and 5 European farms, while farm-scale barriers to adaptation and mitigation measures were surveyed in all project regions. The cost effectiveness of mitigation options was assessed, with emphasis on pasture intensification in Brazil, a cost-benefit analysis was conducted for selected adaptation options and the synergies and trade-offs between A&M options was assessed for dairy farms.
Regional scale A&M packages. Packages of A&M options suited to the level of vulnerability and mitigation potentials of small scale livestock farmers in Africa were designed, mapped and their impacts on GHG emissions was modeled. Regional case studies on adaptation and mitigation were developed in all project regions, including modeling of regional measure costs and uptake and of eco-efficiency. Finally, a white paper on key EU policy issues for the livestock sector was delivered. To date, the project has led to the publication of more than 120 peer-reviewed articles, of which 66 acknowledging AnimalChange support, and registered 293 dissemination activities.
Project Context and Objectives:
The context
Climate change is probably the most serious environmental challenge facing humanity, threatening the well-being of future generations. Tackling climate change has now become extremely urgent, as the door of climate targets is quickly closing: the later the global emission reduction takes place, the greater the efforts needed to achieve stabilization within tolerable levels. At the same time, impacts of climate change on agri-food systems could be “manageable” by 2050, even though costly, but if no action is taken, the period 2050-2080 is likely to be much more challenging. The urgency for reducing greenhouse gas (GHG) emissions and adapting to climate change is an unprecedented challenge for the international community but can be addressed with the right policies that support innovation and investment to reduce emissions, improve resilience and increase productivity, while accounting for all emissions along the supply chain.

Food security is still an issue for about 805 million people, or about 11.3% of the world population, mostly located in the least developed countries of the planet. For these, livestock play a critical role in improving food security, by supplying protein and micro-nutrients, contributing to agriculture productivity and providing income opportunities. Concomitantly, livestock products are among the food items over-consumed (there is no universal recommendation on livestock product intake but recommended daily protein intake range between 50 and 60 g per capita for adults, source EFSA, 2012) and causing overweight and associated health issues among about another billion people. These contrasted demands on the sector are estimated to drive a production increase of about 70% during the next 4 decades, mostly taking place in developing countries and among intensified production systems.

Livestock matters to climate change. The sector is estimated to contribute a significant share to global GHG emissions: ca. two thirds of direct agricultural emissions, and about 14.5% of total human induced emissions when a supply chain approach is considered. Feed production and processing that release mainly nitrous oxide (N2O) and carbon dioxide (CO2), and methane (CH4) from enteric fermentation in ruminants, are the two main sources of emissions, representing 45 and 39 percent of global livestock sector emissions, respectively. N2O and CH4 released by manure storage and processing represent 10 percent. The remainder is attributable to CO2 released by the processing and transportation of animal products. Globally, GHG emissions from livestock could be reduced by one third if less efficient producers would adopt the best practices of their peers, in the same production system and region. Technologies and practices that help reduce emissions exist but are not widely used. Those that improve production efficiency at animal and herd levels, including feeding, breeding, health and reproduction management, also have productivity co-benefits. In addition, the livestock sector could benefit from carbon offset programs that represent potential additional income. This would however require the development of workable and equitable payment schemes, based on proper Monitoring, Reporting and Verification (MRV).

Climate change matters to the livestock sector. Direct impacts on production range from extreme climatic events, droughts and floods, to thermal stress and reduced yields or water availability. Climate change also affects the sector indirectly through productivity and quality of forages and animal diseases, modifying the patterns of affected areas and livestock vulnerability at the same time. Nearly two billion people in the world depend on livestock for a living, among which are 1 billion poor (
Integrating climate change mitigation and adaptation for livestock is a challenge. To address this, AnimalChange, a project funded by the EU, brought together researchers from 25 organizations in 13 EU member states and 6 non EU countries (Senegal, South Africa, Kenya, Brazil and New Zealand) for 4 years. The project led to significant technology breakthroughs and innovation for mitigation and adaptation at animal, grassland and farm level. It also produced cost and benefits assessments of interventions and tested scenarios at regional and global level to evaluate policy options.
The Problem
The demand for livestock products is growing and is expected to increase by 70% by 2050. Much of this is predicted to be in the form of pig and poultry meat and most growth will be in developing countries where livestock production is a key contributor to rural livelihoods. Climate change will threaten both food security and rural livelihoods through changing patterns of rainfall, increasing incidence of extreme weather and changing distribution of diseases and their vectors. But it will also present opportunities.
The global animal food chain and associated land use change is estimated to generate 18 % of global greenhouse gas (GHG) emissions. However, there are huge uncertainties and we cannot adequately characterize trade‐offs in terms of emission reduction and food production and economic development. Thus policies that are currently in place to curb GHG emissions may prove insufficient and ill advised. AnimalChange aims to improve the estimates of these emissions and provide opportunities in livestock systems to reduce emissions not only in Europe, but also Africa and Latin America.
The Vision
It is within this context that AnimalChange provides a vision of the possible futures of the global livestock sector under climate change in order to create a sound basis for the development of strategies and policies to reduce climate change impacts on, and emissions from, livestock systems at farm, sector and regional scales.
The Outputs
To achieve this vision, AnimalChange has created a consistent suite of scenarios, models, experimental results, assessments and policy support tools. The project has delivered:
- reduced uncertainties concerning GHG emissions from livestock and soil organic carbon sequestration in temperate and tropical grasslands,
- improved quantification of the impacts of heat and drought extremes on grasslands and grazing livestock as part of climate impact assessment,
- integrated models providing the landscape of possible futures for the livestock sector under contrasted socio-economic and emission scenarios and their consequences for food supply, GHG emissions and land use at both global and regional scales,
- integrated assessments of mitigation and adaptation options to climate change for both extensive and intensive livestock systems in Europe, Brazil and for regions in Africa,
- breakthrough mitigation options for enteric methane emissions based on a renewed understanding of the effects of additives on the functioning of the rumen,
- advanced farm scale modeling showing the technical and economic potential for mitigation and adaptation in a range of temperate and tropical livestock farms,
- extensive surveys of barriers for the adoption of mitigation options in livestock systems,
- policy packages for livestock systems adaptation and mitigation in Europe, Brazil and regions in Africa.
AnimalChange has also reached out to stakeholders by organizing symposia, training of scientists, technicians and policy makers and forming a network to alert stakeholders of project outputs and events and establishing a stakeholder advisory process so as to guide the research and its dissemination.
The Impact
By providing alternatives to ‘business as usual’ scenarios AnimalChange provides evidence to impact:

- on the sustainability and competitiveness of livestock systems by showing the potential for:

Reducing greenhouse gas emissions from European livestock systems using cost effective methods,

Reinforcing economic and environmental competitiveness of animal production systems under climate change and increasing their resilience to climatic variability.

- on EU and International policy objectives by providing:

Guidance concerning the CAP objectives that target a sustainable development of agriculture and territories

Support to international policies such as the United Nations Framework Convention on Climate Change (UNFCCC)

Support to the European Climate Change Program and to its policy on adaptation to climate change, as well as the European Commission’s new energy and climate strategy

Guidance to development cooperation concerning developing countries that are most affected and that have the least capacity to deal with climate change.

The beneficiaries
Livestock farmers, animal production support industry, consumers, NGOs, scientists and policy makers in Europe, Latin America, Mediterranean Africa and Sub‐Saharan Africa are the key beneficiaries of the project. AnimalChange has, for example:

- Tested mitigation and adaptation options for the benefit of livestock farmers and their advisors.

- Delivered planning tools to incorporate cost effective mitigation and adaptation strategies at the farm level for the benefit of livestock farmers and their advisors, and to support climate policies.

- Improved integrated models to predict the consequences of climate change for a wide range of scenarios for the benefit of policy makers, consumers and farmers.

- Models for improved greenhouse gas inventory methods for the benefit of policy makers.

Web based training courses to improve the understanding of climate change impacts and strategies for the benefit of scientists and technicians in Europe, Latin America and Africa.
Project Results:
• WP2 – Climate and socio-economic scenarios for livestock systems in project regions
1. Livestock production systems database
AnimalChange contributed to the development of a unique, biologically consistent, spatially disaggregated global livestock dataset containing information on biomass use, production, feed efficiency, excretion, and greenhouse gas emissions for 28 regions, 8 livestock production systems, 4 animal species (cattle, small ruminants, pigs, and poultry), and 3 livestock products (milk, meat, and eggs). The dataset highlights: (i) feed efficiency as a key driver of productivity, resource use, and greenhouse gas emission intensities, with vast differences between production systems and animal products; (ii) the importance of grasslands as a global resource, supplying almost 50% of biomass for animals while continuing to be at the epicentre of land conversion processes; and (iii) the importance of mixed crop–livestock systems, producing the greater part of animal production (over 60%) in both the developed and the developing world. These data provide critical information for developing targeted, sustainable solutions for the livestock sector and its widely ranging contribution to the global food system. This work has been published in the PNAS journal.
2. Livestock production systems transitions and greenhouse gas emissions
The above presented dataset has been implemented in the economic model GLOBIOM with the aim to project the future livestock production systems structure and the related environmental impacts, focusing on GHG emissions. Our results showed that by 2030 autonomous transitions towards more efficient systems would decrease emissions by 736 million metric tons of carbon dioxide equivalent per year (MtCO2e•y−1), mainly through avoided emissions from the conversion of 162 Mha of natural land. We found that livestock production systems transitions could contribute some 20% of total emissions abatement in the AFOLU sector, intra- and interregional relocation of livestock production was estimated to potentially contribute another 40%, and all other mechanisms together also some 40%. Hence, this study, also published in PNAS, points to a new important mitigation wedge in the AFOLU sector next to the commonly considered technological options and potentially reductions in consumption: restructuring of the sector towards more GHG efficient production systems within and across the world regions.
3. Multi-scale model cluster
The proposed scenario work in this project faced the challenge of developing an economic modelling framework with global coverage in terms of regions but also land use sectors, and with sufficient spatial (sub-national level for EU countries and project SICA regions) and technological (production system representation) detail to provide robust results at policy-relevant level of resolution. For this purpose the EU focused economic models AROPAJ and CAPRI were linked with the global economic model GLOBIOM. The coupling consisted of two components i) harmonization of datasets around the base year for each of the models, and ii) harmonization of the baseline, in particular between CAPRI and GLOBIOM. This modelling framework provides the EU with an unprecedented capacity for assessment of forward-looking scenarios of the agricultural sector in general, and the livestock sector in particular.
4. Alternative planetary developments to 2050
The objective of the project was to explore a large space of plausible livestock futures in the context of alternative socio-economic development storylines and climate change scenarios, rather than to predict the future of the sector. Therefore we anchored our work in the new IPCC scenario framework constructed along two dimensions, the Shared Socio-economic Pathways (SSPs) and the Representative Concentration Pathways (RCPs). SSPs contain quantitative information on future population and economic development, and qualitative storylines for the different economic sectors and social aspects. Within AnimalChange, we have converted the qualitative storylines for the agriculture and livestock sectors into quantitative drivers to be used in economic models. The aspects covered are: future dietary preferences, incl. the place of milk and meat in human diets, losses and waste management, technological change in the crop and livestock sectors, and productivity growth of grasslands. Particular effort was devoted to the reconstruction of past trajectories in feed conversion efficiencies (FCE) by livestock category, for which a new model – AgRIPE (Agricultural Representative Identity based Pathways) – has been developed. This model than allowed us also to develop SSP specific future FEC scenarios. The results indicate that while in the monogastric sector, further increases in FEC will be limited across the world, there are still large potentials for efficiency gains in the ruminant sectors, in particular in Africa and Latin America.
5. Livestock sector towards 2050
Three out of the five SSPs have been quantified by the multi-scale model cluster: SSP1 – characterized by low population growth and high income growth, in particular in developing countries, with corresponding fast technological change and preferences for sustainable development, SSP2 – where many of the current trends are extrapolated in the future, and SSP3 – characterized by fast population growth and slow economic growth and technological change, and conventional diets. Depending on the SSP, the livestock sector production is projected to grow between 2000 and 2050 globally by 59-91% for milk, 22-97% for monogastric meat, and 23-92% for ruminant meat. The production is much less dynamic in Europe, with 16-36% for milk (–6)-32% for monogastric meat, and (–5)-17% for ruminant meat. Producer prices are projected to decrease or stagnate in Europe, while globally they can increase by up to 39% in the case of ruminant meat in SSP3. Total annual greenhouse gas emissions from agriculture are under SSP2 projected to increase from about 3.6 Gt CO2e in 2000 to 5.3 Gt CO2e in 2050. Direct emissions from livestock production represent 66% of total agricultural emissions. Latin America and sub-Saharan Africa would contribute respectively 17% and 15% of the total direct livestock emissions, which is relatively comparable with the contribution from Europe – 10%. On the other hand, the quasi-totality of GHG emissions from land use change would come from Latin America and sub-Saharan Africa. The total AFOLU emissions would be by 36% and 7% lower compared to SSP2 under SSP1 and SSP3 respectively.
6. Livestock production systems and climate change impacts
In the framework of this project, the first-ever comprehensive assessment of the livestock sector future under climate change was carried out. We found that climate change impacts on crop and grass yields could have only small effect on global milk and meat production by 2050, which remains under any climate scenario within +/-2 percent of the projected production without climate change. However, depending on the scenario, the climate change effects can be more pronounced at the regional scale. In sub-Saharan Africa, the effects are both the most uncertain and potentially the most severe; e.g. ruminant meat production could increase by 20 percent but it could also decrease by 17 percent. The effects on regional consumption are less pronounced because the impacts of climate change are mostly buffered through international trade. Virtually all the negative effects are smaller than 10 percent. Finally we found that also, with respect to the climate change impacts, livestock production systems restructuring may play an important role as an adaptation option. Grass yields benefit more (or are hurt less) from climate change than crop yields. Climate change would hence favor the grazing systems, leading potentially to a change in the current trend towards more intensive systems.

• WP3 – GHG emissions from livestock sector: reducing key uncertainties
WP3 had 5 tasks with individual objectives.

The main objective of Task 3.1 was to quantify the GHG emissions from livestock production chains at regional level in EU-27 and in study regions of Africa and Latin America, and to assess the factors controlling the accuracy of the mean estimates.
Contributions were made to two scientific publications about GHG emissions from agriculture. In the paper by Leip et al. (in press) N footprint results (incl. N2O) for two models (MITERRA-Europe and CAPRI) were compared for six vegetable and six livestock commodities.
A report and paper were delivered on “Current GHG emission from livestock production at regional scale”. Major efforts were made to develop the MITERRA-Global model, which extends the MITERRA-Europe model to entire world, incl. the AnimalChange regions Africa and Latin America. The model is ready and now includes data on feed allocation and N excretion. The development of the model with all input data has been described in D3.5.

The main objective of Task 3.2 was to assess the uncertainties in the relations between animal production (milk, meat, egg, etc.) and GHG emissions at regional level in EU-27 and in study regions of Africa and Latin America.
A review was completed (D3.1) about (i) calculation methodologies of GHG emissions from livestock production, (ii) the uncertainty in modelled GHG emissions from livestock production and (iii) the improvement of modelling of GHG emissions at various scales so as to reduce the uncertainty in the estimates of GHG emissions. The study showed that uncertainty in the activity data can be as large as the uncertainty in emission factors. For CH4 emission from enteric fermentation, there is shortage of data on feed intake and nutrient management at regional scales. Uncertainty in N2O and CH4 emission from manure storages emanate from shortage of data on storage type and size and type of manure stored. Uncertainties related to N2O emissions from soil can be reduced by improving the prediction of denitrification conditions and the N2O : N2 ratio. The results of the review were used also to focus activities related to the quantitative uncertainty assessment.
An overview was given of current modelling expertise in this area and of the advantages and disadvantages of different types of models. The review included the specification of an operational methodology for spatial uncertainty quantification and spatial uncertainty analysis for the regional scale models in AnimalChange.

The main objective of task 3.3 was to analyse the role of grassland management on carbon sequestration in pastures and animal feed producing areas (EU-27, Africa and Latin) with i) continuous measurements of potential C sequestration by eddy covariance technique (France, UK, Hungary and French Guiana), ii) analyses of soil C stocks of contrasted feed production in Africa (NRAT, ISRA, UP) and Latin America (CIRAD, UFRGS) and Europe (France, NL, CH, IR) and iii) Models simulation (PaSim, RothC) on soil carbon dynamics as a function of environmental conditions and agricultural management were carried out and models (PaSim, DNDC).
The analysis of the role of grassland management on carbon sequestration in pastures and animal feed producing areas (EU-27, Africa and Latin), revealed that moderately managed (in terms of grazing pressure and grazing period) grassland are a sink of C (0.70 ±0.15 tC/ha.yr with 95% of sites being between 0.3 et 0.8 t C/ha.yr) in the long term (10 to 50 years). For tropical pastures (with and without deforestation) results provide evidence that an installation and development of C3 species (legumes and weed) increase inputs of C and N to soil and thus to soil C stock. This carbon is mainly sequestered in deep soil layers (0.2-1 m). Attention has to be paid to grasslands installed on organic and drained soils as these tend to loose C over time, whatever management is applied.
Concerning management per se, climate and the inter-annual variability of climate (e.g. dry and wet season) play a major role in the choice of management practices (i.e. mowing vs grazing, winter vs summer grazing) and subsequent C storage potential. For example for grazed grasslands increase C sink activity by 2% with an increase in annual precipitation of 10mm. Under mowing, mean annual temperature seems more important than precipitation.
With respect to climate seasonal variation, in dry warm periods and high intensity of herbage use, grasslands are likely become a source of C. For management in general, a number of paired sites also show a high C sink activity under grazing compared to mowing, underlining the importance to regulate/prevent grassland form high (frequent) biomass removal. In addition, in order to counteract grassland degradation and improve C storage, grasslands may be set aside (i.e. grazing exclosure) or periods of herbage use (by grazing and mowing) may be reduced/moved.

The main objective of task 3.4 was to analyse the methane emissions from feed intake and forages in arid areas. Three types of forages were selected:
-Forage representative of intake by zebus, rather than offered forage, obtained by hand-plucking technique, throughout one year (monthly sampling) in order to determine seasonal changes in methanogenic power of consumed forages.
-Forage categories obtained from the above mentioned samples by manual sorting and weighing: grasses, herbaceous legumes, legume tree leaves, other broadleaves, undetermined dried grass, in order to show whether the association of these categories by animals in their diet results in the additivity of methanogenic power, or not.
-Specific plant species (grasses, herbaceous legumes, legume tree leaves, other broadleaves), chosen according to their content in fibre, lignin or tannins determined by NIRS, or according to their NIRS spectrum, in order to evidence the variability of methanogenic power, and to establish relationships with their chemical composition.
A total of 180 forage samples was assayed. In vitro determinations have been made, and analysis of results is in progress.

The main objective in task 3.5 was to reduce uncertainties in CH4 and N2O emissions from manure management.
A protocol for analysis of manure quality, storage and measurement of GHG emissions has been developed; this protocol would identify places in Africa for sampling of manure and data on feed quality and GHG measures at Alterra. The approach was started up twice and failed twice due to logistical problems in the samplings of data and manure in Africa. We have developed an alternative approach.
Experiments were conducted to assess animals’ GHG emission from different long-term livestock systems in the Pampa Biome of South America. These results have been used to assess uncertainty in the in the relations between animal production (meat) and GHG emissions. Our findings are that GHG emission per animal is highly variable depending on animal intake rate, production system and season, which affects the quality of the feed. Four experiments were conducted using SF6 facilities to measure GHG emission and Alkanes methods to measure intake. Two of them involved beef cattle systems, one on native grazing areas with different grazing intensities and one on cultivated tropical pastures. The two other experiment involved crop / livestock integrated systems with respectively beef and sheep.

• WP4 – Climate change impacts on livestock systems: reducing key uncertainties
The objective of WP4 was to reduce uncertainties in impact studies on grassland-livestock systems via experimental and modelling work. This WP has provided an integrated framework to test hypotheses, by combining: i) experiments on grassland ecosystems aimed at testing the role of increased climatic variability in interaction; ii) impacts on animal health, encompassing the effects of increased heat stress and the impacts of increased occurrence of infectious diseases and gastro-intestinal parasites; iii) mechanistic models capturing the role of extreme events such as heat waves and severe droughts on pastures and crops.
Grassland manipulation experiments have been conducted in Europe, Africa and Latin America to assess the impacts of changes in rainfall and of increased climatic variability and extreme events on grassland productivity. In Europe, sensitivity to drought was especially observed in Ireland compared to two Swiss sites. Overall, leguminous species (deep-rooted Cichorium intybus) suffered less than non-N fixing species, and mixtures performed better than monocultures. In the sub-tropical conditions encountered in South Africa, changes of precipitations had a strong effect on herbage yield. In Brazil and Senegal, biomass reduction was less marked with rain reduction. The studies carried out in the Ecotron of Montpellier (France) showed that elevated CO2 concentrations increase both NEE (CO2 net exchange ecosystem) and GPP (gross primary production) before and during moderate drought, and after rehydration, which leads to higher resistance and recovery to drought under CO2 enriched conditions. In addition, induced-elevated CO2 nitrogen decrease was compensated by induced-extreme drought nitrogen increase, whereas fibre increased, meaning that an increase of forage digestibility is expected under future climatic conditions. Some evidence of above ground biomass with elevated CO2 was provided by open-top chamber experiments run in Hungary, with changes in N2O emission with elevated CO2.
To assess climate change impacts on animal health (direct effects of excess heat, infectious diseases and gastro-intestinal parasites), grazing trials were conducted in French West Indies to reduce parasitism: 1) with young goat, reared mixed with cattle, to help reducing the impact of parasitism and improve performance growth; 2) with sheep conducted on plots receiving vermicompost containing earthworms, which through their action on the organic matter faeces excreted, have helped to reduce the parasitic load, in addition to have a positive effect on the organic matter of the soil. It was demonstrates that non-climatic factors can be just as important as meteorological factors. Combinations of variables linked to animals affect the faecal egg count in Creole goats, e.g. sex, age, rearing mode and average daily gain. Increasing humidity and number of rainy days with more than 5 mm precipitation are significant drivers of parasite burden on kids, expressed as the number of eggs per gram of faeces.
The threat to animal production systems of vector-borne/environmental pathogens and their spread in relation to changes in climatic conditions was also assessed. For high priority livestock diseases/disease vectors, current climate suitability was mapped, and their potential climatic distributions modelled under different climate change scenarios. Geographical locations of tick disease vectors, vector-borne and infectious diseases were obtained from intentational datasets. Advanced methods were used to evaluate species distribution models based on current climate data, before projecting those which exhibited strong climate signals, using ISI-MIP projected climate. It appeared a general northward shift in suitable climate in the EU as predicted for tick species under scenarios RCP 4.5 and 8.5. Of the diseases modelled, those which demonstrated strong climatic controls are all predicted to experience increased European climatic suitability by the end of this century, while the degree of increase and northward shift of change vary considerably between the two scenarios.
Simulation results were illustrated for a spectrum of models to assess the impact of climate change at different regions and over different crop and grassland systems. The effort was to assemble approaches, applications and prospective developments as a review of the international effort towards the documentation, in a different and much larger scale than today, of impact models of interest for grassland-livestock systems.

An emerging challenge has been the upscaling of model estimates, e.g. the determination of model parameters for large spatial units. The INRA’s experience with the grassland-specific model PaSim was documented, in which advanced techniques (sensitivity analysis and Bayesian statistics) have been applied to identify and calibrate a set of relevant parameters at the European scale. PaSim mechanistic view of grassland C and N fluxes was exploited by CEA to improve ORCHIDEE, a dynamic global vegetation model extensively used in impact studies, which was successfully evaluated in Europe against data of different nature. For both PaSim and ORCHIDEE, the first experiences run by EMBRAPA in Brazilian beef production areas have also been documented, which have opened to the introduction of new approaches in view of further testing. Modelling solutions also emerged for analyzing climate change impacts in Africa. South Africa provided suitable datasets to assess, in a comparative fashion, a locally-developed model (SWB-SCI) and a flexible tool such as STICS (developed by INRA and mainly evaluated in Europe) for C4 crop and grassland simulations. In Africa, large portions of surface lands are covered by rangelands, complex adaptive systems in which the production of domestic livestock is based on natural or semi-natural plant communities (containing both grasses and woody plants), and where grazing effects are inherently variable and difficult to conceptualise and implement. A contribution to rangeland modelling was given by the experiences gained by ILRI with the model G-Range.

• WP5 – Climate change vulnerability of livestock systems
The vulnerability of a system to climate change can be defined from the residual impacts of climate change after autonomous adaptation. Vulnerability, therefore accounts for a combination of risks, sensitivity and adaptive capacity. Two complementary approaches of vulnerability were compared:
- a probabilistic risk assessment, which combines probability distribution functions (PDF) of climatic hazards (exposure) and sensitivity to those hazards. The PDF of climate change impact is then calculated from the probability of sensitivity conditional to exposure. In this way, the framework includes both the effects of variability (in climate) and of uncertainty (in model parameters and structure). Adaptation can further be considered to derive the probability distribution function of residual impacts after adaptation, which defines a probabilistic vulnerability (see Van Oijen et al., 2013).
- a vulnerability index (derived from Luers 2003) which varies according to the ratio between future and current time periods for the coefficient of variation and mean of a given state variable.
Vulnerability of grassland and crop production was calculated for a combination of impact models and of calibrated climate models forced by representative carbon pathways (RCP4.5 and RCP8.5) leading to gridded projections over Europe, Brazil and Africa.
Results show increased risks of pasture failure by the end of the century in Central Brazil and Southern Africa under RCP4.5 and RCP8.5 while the frequency of forage deficit years increases in Europe especially under RCP8.5 indicating increased drought vulnerability.
With the Orchidee-GM model, for high latitudes of the North Hemisphere, an increase of grassland productivity and grazing animal stocking density is projected over the century. In contrast, in south and central Europe and United States a decline is simulated. For Europe, this decline is observed for both scenarios but mainly with the HADGEM2 climate model. For south-eastern United States, oppositely, there is a shift between the two scenarios: for RCP4.5 there is a relatively large increase of animal productivity, whereas there is a decrease in the RCP8.5 scenario. For the tropics, almost all regions show a decrease in grassland productivity and animal stocking density except in mountainous regions. This decrease is still moderate for RCP4.5 scenario but becomes very severe with the scenario RCP8.5.

The vulnerability index by Luers et al. (2003) was applied to projections for dry years (25% quartile), showing increased grassland gross primary productivity vulnerability by the end of the century especially in the Mediterranean region. In contrast, reduced climate change vulnerability was found at high latitudes in Northern Europe. With dry-matter production, increased vulnerability of production was found especially for Central-Eastern Europe and for the British Isles according to PaSim. The application of the Orchidee model to Brazil indicated a largely increased vulnerability for grassland production and animal stocking density over most of the tropical area of the country.

• WP6 – Breakthrough biophysical mitigation options at field and animal scale
WP6 objectives were to identify new and forthcoming mitigation options and quantify both the potential size of the reduction in GHG emissions obtainable and the uncertainty associated with this reduction. Our activities were focused on three interrelated subtasks addressing:

• Pasture based animal production systems
• Intensive animal production systems
• Horizon scanning for new mitigation options

In the first subtask we have collated and analysed data from a wide variety of sources involving the use of novel grass varieties and grass / legume mixtures in order to understand the potential to replacing mineral N fertilizer use and thus reduce N2O emissions in pasture based systems. We have collated data on the use of high sugar grasses and the potential of these to both decrease N excretion from animals (through better microbial protein synthesis in the rumen) and decrease methane emissions. We agreed a common protocol for measuring C sequestration and made measurements and collated data form over 30 managed grassland sites.
The major finding from our work can be summarised as follows:

1) C sequestration: In terms of livestock pasture management, extensive grazing was observed to result in the highest annual C balances compared to cut or cut and grazed systems. Stocking rates of circa. 1.5 LU ha-1 was observed to lead to higher C storage compared to higher 2.5 LU ha-1 stocking rates. Improved pasture management, especially organic N fertilisation led to enhanced sequestration (0.2 t C ha-1 yr-1).
2) Use of Legumes: Grass-legume mixtures were observed to have a higher total N compared to grass monocultures was of 80-100 kg N ha-1 yr-1. Comparisons of grass-legume emissions within systems trials showed that emissions can be reduced considerably. Soil N2O emissions from white clover/ryegrass systems receiving no fertilizer N or 58 kg N ha-1 yr-1 were found to be 16-21% lower relative to a conventional grass/fertilizer system (226 kg N ha-1 yr-1).
3) High Sugar Grasses: Increases in productivity and reductions in both N2O and CH4 emissions (of circa 15-20%) have been reported in animals consuming HSG. However results are inconsistent and this may reflect differences between cultivars of HSG and/or the growing conditions used.

In our second task we have reviewed and summarized potential mitigation strategies for different manure handling systems in Europe and a case study based on Mali has been developed. These studies have raised the issue of what scale manure management should be considered at, farm, region or country, and this is being addressed. Experiments to determine the effect of diverse feeding strategies with different protein sources on emissions of methane and ammonia in pigs have been carried out and provide novel information on nitrogen utilization and excretion (fecal or urinary) in monogastrics. A comprehensive meta-analysis of published studies on the use of additives to decrease methane emissions from ruminants has been completed, the review has considered both the size of the effect that might be expected and variability in the response. Two additives (nitrate and linseed oil) were selected from meta-analysis of literature. Trials in sheep, cattle and goats have been completed and provide a unique data set through which to understand the variability in responses between individual animals and animal type.

1) Protein sources for monogastrics. It was observed that dietary CP can be reduced from 159 to 136 g CP/kg diet in growing pigs without impaired animal performance and carcass characteristics, as long as sufficient amounts of indispensable amino acids were provided. Slurry pH and NH3 emission was reduced linearly with decreasing dietary CP concentration, but CH4 was not affected. CH4 during housing showed that the daily emission per pig increased with increasing body weight whereas no consistent effect of CP level could be detected.

2) Mitigation of enteric methane production from ruminants. Experiments in the UK, New Zealand and South Africa have shown that 2% nitrate in the diet DM lowers methane emissions per cow per day by 21% for both experiments and between 13 and 16% per kg DMI, but 4% linseed oil in the DM did not reduce emission. The Mitigate meta-analysis database has been made publicly available: http://mitigate.ibers.aber.ac.uk with functionality improved to allow authors to add own data to the online database. Analysis of data from both within AnimalChange and other published studies suggest that in sheep and cattle methane production can be significantly decreased with little effect on rumen function and diet digestibility. Furthermore a number of in vitro trials have been carried out investigating the interaction of different dietary additives. Data suggest that in many cases additives are indeed additive in their ability to mitigate against ruminal methane production and that if one were to combine sufficient approaches then the zero methane ruminant might indeed be achievable. However, whilst perhaps technically possible, the zero methane ruminant presents some very real practical issues associated with diet formulation, additive delivery and cost.

In our third task we provided a framework to undertake horizon scanning for new mitigation technologies and to understand the variability in response to existing mitigation strategies. We have established programs of work to review and evaluate new developments in animal genetics and immunological control of rumen function and to consider their usage. Rumen samples from task 2 provided a valuable resource to understand the role of microbial populations in determining the success or otherwise of mitigation strategies. Also a comprehensive program on the use of novel crops and plants to decrease emissions from tropical animal production has been completed. Major findings include:

1) Use of novel plants to mitigate enteric methane. A unique database has been established based on the screening of tannin and saponin rich plants from Africa and South America for their effects on rumen fermentation and in a number of cases in vivo rumen effects and methane production. This database should both improve the accuracy of prediction of greenhouse gas emissions from animals eating such plants and also the potential identification of novel mitigating agents.

2) Rumen responses to methane mitigation. The transcriptomic analysis of the rumen suggest that the reduction in methane in cattle fed nitrate was as a result of the reduction in numbers of Archaea, due the toxic effects of Nitrite in the rumen (after conversion from Nitrate). The analysis revealed the adaptability of the rumen community to high levels of stress and the metabolic pathways available to capitalize on these changes. This flexibility suggests that it is likely that rumen function can be maintained while under stress from mitigation strategies.

3) Using immunological control and selective breeding to decrease CH4 and N2O emissions. The use of vaccines to control rumen methanogens is likely to be cost-effective and one of the few options which would be practical in pasture-grazed animals. However as yet no definitive proof that the approach will be successful in driving large decreases in methane emissions has been published. Notwithstanding this however, it seems likely that vaccine, particularly those based on specific protein or sub-cellular fractions key to the metabolism of rumen methanogens, will be “universal” and not geographically limited. We reviewed the decreases in GHG emissions that could be obtained through improvements in animal genetics and concluded that altering selection objectives in ruminant breeding programs to target environmental goals could enhance the reduction in GHG emissions at a relatively small economic cost. Genetic improvement tools provide a useful and cost-effective mechanism to help livestock agriculture meet the challenges of the reducing GHG emissions.

• WP7 – Breakthrough biophysical adaptation options at field and animal scale
WP7 objectives were to develop and test new techniques for adaptation of livestock production systems to future climatic conditions (global warming, higher variability of climate and increased frequency and severity of droughts). These adaptation options targeted the most vulnerable components of livestock systems.
A main objective was to examine the adaptation of grassland production by an optimal use of sward functional biodiversity and specifically of legumes. (i) The adaptation potential of mixing legumes and grasses as well as cultivars with a different growth pattern, adapted to temperate or Mediterranean climate, was tested in three new experiments in temperate (France) and Mediterranean (Portugal) climate. (ii) A database was developed based on a completed three year continental-scale experiment (30 sites, 16 European countries and Canada) that had tested the effect of mixing different functional types of plants (legumes and grasses, fast establishing and persistent species). This database was made available to the whole AnimalChange consortium immediately and in the last year of the project was made publicly available by publishing in Ecological Archives. Data analyses resulted in several publications in high quality scientific journals. (iii) In a desk study more than 70 experimental and review articles were used to summarize the potential role of forage legume intercropping specifically for African conditions.
Mixed swards showed consistently higher yields and lower weed invasion than monocultures. All these studies found that such exploitation of functional biodiversity is a major contribution to sustainable intensification of agriculture. The contribution of legumes was found to be a key element, but also other functional differences among species as e.g. growth patterns that resulted in more yield with the same input of resources. Of special interest is the N gain of grass-legume mixtures as compared to grass-only stands, due to the legume’s symbiotic N2-fixation. This increase in N-efficiency is an important mitigation option of climate change (N fertilizer replacement).
New experiments were carried out according a common protocol to test the effect of drought stress on monocultures and mixed swards (Ireland, Switzerland; tight collaboration and data sharing with WP4). These experiments reveal that functional types of N2-fixing legumes and of deep rooting species are more resistant to moderate drought stress, and thus cropping these species in mixed plant communities is an important adaptation option to climate change. At least as important is that although mixtures and monocultures both yielded less under drought than under control, the mixtures’ yield advantage compared to monocultures was so great that, the mixture yield in the drought treatment was comparable to the average monoculture yields in non-drought conditions.
Due to this high adaptation and mitigation potential of legumes quantified in these many experiments under so diverse climatic conditions, it is a major outcome of AnimalChange that legumes represent a major adaptation and mitigation option to climate change.
Measurements of climate change on forage quality were carried out on plant samples of the AnimalChange experiments and an extensive literature review and meta-analysis was published. The consistent result is that no dramatic change in forage quality was induced by climate change factors. The most important effect seems to be a shift of species composition in grasslands due to the changed environmental conditions.
A main objective on animal scale focused on the development of adaptation options to reduce the effect of excess heat on pig and sow production. Model development was successful, and based on this a decision tool was developed.
The work package successfully developed adaptation options to reduce the effects of increased gastro-intestinal parasitic spread on small ruminants. Mixed grazing of goats and cattle is an efficient way for controlling helminths and improving livestock performance. Using HPLC profiles it was possible to build a typology of secondary plant metabolites to better predict the actions of plants on production, health, and parasite burden of livestock. Finally the potential of using vermicompost as a way to reduce the gastrointestinal parasitism and increase biomass production has been confirmed.
In all three tasks, progress was as planned and the tight collaboration among several of the partners is a highlight. Over the whole project duration, WP7 activities were successful and all the planned activities were completed.

• WP8 – Integrating adaptation and mitigation options
The aim of WP8 was to test the implications of mitigation at the field/animal scale on the potential to adapt to climate change, and to test the implications of adaptation on the potential to mitigate greenhouse gas emissions. WP8 provided an integrated framework to select and evaluate alternative adaptation and GHG mitigation measures via an array of biophysical models covering a variety of scales from the animal to the grassland ecosystem, while also accounting for management intensities and regional differences. The objective of Task 8.1 was to gather site-specific information required to address the major mitigation and adaptation options for a range of systems and regions. The objective has been reached. The main outcome of this Task has been a qualitative overview of mitigation and adaptation options at the field/animal scale and their synergies and trade-offs, delivered in year 1, which has become the basis for modelling the effect of different options at field/animal scale, farm scale and regional scale in different Work Packages. The objective of Task 8.2 and 8.3 was to identify the effects of mitigation and adaptation options on GHG emission by employing process-based models in, respectively, intensive ruminant production systems and extensive pasture based systems. The objective of Task 8.4 was to make a final evaluation of the various adaptation and mitigation options tested for a limited number of sites and of farming systems. This evaluation aimed at assessing how robust these options are for an extended range of farming systems and of sites. The objectives have been reached. Three basic models (DNDC by Teagasc, PaSim by INRA and Tier 3 approach by DLO) were identified as suitable to test a certain number of options at specific test sites reflecting conditions representative of alternative agro-ecological zones. The three models reflect different levels of detail: DNDC is a detailed soil model, PaSim is a grassland-livestock mechanistic model and the Tier 3 approach focuses on the animal. The inputs and outputs of these models have been linked (and adapted or extended when needed) to allow for a prediction of the effect of mitigation options on the various GHG emissions occurring in the farming system (available forage, diet consumed, excretion, manure storage, and manure application). Modelling actions (identification of impact models and calibration) were developed to support the simulation of GHG emissions from grassland-livestock systems for a selection of sites and conditions. Simulations included the emission of CH4, N2O and CO2 at the level of the animal, field and manure storage. Process-based models for each component were used in combination and compared to the assumptions adopted with IPCC Tier 2 approaches. Farm adaptation and the application of mitigation measures were integrated through simulation studies with process-based models (cattle, soil/crop, manure) for a selection of farm cases. The effect of climate scenario on trade-offs between the various sources of GHG emission were identified. Focus was laid on those mitigation options that are most relevant with respect to accounting for trade-off mechanisms between different GHG sources: stocking density, N fertilization rate, legumes, full grazing vs. restricted grazing. PaSim was employed to generate maps of GHG emissions at European scale under management and climate scenarios. Also, DNDC has been run for a range of mitigation options under various sets of weather data. The final work in this Work Package was to quantify the importance of the most relevant effects involved with on-farm mitigation. A final evaluation of the various adaptation and mitigation options was made for a number of sites and of farming systems in cooperation with CP3 using the FarmAC model.

• WP9 – Farm scale modeling methodologies for mitigation and adaptation
WP9 was primarily a technical workpackage, designed to enable the greenhouse gas (GHG) budgets of the model and showcase farms to be estimated. It was originally intended that one or more existing complex, dynamic farm models would be used for this purpose. However, it became apparent that this was not a viable option because:
- the complex dynamic farm models available were too highly adapted to European conditions
- the quality and quantity of input and parameter data from the model and showcase farms, particularly those outside Europe, would not be sufficient to satisfy the demands of the complex dynamic models.
As a result, it was decided to develop a new model, specifically to allow the GHG budgets of a wider range of livestock farms to be assessed. The model was intended to be simple but became progressively more complex as the project progressed. This was a result of pressure from users, who found that the simple approach adopted to describe some processes led to incorrect or inconsistent responses. The final model contains the following elements:
- A static ruminant livestock production model based on energy and protein budgeting and an enteric methane emission model based on the IPCC (2006) Guidelines
- A static model of C and N emissions from animal housing, manure storage and grazing, based on the IPCC (2006) Guidelines and the EMEP/EEA air pollutant emission inventory Guidebook.
- A dynamic model of C and N transformations in fields, including gaseous emissions, nitrate leaching and C and N sequestration in the soil.

• WP10 – Integrating adaptation and mitigation at farm scale
Located at farm scale, WP10 aimed to select and describe most relevant livestock systems in the agroecological zones of the project, identify and use case study farms for experimenting practical measures and for discussion and testing in modelling, and integrate the adaptation and mitigation measures in the decision process of farm management. Further it aimed at extending the spatial scale of the classical farm to include issues such as animal mobility that are relevant to the regional scale. In parallel, an action was dedicated to the evaluation across the panel of selected farms of the main socio-technical barriers to adaptation and mitigation measures proposed in the project.
In the first year the WP identified 24 model farms and experts representing current and future livestock farming systems across EU and SICA partner countries (D 10.1). In addition, it contributed to a network of 20 real showcase situations. For each main agroecological situation and system, a list of preferred mitigation and adaptation options has been discussed and progressively built, maintaining for each situation the five main M/A options based on the outputs of WP8 ‘Integrating adaptation and mitigation options’. Surveys and discussion meetings have been organized with the panel of showcase and model farms to define the local relevance, the prioritisation of options and the parametrisation of on-farm M/A strategies. In preparation of modelling exercise, a compiled database on LCA coefficients has been delivered (D10.2). The modelling exercise took in itself an important and not really forecasted effort of collaboration inside CP3 to firstly agree on genericity of terms and develop (WP9 team) a simplified and original interactive “FarmAC” platform linked to major existing models routines (Fasset, Melodie, Farmsim), sufficiently generic to cover the data collection on the very large panel of situations present in the project. The interface has been used to compile the data and analyse, on a selected subset of farms, the effects of mitigation and adaptation options on GHG emissions and productivity parameters in different parts of the world (Europe, Africa, Brazil), with consistent results described in a report D10.3. The work focused then on parametrisation of initial computing routine to the very large set of situations. The large partner participation to the regular CP3 meetings was useful to secure that output from the simple model’ Farm AC’ as well as collect data needs for this model (WP9 and 10), which fits as far as possible with the economic MACC approaches (WP 11). While the approach doesn’t cover the whole range of the 24 farms identified during the project, it however simulates major and exemplary situations and clearly shows the proof of concept of the FarmAC tool to simulate a diversity of cases. Furthermore, the mid or long term modelling approaches to complete the analysis on the integration of M/A options and assess the effects of animal mobility on whole farm production, GHG emissions and soil carbon storage, the choice has been made to focus on the analysis of actual real data on a typical pastoral/sedentary meat sheep farming system set of farms in the French Mediterranean area. Results of the in-depth study are detailed in D10.4 and show the positive LCA effect of moving animals.
Regarding M/A technical and social limitations to adoption, an important activity among of the partners has been developed to collect data on M/A perceptions and barriers in the main agroecological zones of the AnimalChange project. A questionnaire has been shared for discussion and a large scale survey organised. The data collected on 176 farms on 14 locations in the three continents have been analysed. A detailed report showing the varying perceptions on CC and M/A has been provided to build the Deliverable 10.6.
Even if delays were important, most of the tasks’ slightly adapted objectives have been maintained. The CP3 meetings (WP 9-10-11) strongly helped to progress interactively between WP, tasks and partners. The important efforts made by AU and farm experts and the large steps on the operationalization of FarmAC were critical for the different tasks. The implication of the partners on implementing their data in the model and deploying the barriers survey on their farms were highly significant and largely encouraging regarding the general collaborative scope of the AnimalChange project.

• WP11 – Filtering out options
WP11 sought to provide the core economic analysis to define the efficiency (i.e. the balance of costs and benefits) of implementing different mitigation and adaptation options. As such, the WP had to initially draw on research into the technical effectiveness of measures that was largely conducted in previous WPs. The WP therefore has an important role in maintaining consistency between scientific and economic work plans. It was important that the initial science was based on some prior notion of measures that might be economically worthwhile. It also made sense for the economic analysis focused on measures that are actually technically (scientifically) feasible. Accordingly, much of the focus of the WP was on making sure that there is a common understanding of this objective and some consistency across WPs. This includes a clear view of what farm systems and data are actually being considered across WPs. The level of our ambition for this WP had to be revised to accommodate limitations in the farm scale modelling. The conclusion was that the project simply could not characterise all farm systems. Rather the project had to be pragmatic in the use of data that was already available to partners.
Because WP11 had one of the earliest deliverables (D11.1 M8) it quickly became clear that this deliverable should be used to scope the economic, social and political context of the consortium. This was important since the project meetings suggested a divergence of views on this topic, with no obvious deliverable elsewhere to set the context.
This objective was met, but with some minor deviation from the original plan, which suggested that the deliverable would already be able to specify region-specific mitigation and adaptation measures. However, it was immediately clear that the data and information to do this were not yet available from work to be conducted in previous WPs. Summaries of other tasks are as follows: 11.2 sought to clarify the application of MACCs, 11.3 defined adaptation measures and alternative adaptation approaches, 11.4 considered how mitigation and adaptation measures can be jointly screened to determine the nature of trade-offs and synergies in the application of measures simultaneously. All of these WPs have drawn on prior work conducted in WP 8,9,10.

Note that the tasks in this WP were apparently shared with a number of organisations. In practice the input of some of these partners was negligible (either not offered or not requested by SRUC). Although some partners were initially suggested as contributors at the time of project formulation, this was never subsequently agreed or confirmed. As an example, we have never confirmed arrangements with AgResearch. We made a fruitful arrangement with EMBRAPA for a researcher exchange to contribute to task 11.2. As things stand, the ‘passive’ role of these partners was not a problem from the delivery of stated outputs.

• WP12 – Exploring adaptation and mitigation options at a regional scale
WP12 is combining information from field/animal level, farm level and regional/global level to design adaptation/mitigation combinations suited to the levels of vulnerability and mitigation potentials in the main regions/systems included in the project and to assess the likely effects of implementing these mitigation and adaptation options.
WP12 provided a comprehensive assessment of exposure to hazardous climate conditions for different livestock production systems in different regions. By combining this assessment with a broad differentiation of sensitivity of these systems to such hazards we have been able to give general recommendations for adaptation planning. By incorporating the full set of projections from different climate models we were able to point out where projection uncertainty poses a specific challenge to adaptation planning.
We modelled some ’packages of options’, or so called, specific interventions, in selected livestock production systems in both East Africa and Europe. The scope of this modelling exercise was to highlight whether there was potential for improving producion and at the same time reduce green house gasses (GHGs) emissions. For East Africa the following options were modelled: improved animal husbandry and health (improve fertility, reduce mortality rate); improved feed quality (‘processing crop residues and adding maize to the ration’); improved grassland management (‘improved grazing management’; ‘increase legumes in grassland’). For Europe we modelled: feed supplementation (’feeding more fat’); manure management (’anaerobic digestion and biogas production’); improved grassland management (‘improved grazing management’ and ‘increase legumes in grassland); energy use efficiency.
All the tested options were be promising in terms of reducing emissions and increasing productivity. We therefore provided the evidence-based potential of selected options to address both an environmental footprint and food security simultaneously. However, the main drivers of GHGs emissions and the applicable technical interventions to address them, highly differ from developing and developed countries. The uptake of these options could potentially increase resilience of farmers through improved livelihoods. However, there are several additional actions that would concur to reduce the vulnerability of farming communities to climatic variability and climate change, among the most important ones are: asset and livelihood diversification; drought cycle management; index-based livestock insurance; and safety nets.

• WP13 – Economic and social analysis of mitigation and adaptation options with climate constraints
Overall objectives of WP13 were to generalise mitigation/adaptation information including barriers to a regional scale (emerging from largely farm scale data). To do this a number of case study regions were explored to generate data and provide a basis for understanding regional differences with respect to behavioural, institutional and biophysical constraints.

In terms of specific tasks:
T13.1 this task provided a representation of mitigation potential at regional scale (EU dairy pasture restoration in Brazil, legumes in France) as well as providing an overview of alternative methodological top-down and bottom-up approaches to estimating this potential. The deliverable was delayed for 3 months.

T13.2 this task was split in terms of focus on three separate case study elements that are geographically bound. The first element (UPM) focuses on the interface of climate change (responses) in the context of a water-stressed environment (southern Spain). The issue here was the interface between water policy and climate. This has been published as a paper in Environmental Science and Technology. UPM also have conducted work in in Kenya, with support from ILRI. This work was presented in Edinburgh in an international conference on Sustainable Intensification. A third case study focused on farmer behaviours and responses to climate change, a catchment study in Scotland. This work has already generated three published papers on behavioural and institutional barriers and a further paper has been submitted focused on uptake of a number of MACC based mitigation measures.

A final year PhD student from UPM spent 3 months at SRUC working with partners to develop a behavioural modelling approach to understanding mitigation and adaptation within Sub-Saharan Africa. These approaches have been in response to the identified disjoint between the messages from place-specific case studies and more general regional applicability. This is relative to the task being just about methodological development. This issue was raised in relation to the link through to WP 14.
T13.3 this task aimed to develop a flexible model to predict regional measure uptake. The model had to incorporate behavioural rules encountered in 13.2. Agent-based modelling has been developed for an intensive dairy catchment in Scotland but has been found somewhat impractical for regional application (i.e. extrapolation from the catchment study in 13.2). The framework on policy potential was applied to a number of country level data sets to measure policy potential. This has been developed into a paper which has been submitted to Environmental Science and Policy.

A meeting held in Amsterdam on 7th November 2013, which resulted in a revised strategy and work plan for work at regional scale in CP4. Given the time frame of CP3 and the actual expected outputs presented during the meeting, the strategy was revised and it was decided that modelling in CP4 should be independent from CP3 results. This has also led to a delay in the delivery of D13.4 which applied the modelling framework, identified in D13.3 to a number of candidate regions.

Significant highlights of the WP are the number of journal publications and the level of partner cooperation.

• WP14 – Evaluating policy options for adaptation and mitigation
The objectives of WP14 were to support the policy making process through the implementation of analyses to provide reliable information and the direct interaction with policy makers, ensuring that constraints and objectives arising from the policy arena are fed into the analytical work and that the technical results directly feed into the policy making process.
The information produced in the WP is relevant for quantifying the ‘policy potential’ of measures. The technical mitigation potential of options assessed at different levels in the project (animal/field, farm and region) is theoretically higher than the policy potential because of constraints on producers and governments, institutional, behavioral and economic.
First, case studies were developed to explore how to achieve mitigation in practice in different production systems and regions in the Global Livestock Environmental Assessment Model developed by FAO. They evaluate the mitigation potentials of specific technical interventions in selected production systems and geographical areas. Results indicate that the mitigation potential exists for all systems and regions, that it ranges between 14% and 41% of baseline emissions (2005) and that a large number of options also result in increased productivity and production. Results are presented in D14.4 and D14.2.
A review of policy options available and the development of 3 policy “storylines” for the EU were elaborated with and presented to the members of the Policy Committee and the Stakeholder Platform of AnimalChange during 4 meetings over the project duration. They are presented in D14., together with 2 scenarios for grassland intensification in Brazil.
The main factor considered for the selection of the mitigation policy scenarios in Europe was the cost-effectiveness and the levels of compensation of the implementation cost. Under scenario 1 (obligation to implement all considered mitigation technologies up to their feasible limits without financial compensation) total emissions from agriculture are reduced by 10% in EU27, with a small shift from milk (-1%) to beef (+2%) production, a slight increase in poultry and decrease in pigs. Under scenario 2 (full compensation), there is a net gain for those farmers who bear lower implementation costs. Emissions decrease less than in scenario 1, because animal herds and production is growing. Emissions from agriculture decrease by 7 to 8%, while the dairy cow herd increases by 1 to 2%, the beef cattle herd by 5 to 6%, the pig herd by around 2%, and the poultry herd between 2 to 5%. Under scenario 3 (famers are paid 50 EUR per ton of CO2eq reduced), achieved emission reductions are lower than in scenarios 1 and 2, amounting to 4 to 5%. The dairy cow herd reacts only slightly with a reduction of 1%, while the beef cattle herd might go down by around 2%.
The potential of rangeland intensification for greenhouse gas mitigation in the global context was explored by Cohn et al. (2014) with GLOBIOM. They considered two types of policies – a payment to the farmer per hectare of semi-intensively managed rangeland (subsidy) and a tax on the farmer per hectare of extensively managed grassland – with the same objective to increase land productivity and hence avoid deforestation. Results show that unilateral policies, in particular in regions where livestock production is extremely GHG intensive compared with the rest of the world but connected to it through international trade, have great potential to be effective and lead to overall emissions reductions. Results are presented in D14.4 and D14.2.
Finally, a methodological framework for assessing adaptation interventions in livestock production systems in the dry lands of Africa was developed, considering that these systems are among the most exposed ones to climate change and that they provide livelihoods to a large number of the poor. The framework, developed by FAO, the World Bank, ILRI, IFPRI, CIRAD and Action Contre la Faim, and illustrative results are presented in details in a draft paper included in Annex 3 of this deliverable. This work also identifies data and methodological caveats for assessing resilience in livestock production and illustrates that livestock can act as a buffer of climatic variability.
Results of the analysis and the policy dialog developed in WP14 were used to produce a White Paper (D14.3) “Key EU policy issues for livestock and climate change”, with contributions from the Executive Committee of AnimalChange in collaboration with the Policy Committee and the Stakeholder Platform and from project participants. It consolidates the policy-relevant results of the project and proposes a vision for climate smart livestock. It moves from an assessment of current initiatives to a staged proposal of what can be done, in the short and longer term. Potential roles for main stakeholders are identified, including the EC, national and local governments, private sector and academia. Key results from AnimalChange underpinning the paper are presented in boxes.

• WP15 – Data Infrastructure and Model support
This workpackage has established the data infrastructure of AnimalChange and facilitated the link between other project components. It has received input data from partners and provided output data to the scientific community and to stakeholders. This work has required a clear understanding of the data involved in the project. After a first step in which two documents describing in detail the simulation models and the experimental sites implied in AnimalChange have been produced and made available on the web site, the web interface, and the database receiving the data have been developed and made functional. Attention was paid to flexibility during the development phase, because of the high risk of specific needs or modifications occurring at later stages. The web interface and the database receiving the data are functional. Moreover, the data infrastructure of the project was established through a series of interconnected databases, linking the activities run in the other project components, and thus facilitating the flow of outcomes to the scientific community and stakeholders.
A series of methods and tools were also developed and provided to partners to facilitate the use of simulation models. In particular, a framework for model calibration based on Bayesian methods was developed in the forms of R-language scripts for coupling to simulation models (Ben Touhami, 2014, PhD Thesis), supplied together with practical guidance (also with the support of training and scientific sessions offered to partners). Based on this framework, updated parameterizations of grassland models were generated using dedicated grassland manipulation experiments (precipitation reduction) and long-term observational flux data from European grassland sites (Ben Touhami and Bellocchi, 2015, Ecological Informatics, submitted). Preceded by a sensitivity analysis to identify the most relevant parameters in grassland systems (Ben Touhami et al., 2013, Ecological Modelling), the application of the Bayesian framework has allowed reducing uncertainties in the simulated production and GHG outputs obtained with the model PaSim in European grasslands, generated by WP5 (without adaptation-mitigation options) and WP8 (with adaptation/mitigation options).
The integration of data-driven and theory-based information was completed by the development of quantitative approaches and tools for vulnerability assessment to climate change. A set of indicators for vulnerability analysis was produced and evaluated in a range of conditions (Lardy, 2013, PhD Thesis). In Lardy et al. (2014, Environmental Modelling & Software), a generic quantitative approach was illustrated to evaluate vulnerability to climate change, with an application to soil organic matter in French grassland soils. The conceptual approach was accompanied by a numerical evaluation of vulnerability (meta-modelling built on mechanistic modelling of grasslands with the model PaSim), which has shown that soil organic matter can increase in the future in French grasslands. The environmental vulnerability would therefore decrease, yet likely be accompanied by increased inter-annual variability. Based on these studies, dedicated Java-based tool for vulnerability assessment was developed (Lardy et al., 2015, Computers and Electronics in Agriculture, submitted) and supplied to partners (together with practical guidance and the support of training and scientific sessions), which was applied to vulnerability studies by WP5 and WP8. The tool for vulnerability assessment has been the basis for the development of an integrated decision support system for vulnerability studies, in the form of generic model-based platform (Eza et al., 2015, Ecological Informatics, submitted).


Potential Impact:
• WP2 – Climate and socio-economic scenarios for livestock systems in project regions
The newly developed global livestock production systems dataset opens the possibility to view the livestock sector in unprecedented detail, but still in a consistent framework. The dataset has been shared with several researchers worldwide which see it as the missing piece in a wide range of studies they planned to do. Many global economic modeling teams outside of the project requested access to the dataset. Its adoption by these teams will change the way how livestock sector is modeled.
The GLOBIOM-CAPRI multi-scale model cluster further developed in the course of this project is now being applied by the European Commission for assessment of EU policy strategies with respect to climate change mitigation and adaptation. It will be further developed and expanded for new functionalities in the future; for instance in the framework of the H2020 SUSFANS project dealing with future challenges to EU food security.
The quantified scenario drivers based on the Shared Socio-economic Pathways (SSPs) benefited strongly from the new IPCC scenario development process. But the IPCC scenario process strongly benefited also from the specific developments in AnimalChange, because the quantified scenario drivers are through GLOBIOM used also in the IPCC mitigation scenario quantification. In addition, the Impact, Adaptation, and Vulnerability (IAV) community expressed the need for development of so called Representative Agricultural Pathways (RAPs), and the AnimalChange work has a great potential to contribute also to this process.
The scenario results themselves start to feed into reports likely to inform the policy making processes outside of the EU. As an example, we can mention the African Livestock Futures report cosponsored by the office of the Special Representative of the UN Secretary General for Food Security.
Last but not least, our scenario results provide new insights about the substantial potentials for mitigation of and adaptation to climate change coming from the optimization of livestock production systems structure. Adding this new layer of detail to the discussion of the livestock sector future in the context of climate change will lead to new and sometimes surprising conclusions.

• WP3 – GHG emissions from livestock sector: reducing key uncertainties
The experimental results and the modelling approaches and results that have been developed in WP3 offer benefits to a variety of end-users; these include both sector specialist and farmers, food and livestock supporting (feed) industry, local, regional and national and EU policy makers.
The uncertainty quantification of estimates and calculations of greenhouse gas emissions has potential to direct further work on data sampling and setting up statistics and surveys. Our analysis can be used to determine how to contribute to reducing the uncertainties in emission inventories. And to allow these emission inventories to become region or continent specific. As such these further efforts in data and statistics will likely be more cost effective when based on the uncertainty analysis in AnimalChange.
Identifying and reducing the uncertainties will likely contribute to making the sector involved (e.g. agriculture and livestock farming and farmers) less restrictive in committing to greenhouse gas emission projects and targets as impact and effect are more clear than before. Incentives and rewarding mechanisms can be put in place with greater confidence by either policy or industry.
Our rigorous assessment of the greenhouse gas emissions across regions and continents based on the improved Miterra Europe and Global framework now give access to emission profiles for sectors, subsectors in the livestock production and specific groups of products as well. This profiling does identify where in the livestock production and products major emissions are and what actions would help reduce these emissions.
Our models have made now available data and process understanding that would be useful to further develop greenhouse gas emission reporting and monitoring schemes and approaches. This will require to fil existing data gaps, improve understanding on selected processes contributing to emissions of greenhouse gases and derive values for emission factor values (tier 2) or agreement of application of process based models for calculation of greenhouse gas emission for inventory preparation.

• WP4 – Climate change impacts on livestock systems: reducing key uncertainties
The experimental results and the modelling approaches developed in WP4 can be beneficial to a variety of end-users.
The role played by legumes and mixed-species swards to face drought events can inform subsequent studies and are relevant to decisions in the context of wide adaptation strategy options applied to grassland-livestock systems. The information provided by the maps of geographical locations of animal disease vectors (under climate change conditions) are readily exploitable. This is also true for the set of solutions identified (e.g. vermicompost) to reduce parasitism in grazing animals at tropical sites.
The agro-environmental modelling community would also benefit from both the developments in modelling approaches and improved parameterization for large scale simulations of crops and grasslands at a variety of contrasting sites.

• WP5 – Climate change vulnerability of livestock systems
This WP provides detailed maps showing the changes in productivity of grasslands, and to a lesser extent of feed crops, under various climate change scenarios. Moreover, it shows the risks associated to changes in climatic variability in a spatially explicit way, by indicating the risks created by heat and drought events. By expressing some of the results directly as number of grazing days per year and as potential livestock stocking density, the results can be of direct interest to both private and public stakeholders of the sector. The information provided by these maps is readily exploitable. They can be used to develop planned adaptation both in regions with increased opportunities and in regions with increased vulnerability under contrasted climate change scenarios.

• WP6 – Breakthrough biophysical mitigation options at field and animal scale
WP6 has:
- provided clear evidence of the ability of soils to sequester carbon and how this is affected by different grazing and management regimes.
- quantified the reductions in greenhouse gas emissions that can be achieved by increasing the use of legumes in grazed pastures.
- established an online updatable database of experiments carried out to decrease enteric methane emissions and provided online tools to allow user-driven meta-analysis to be carried out.
- demonstrated and quantified the usefulness of nitrate as a dietary supplement to decrease enteric methane emissions by dairy cows.
- established that methane production in the rumen can be inhibited with no apparent deleterious effects on rumen fermentation.
- established a unique database of the effect of tannin and saponin rich plants from Africa and South America on rumen fermentation and methane production.

• WP7 – Breakthrough biophysical adaptation options at field and animal scale
The size of the beneficial effects of legumes was unexpectedly large (size of impact), covered adaptation and mitigation options to climate change, addressed sustainable intensification of livestock systems (multifunctionality of impact) and was evident under highly diverse climatic conditions from Africa to the whole of Europe, from Mediterranean to Arctic climate (geographical scale of impact). It is of primary interest that the use of legumes in these systems represents a ‘ready-to-use’ technology. It is also a success of this WP that the adaptation of animal production to heat and parasites yielded ready-to-use solutions to be applied in practice.
Since this WP was dealing with the field and the animal scale, the developed technologies are primarily of beneficial use to the farmer / manager of livestock enterprises, to the industry developing and trading production resources (e.g. seed market) and to extension services. Since the use of functional biodiversity (i.e. mixtures instead of monocultures) is a highly topical issue in ecology and agriculture, the knowledge gained on ecosystem functioning and the databases developed for modelling approaches are of high value for the scientific community.
According to these multiple end users on target, dissemination strategy was diverse. An exceptionally rich output in scientific papers in highly ranked international journals were a cornerstone of dissemination. A second key focus was on dissemination to practice via newspaper articles, teaching in universities, colleges and to extension specialists as well as via demonstrations on field days for farmers and the industry.

• WP8 – Integrating adaptation and mitigation options
The results of this WP highlight the importance of integrating adaptation and mitigation options at the field/animal scale when studying strategies for specific regions and livestock production systems. Many trade-offs occur between GHG emissions of animal, manure or soil origin. The size and direction of these trade-offs (or sometimes synergies) are intricately related to farm management and several important trade-offs have been identified that would require attention when addressing mitigation on (grass-based) livestock farms under climate change. The biochemical, biophysical, biogeochemical process-based models representing the mechanism involved with GHG emissions are in principle well suited to delineate the effects of mitigation measures on GHG emissions, and to quantify the background and variability of potential trade-offs. They should best be applied when improving methodologies for farm surveys or GHG inventories on accountability of important sources of variation in GHG and in size of trade-offs in particular (in contrast to continuing the use of generic higher Tier approaches). Results of WP8 were widely spread to the international scientific community and presented at many meetings. The qualitative overview of mitigation and adaptation options was used in many WPs as a starting point for further modelling. Results were also presented in stakeholder meetings.

• WP9 – Farm scale modeling methodologies for mitigation and adaptation
The FarmAC model has proven to be useful for the quantification of production and C and N flows on a wide range of European, South American and African farms. In a number of situations, the calibration of the model highlighted a lack of empirical data and a lack of understanding of some processes driving production and GHG emissions. In the longer term, the FarmAC model is likely to be of particular use in non-European situations, since it combines a relatively simple yet flexible approach to modelling livestock production and emissions from manure management, with a more complex, dynamic approach to modelling C, N and H2O flows in the soil. This is because for low input systems, the dynamics of organic matter in the soil and the fate of crop residues are of particular importance.

The domain name www.farmac.dk has been acquired and this will form the focus for future work. Aarhus University is committed to continuing support for the model, now that the project has been completed.

In order to disseminate results, training sessions for FarmAC users were organised in Paris November 2012 and in the Netherlands in January 2013, January 2014 and October 2014. In addition, AU participated in a regional workshop in February 2015 in Brazil. Extensive email and Skype consultations were undertaken during the period November 2012 to February 2015.

• WP10 – Integrating adaptation and mitigation at farm scale
In WP10, the large participation (12) of the partner institutions in the regular meetings as well as their on-field work with local stakeholders brought representations from all the agroecological zones and contributed greatly to dissemination of concepts on the M/A questionings when levelled at whole farm management scale. Providing real case and/or model farms, formalising data in the FarmAC, selecting options for M/A, modelling at different extent their own farms, surveying farmers on their perceptions on CC, animal impacts, mitigation adaptations, are numerous actor activities in WP10 that participated in sharing and improving scientific knowledge, disseminating ideas and open minds on possible future in the local livestock systems. It is particularly the case in SICA countries (North and Sub Saharan Africa, Brazil), where adaptation in a major concern and where mitigation is often considered a concern of highly developed countries. Development of the AnimalChange project under this task was helpful to foster the need for data, methods, models and whole farm simulation to argue on the effectiveness of mitigation options. The training organised in Dakar (Senegal) associating North African countries was a good example of dissemination.
Further project ideas such as the impact of animal mobility in mitigation and adaptation have been generated for national and international calls upon the first results of the project.

• WP11 – Filtering out options
WP11 has made a significant contribution in terms of regional training workshops :
Cost benefit analysis of mitigation strategies: win-win solutions; EU FP7 Project – AnimalChange Workshop on Livestock and Climate Change; attended by policy makers and researchers http://www.animalchange.eu/Content/Nairobi2015.html
2-3 February, 2015 Nairobi, Kenya

Cost benefit analysis of mitigation strategies: win-win solutions; EU FP7 Project – AnimalChange Workshop on Livestock and Climate Change attended by policy makers and researchers http://www.animalchange.eu/Content/Budapest2014.html
30-31/10/2014 Budapest, Hungary

Regional workshop Curso em Produção Animal e Mudanças Climáticas / Curso de Producción Animal y el Cambio Climático / Training course on “Livestock and Climate Change” in Latin America (February 2015)

• WP12 – Exploring adaptation and mitigation options at a regional scale
The results of this WP highlight the importance of targeting mitigation strategies to specific regions and livestock production systems. Interventions should be carefully designed and targeted, and this requires a better understanding on how potentially up-scale the results obtained from individual trials. In addition, we illustrated how important, firstly an understanding of adoption rates are and secondly the role of public and private sector policies in supporting the adoption of these options by targeting solutions to current barriers of adoption and foster adoption of those strategies that increase risks and costs to farmers.
Results of the activities undertaken in WP12 have been presented on the occasion of regional trainings, workshops and conferences and disseminated through reports, peer review publications, seminars, blogs, etc.

• WP13 – Economic and social analysis of mitigation and adaptation options with climate constraints
The significant highlights of the WP are the number of journal publications (13 aknowledging AnimalChange support) and the level of partner cooperation.
This WP highly contributed to the impact of the project on the sustainability and competitiveness of livestock systems by showing the potential for:
- Reducing greenhouse gas emissions from European livestock systems using cost effective methods,
- Reinforcing economic and environmental competitiveness of animal production systems under climate change and increasing their resilience to climatic variability.


• WP14 – Evaluating policy options for adaptation and mitigation
WP14 provided policy support both through the implementation of analyses to provide reliable information and the direct interaction with policy makers, guarantying an optimum uptake of findings in the policy making process, both at EC level and at member state level. The permanent link with the Stakeholder Platform also ensured that results from the project reached out to representatives of the industry and the NGOs and that in return their inputs fed into the policy making process. The policy White Paper aims at consolidating policy relevant results from the project with a vision for climate smart livestock. It was presented during various regional workshops and conferences and will potentially be largely disseminated.

• WP15 – Data Infrastructure and Model support
The tools developed in WP15 (database, Bayesian calibration scripts, tools for vulnerability assessment) can be beneficial to a variety of end-users.
The database allows each partner to both store their own data linked to AnimalChange activities and get data, with the agreement of the data owner.
The agro-environmental modelling community would benefit from both the scientific outcomes and the artifacts (generic Bayesian calibration tools) produced by WP15 (and supplied for free to interested users). It was shown that Bayesian calibration can help reduce uncertainties associated with model parameters by updating a distribution function from prior knowledge yet under altered climate conditions. These results can inform subsequent large-scale calibration efforts or uncertainty analyses.
The tools for vulnerability analysis (freely supplied to interested users) hold potential in terms of usefulness for science-stakeholder dialogue, as a support to present and disseminate results. In particular, an approach based on synthetic (yet simple) indices can help in easy communication of vulnerability analysis results to stakeholders of different nature.

COMMUNICATION AND OUTREACH
• WP16 – Knowledge Interaction, Training and Dissemination
The overall objective of WP16 is to make an analysis of the research and technical outputs of the project, and to ensure that they are effectively disseminated among the following stakeholder groups: i) animal farmers, ii) animal production industry including extension services, iii) consumers; iv) NGOs, v) scientists, v) policy makers. Specifically the component aims to:
- Disseminate AnimalChange outputs within the EU and globally in particular to developing countries both in Africa and S. America, in the African Mediterranean Partner Countries and South-Eastern European countries using stakeholder representatives and multinational agencies
- Create a participatory framework that will allow meaningful dialogue between the Work Package leaders and the stakeholders to ensure that the project meets their needs.
- Organise regional workshops in developing and disadvantaged countries and a final conference to enhance the relevance of the research to stakeholders and to agree on future implementation strategies
- Provide innovative training elements to take into account the wide nature of stakeholders both in terms of geography and existing levels of capability
- Improve inventories in non-Annex 1 participating countries so that baseline inventories and the impacts of mitigation and adaptation efforts can be more easily traced.

A stakeholder platform of key end users was formed to create a participatory framework linking key end users to the project team. It comprises representatives of European organisations such as Copa/Cogeca, FEFAC; International bodies such as the International Dairy Federation and the International Meat Secretariat and also NGOs such as Vétérinaires Sans Frontière, Farm Africa and IMAFLORA (S. America). The platform met with the project team in Brussels on 21st September 2011 and was briefed on the aims and outcomes of the project. They subsequently produced a report of their recommendations in relation to their role, their expectations of the project outputs and on the process of consultation between the project team and the stakeholder platform. A dissemination and training plan was produced based on the description of work in WP16 but importantly incorporating the above recommendations of the stakeholder platform. It also includes key actions to implement the plan during the first 18 months of the project. It was updated regularly following annual meetings between the stakeholder platform and the project team. The platform also met with the project team in the Annual Meeting in Dublin in June 2013 and in Madrid in May 2014. They subsequently produced reports of their recommendations highlighting the good progress of the project but pointing out the need for more collaboration between workpackages. Following Madrid, they confirmed that the project was making an important effort to contributing to more climate-friendly farming and to productivity growth in order to meet the world food demand, but they are concerned about the ability of some tasks to meet deadlines in the final period, and there was not enough focus on Africa and L. America and smallholders. The stakeholders also attended two sessions of the WP14 Policy committee and contributed to their recommendations.

The project has held 5 scientific symposia. The first was at EAAP Stavanger 2011 jointly with FP7 project SOLID and featured Lifecycle Assessment of Livestock Production. The second symposium on ‘Livestock and climate change: options for mitigation and adaptation’ was presented at EAAP Bratislava in late August 2012. A session on “Adaptation” was organised at the Greenhouse Gas and Animal Agriculture conference in Dublin in June 2013 and the project contributed to the conference on “Sustainable Intensification- The pathway to low carbon farming” in Edinburgh, UK in September 2013 and in particular a session on “Sustainable Intensification Platforms and Animal Change”. All the presentations are available on the project website www.animalchange.eu. Finally the project organised a conference on “Livestock, Climate Change and Food Security”. This was held in Madrid, on the 19th & 20th of May 2014. A summary of the presentations was presented in an abstract book that contains the 114 offered abstracts and 13 of the invited keynote speakers as can be found on https://animalchange.wordpress.com/abstract-book/.

Two leaflets were produced. A project brochure outlined the aims of the project, the structure and the main outcomes and their likely impact. In addition a vision document highlighting the outcomes and impact has been produced at the request of the Scientific Advisory Board and the Stakeholder Platform. These leaflets have been widely distributed through the database and at all conferences and symposia organised by the project and at other relevant meetings. They are also available to download from the website.

The project organised five regional workshops in Hungary, South Africa, Senegal, Kenya and Brazil that targeted a wide audience from industry, extension services, NGOs and policy. The aim of the workshops was to disseminate the project outputs that were relevant to the above regions and to discuss future priorities for research and for implementation of project outputs. The presentations from the symposia and workshops are available on the project website (http://www.animalchange.eu).

The development of e-learning courses was initiated and target audiences, content and teams have been agreed upon. Attendance by FAO and EAAP at Global Research Alliance (GRA) workshops in Nairobi (joint with ILRI) and Accra has clarified the needs of the target countries and regions in the project for national inventory support. The most efficient means of achieving the outcomes are a) to link with the Livestock Research Group initiatives in this area within the GRA and b) to focus the support as part of the regional face to face training events being organised by the project. Four face-to-face training courses were delivered in S. Africa, Hungary, Senegal and Brazil. Young scientists were supported by the project to attend these meetings. To give wider access, 39 web based e-learning materials in a variety of formats were developed. These courses, which will be available through the project website, will improve capacity and enhance the implementation of project outputs.

Three categories of actions to reduce the skill and technical shortages were delivered to allow improved inventory methods to be adopted particularly in non-Annex 1 participating countries. The project developed a scientific network to improve emission factors and activity data bringing existing networks together to improve communications across the regions, improved links between science and policy to transfer knowledge and develop feedback loops between science and governments, industry and the private sector and training and capacity building on measurement techniques to improve emission factors.
Discussions with the stakeholder platform and the scientific advisory board together with views from the regional workshops led to a definition of:
• the key outputs from the project that are of relevance to the stakeholders,
• their likely impact and
• a strategy for the dissemination and implementation of the outputs post project

The outputs of immediate relevance to stakeholders are available on the project website.

List of Websites:
Project website address: www.animalchange.eu

To support dissemination and to host the web-based training courses, an open public website was created (www.animalChange.eu). This informs on the aims and structure of the project, project outputs, proceedings of project symposia/workshops and news of the project and forthcoming conferences. Public deliverables are also published on the website. A project brochure was produced outlining the aims of the project, the structure and the main outcomes and likely impact. It has been widely distributed at conferences and is also available on the website. In addition a vision document highlighting the outcomes and impact has been produced at the request of the Scientific Advisory Board and the Stakeholder Platform. Two external newsletters have been produced highlighting particularly important results of AnimalChange research.