INNOVATION IN INTERDISCIPLINARYMETHODS - THE RELU EXPERIENCE

RURAL ECONOMY AND LAND USE DATA SUPPORT SERVICENOVEMBER 2011

First published November 2011

Editor: Veerle Van den Eynden

Published by:Relu Data Support ServiceUK Data ArchiveUniversity of EssexWivenhoe ParkColchester, CO4 3SQ

ISBN:  1-904059-84-8

Designed and printed byPrint Essex at the University of Essex

© 2011 University of Essex

The Rural Economy and Land Use (Relu) Programme hasbeen not only an important research programme in itsown right, producing significant results and impact, but alarge-scale experiment in how to do science.Interdisciplinarity was fundamental and this posedparticular challenges, both in the design andimplementation of projects, and in managing andintegrating data.

The researchers have risen to these challenges in novelways, experimenting creatively with their methodology.Also the setting up of the Relu Data Support Service toallow interdisciplinary data management and archivinghas provided an important model for future programmes.

Thus, during the course of the Relu programme we havebeen able to develop and learn about the mechanics ofdata integration and create a legacy, not only ofsubstantive findings, but of methods that can be appliedin future research.

Jeremy Phillipson, Rural Economy and Land UseProgramme

INTRODUCTION 1

COMMON UNDERSTANDING 3

STAKEHOLDER ENGAGEMENT 6

QUANTIFYING 11

MODELLING 16

VISUALISATION 20

CONTENTS

REACHING OUT ACROSS RESEARCH DISCIPLINES,ENGAGING STAKEHOLDERS, GAINING KNOWLEDGE,STUDYING COMPLEX REALITIES.

1INTRODUCTION

Studying the complex and pressing problems Relu aimed toaddress - the social, economic, environmental andtechnological challenges facing rural areas in Britain -needs an interdisciplinary approach to consider all angles1.The research focused on sustainable food chains, themanagement of animal and plant diseases, adaptation toenvironmental change, and integrated land and watermanagement.

Equal importance was placed on researchers workingclosely with and engaging stakeholders, to ensure researchwas grounded in the realities of the countryside; and for theresearch findings and outcomes to be policy-relevant andapplicable to rural areas and the people who live there.

The Programme required that in all research projects itfunded, social and natural scientists would work togetherto investigate the chosen topic. The research teamsthemselves decided and developed which interdisciplinaryapproaches and methods they would employ.

The Relu Programme was funded by the Economic andSocial Research Council (ESRC), Biotechnology andBiological Sciences Research Council (BBSRC) and NaturalEnvironment Research Council (NERC), the ScottishGovernment and the Department for Environment, Foodand Rural Affairs (Defra).

INTRODUCTIONFROM THE START OF THE RURAL ECONOMY AND LAND USE (RELU) PROGRAMME,INTERDISCIPLINARITY WAS SEEN AS A KEY PRIORITY FOR THE FUNDING RESEARCHCOUNCILS AND AN ESSENTIAL APPROACH FOR THE STRATEGIC RESEARCH THE PROGRAMME WOULD UNDERTAKE.

1 Lowe, P. and Phillipson, J. (2006), Reflexive Interdisciplinary Research: The Making of a Research Programme on the Rural Economy and Land Use.Journal of Agricultural Economics, 57: 165–184. doi: 10.1111/j.1477-9552.2006.00045.x

WHY INTERDISCIPLINARITY? THE RESEARCHERS’ VIEW

“Scientific understanding of plants and insects and theirinteraction was essential to the expertise of our project,but it was also necessary to draw on the knowledge ofpolitical scientists of regulatory processes and the way inwhich their design and implementation is affected bystakeholders, including private governance initiatives byretailers.”Wyn Grant, Warwick University, The role of regulation indeveloping biological alternatives to pesticides

“The issue of ‘quality’ within food chains cannot beinterpreted solely from natural or from social scienceperspectives. Identifying the linkages and networkedpractices that join the ecological quality of grassland tothe feeding patterns of stock animals; the nutritionalproperties of grasses to the perceived taste andmetabolic capacities of meat; the societal imaginaries ofhigh landscape value farming with the economic realitiesof upland husbandry; all required an interdisciplinarypractice and a lively process of knowledge creationthrough which socio-natural entities were defined andexplored.”Henry Buller, University of Exeter, Realising the linksbetween quality food production and biodiversityprotection

“Ecologists think of communities in different ways tosociologists. To an ecologist a community is anassemblage of populations of different species,

interacting with one another within a sharedenvironment. Species interact by competition (lose/lose),predation (win/lose) and mutualism (win/win), governedby environmental circumstance. To a sociologist, acommunity is a group of diverse people living within adefined geographical area. Individuals interact alongsimilar lines as species (competition, predation,mutualism), but are governed by mutually agreed rules,shared interests and shared beliefs. Ecologists andsociologists then need to link their differentunderstanding of ecological and sociologicalcommunities.”Jeremy Franks, Newcastle University, Collaborativeconservation in agri-environment schemes

“The challenge was to capture the knowledge of localmanagers and utilise this in a model of deer distributionto generate predictions that fitted the observed data.” Justin Irvine, James Hutton Institute, Collaborative deer management

Between 2004 and 2012, Relu funded a total of 94projects. The projects engaged around 4,000stakeholders and the programme forged close networkswith key individuals and organisations. Relu broughttogether researchers from 40 distinct disciplines fromacross the social, environmental and biological sciences,ranging from anthropology to water engineering andsociology, with the largest numbers coming from ecology,economics, human geography, environmental modelling,sociology, hydrology and crop sciences. Relu teams allincluded natural and social scientists and mergedqualitative and quantitative disciplines.

There is now a unique opportunity to analyse and reflecton the variety of innovative interdisciplinary methods andapproaches to data integration that emerged. Specifictypes of approaches and methods can be distinguished,with many projects applying a combination of many suchapproaches.

The research data from Relu projects have been archivedat the UK Data Archive and the Centre for Ecology andHydrology’s Environmental Information Data Centre andare available for future research. Research outputs andpublications are held in the ESRC research catalogue. Dataand outputs can be explored and accessed via the Reluknowledge portal.2

2 INTRODUCTION

2 Relu knowledge portal: relu.data-archive.ac.uk

3

COMMONUNDERSTANDING

3

INTEGRATING DISCIPLINARY RESEARCH PERSPECTIVES AND METHODS

This was frequently realised through discussion, reflection,shared evaluation of findings and joint publications.Researchers developed common frameworks for theresearch and jointly designed and adapted methods.Some researchers took this even further and engaged in

cross-disciplinary activities, swapping roles during datagathering or carrying out fieldwork together. This led tomore holistic assessments of problems, systems andtechnologies, and more rounded framing andinterpretation of core concepts.

MANY PROJECTS IMPLEMENTED A COLLABORATIVE APPROACH AMONG RESEARCHERSAND STAKEHOLDERS TO CREATE A MUTUAL UNDERSTANDING OF THE ISSUES STUDIEDAND OF EACH OTHER’S METHODS AND LINES OF THOUGHT.

4 COMMON UNDERSTANDING

CASE STUDYTHE TRADING ZONE OF MICROBIOLOGY AND POLITICS

More sustainable means of plant protection are neededthat enable food security whilst also protecting theenvironment.

Although chemical pesticides have often attractedcontroversy, viable biological alternatives have been slowto come onto the market. Commercial take-up has beenrelatively limited, in part because of the limitedavailability of suitable products. One reason for this isthat the regulatory process was designed to cope withsynthetic products and had difficulty in adjusting to thedifficult requirements for registering biopesticides.

Scientific understanding of plants and insects, and theirinteraction with and possible impacts on beneficialinsects, was essential to the expertise of this project. It was also necessary to draw on political scientists’knowledge of regulatory processes and the way in whichtheir design and implementation is affected bystakeholders, including private governance initiatives byretailers.

The microbiologists and political scientists engaged in theproject needed to have an understanding of each other’sdiscipline to facilitate interdisciplinary working. The twoliteratures were written in very different ways, withpolitical science literature being more discursive than thetersely written life sciences literature. Each discipline hadits own particular terminology. It was therefore necessaryto establish a ‘trading zone’ to facilitate understanding ofeach discipline’s methodologies.

Each group of natural and social scientists read selectedarticles from the other discipline relevant to the projectand then reported back to the next team meeting on theirunderstanding and interpretation of the article and whatthey thought its significant points were. This helped toclear up any misunderstandings about terminology andalso permitted a fuller understanding of the substantivegoals and methodological procedures of each discipline. Ithelped to create a shared interdisciplinary space in whichboth disciplines felt comfortable.

It facilitated considerably the greatest practical challengeof all: writing coherently and accessibly together,particularly for the book the project produced. It helpedthe team reach out successfully to diverse audiences ofstakeholders and academics.

The role of regulation in developing biologicalalternatives to pesticidesWyn Grant, Warwick University

5COMMON UNDERSTANDING

CASE STUDYSWAPPING ROLES AND MUCH MORE

The researchers set out to provide data and methods tosupport decisions by organisations with responsibility forthe management of floodplains as they seek to balancecompeting demands such as food production, natureconservation and flood risk management. Suchorganisations include government agencies, regionaldrainage organisations, farmers and their associations,and conservation bodies.

The project required continual integration of social,economic, ecological and engineering perspectives, withthe management of flooding as the core focus. The teamdeveloped analytical methods that combined biophysicaland ecological assessments, hydrological modelling,appraisal of engineering options, and economic valuationof outcomes. Interdisciplinary integration was advancedby close and shared working amongst researchers. Theyswapped roles during data collection and analysis, didteam field visits and jointly prepared and deliveredpresentations and publications to report findings.

Using eight floodplain sites in England, they carried outdetailed monitoring of floodplain water levels to constructa hydrological model that predicted water table levelsbased on rainfall and site conditions. Linked to this, thetolerances of different types of agricultural crops and‘natural’ vegetation species to seasonal flooding andwater table height were derived based on scienceliterature and site observations. This informed the analysis

of land use scenarios which prioritise different outcomes,such as food, nature conservation or flood control.The project developed a framework of 14 indicators ofecosystem services such as food production, carbonstorage, flood storage, biodiversity and recreation; whichclearly identifies the potential synergies and trade-offsamongst different benefits as land use changes.

Together with the Relu project on farming andbiodiversity, a novel stakeholder mapping tool wasdeveloped to assess the range of interests and influencesin floodplains. This demonstrates how stakeholderinterests tend to focus on particular ecosystem servicesand how the interactions amongst stakeholder interestsmight be managed so as to realise more value fromfloodplains.

Integrated management of floodplainsJoe Morris, Cranfield University

6

STAKEHOLDERENGAGEMENT

STAKEHOLDER EXPERTISE INFLUENCING THE RESEARCH PROCESS

Stakeholders were involved in data collection, groundtruthing research findings, modifying and validatingmodels and scrutinising findings. through theirinterdisciplinary teams the researchers developed new

methods of participatory research. ecologists said thatbeing able to work more closely with stakeholders wasone of the main benefits of working with social scientists3.

7STAKEHOLDER ENGAGEMENT

INTERDISCIPLINARY WORKING ENABLED CLOSER ENGAGEMENT WITH STAKEHOLDERS INTHE RESEARCH AND INCORPORATION OF THEIR EXPERTISE.

3 Lowe, P., Whitman, G. and Phillipson, J. (2009), Ecology and the social sciences. Journal of Applied Ecology, 46: 297–305. doi: 10.1111/j.1365-2664.2009.01621.x

CASE STUDYSTAKEHOLDERS GROUND TRUTH ECOLOGICALMODELS

Most kinds of natural resources are best managedcollaboratively. In a free-for-all, a resource is likely to beover-exploited with each user attempting to extract hisor her maximum benefit in the short term. But simplyknowing that collaboration is a good idea does notguarantee that it can be achieved.

Red deer management is an excellent case study toinvestigate collaboration because deer provide bothsocietal benefits and costs: wild deer are not owned byanybody but as they move around, they cross boundariesand provoke potential conflicts between neighbouringowners who have differing management goals.

Researchers have developed many models to predictwildlife use of habitats. These are often of little value forlocal management because their predictions do notmatch observations, largely because they do not takeaccount of the local management actions. For models tobe credible tools to develop collaborative solutions forwildlife management, they need to bring togetherscientific knowledge with the wealth of insights held bythose who manage these resources.

The challenge was to capture the knowledge of localmanagers and use this in a model of deer distribution tothen create predictions that fitted the observed data.The credibility of the deer distribution predictions in theeyes of managers would be enhanced, making it a usefultool to explore potential conflicts between neighbours orbetween local practice and national policy objectives.

The team developed a participatory approach tointegrate deer managers’ local knowledge with scientificunderstandings and ecological spatial data in a simpleGeographic Information System (GIS).

Managers’ knowledge on deer habitat use in relation toshelter and forage, together with local information onpaths, fences and habitat changes was used to change

the way in which the GIS developed deer distributionpredictions. The results fitted very well with observeddata and were much better than predictions from amodel based only on the existing scientific data. Theyclearly showed the value of using local knowledge.

This approach allows knowledge from different sourcesand at different spatial scales to be combined to giverealistic predictions of deer distribution. Suchparticipatory interdisciplinary approaches to wildlife-habitat models can improve communication andconsensus across ownership boundaries where differentmanagement objectives exist and can therefore removekey obstacles to collaborative natural resourcemanagement.

The team also explored the perspectives of researchersand stakeholders on the successes and challenges of thisway of working. Perceived benefits of a participatoryinterdisciplinary approach included improved socialnetworking, social and technical learning and academicachievements. Challenges included the time and cost ofintensive engagement, the building of relationshipswithin the constraints of the research project, meetingdiverse expectations and the difficulties of integratingdifferent forms of knowledge.

Collaborative deer management Justin Irvine, James Hutton Institute

8 STAKEHOLDER ENGAGEMENT

CASE STUDYPARTICIPATORY MODELLING TO SUPPORTCATCHMENT MANAGEMENT

Reductions in water pollution have so far been achievedmainly through regulation and investment in waste watertreatment, but the underlying water quality problem inmuch of the UK is diffuse pollution derived from currentand past land use, plus atmospheric deposition.

The project focused on how to improve the ecologicalquality of rivers and lakes and worked in two case studycatchments: Tamar and Thurne. A key proposition hasbeen that stakeholders with differing understandings andvalues must weigh up catchment management options.This requires a shared knowledge base and commonunderstanding of processes of water quality degradation.

The team developed a participatory modelling approachand built an Extended Export Coefficient Model in whichthey incorporated local knowledge. Stakeholders wereinvolved in developing the model and in testing andapplying it, through a series of workshops and eveningmeetings. Stakeholder analysis and ‘circuit riding’through face-to-face meetings and telephoneconversations by a social scientist, first built interest andtrust in the process. Continuity of engagement of keyrepresentatives of varied stakeholder groups wasachieved through the series of meetings. These followedan adaptive planning and management cycle ofvisioning, catchment and pollution characterisation,pollution source and pathway modelling, scenariodevelopment and implementation planning.

Graphical modelling of catchment processes was used toclarify expectations and create a shared understanding.In the words of a leading Thurne farmer: “After living andfarming in the area for so many years this diagram hasbrought home to me for the first time the importance ofthe pumps in the Thurne catchment and that otherwisesurface inflows are relatively insignificant. It does providea good means to capture local understanding of thecatchment.”

Whilst the mathematics of the modelling depends onexpert knowledge, key assumptions, sources ofuncertainty and limitations were all scrutinised bystakeholders. The local knowledge of stakeholders, inparticular farmers, was essential to ground truth data likeagricultural census data. This knowledge also allowed theinclusion of the impact of farmer adoption ofmanagement practices, best suited to local conditions,into the model.

Using farmers’ knowledge for otherwise unknownparameters, and ground truthing data and outputs withfarmers - who had previously been disengaged - built trustand demonstrated social learning. This built ownership ofthe process and a commitment to collective action.Stakeholder testing also helped the design of user-friendlyinterfaces for running scenarios and showing outputs.

Expertise from the social sciences helped design andfacilitate the processes of data collection andengagement. Advanced statistical methods were used toaccount for model uncertainties but the model’s credibilityand its probabilistic predictions were improved by themodeller explaining these in lay terms, and being opentowards interpretation and suggestions by stakeholders. In the words of a Tamar farmer: “How on earth could youhave come up with a single number as a result anyway?”

In the iterative and participatory planning process themodel was the essential tool that enabled participants toframe the scale and severity of selected water qualityproblems. Management scenarios were explored in realtime, stimulating dynamic and engaged debate.Management options were then costed and a collectiveassessment made of governance and implementationarrangements.

The lessons from this experience were combined withwider interdisciplinary assessments of internationalexamples of catchment management programmes thatintegrate the best science with effective communicationtools and decentralised and collaborative modes ofgovernance. This provides guidance for catchmentmanagement in the UK and other areas of intensiveagriculture and dense rural settlement.

Catchment management for protection of waterresources Laurence Smith, School of Oriental and African Studies

9STAKEHOLDER ENGAGEMENT

CASE STUDYCOMPETENCY GROUPS - AN EXPERIMENTAL METHODFOR COLLABORATIVE ENVIRONMENTAL RESEARCH

This project sought to understand how and why floodrisk management, and the forecasting technologies onwhich it relies, become matters of public controversy.

It combined the ethnographic techniques of science andtechnology studies with hydraulic modelling andexperimented with a new method of bringing theknowledge of local people with experience of flooding tobear on the modelling of flood risk - competency groups.This method was trialled in two localities in which floodrisk management was already in dispute - Ryedale inYorkshire and the Uck catchment in Sussex.

The competency group approach is designed to ‘slowdown reasoning’ in the event of a knowledgecontroversy, enabling those affected by flood tointerrogate the expert knowledge claims and practicesthat inform existing flood management policies and totry out alternative ways of understanding and mitigatinglocal flooding problems. It centred on bi-monthlymeetings over a 12-month period in which hands-oncomputer modelling became the key practice,supplemented by field visits, the production and analysisof video and photographic materials and othercollaborative research activities.

The groups combined the different experiences and skillsof the natural and social scientists in the project team(university members) with those of volunteer residents

affected by flooding (local members) by working closelywith various materials and artefacts that embody expertknowledge claims – flood maps and computer models.This way of working also emphasises the importance ofproducing new materials and artefacts to help the group’sown knowledge; and propositions ‘travel’ and thereforemake a difference to public debate and policy-making.

The approach requires a sustained commitment from allto negotiate the different modes of reasoning of fellowparticipants and to appreciate the different kinds ofexpertise brought to the collaborative production ofknowledge. The project produced a web-resource to helpothers in trying out competency groups.

Understanding environmental knowledge controversies:the case of flood risk managementSarah Whatmore, University of Oxford

10 STAKEHOLDER ENGAGEMENT

CASE STUDYMODELLING ENVIRONMENTAL FACTORS FOR THE COMMUNITY

Algal blooms are excellent indicators of declining waterquality. Their presence indicates that there are excessivenutrients in the water body. In Loweswater, theoccurrence of regular potentially toxic blooms of blue-green algae and their impact on visitors and locals, madeit important for scientists and local people to try tounderstand more about the sources of those nutrients.Then locals could begin to address potential causes andtake on the management of their catchment area.

The algal problem at Loweswater clearly requiredstudying linkages between humans and the land and alsobetween land and water. This meant scientists fromdifferent disciplines had to step outside of theirdisciplinary boundaries and focus on how the specificsystems that they specialise in are connected to, andinteract with, other systems.

Scientists, local residents and agencies worked togetherto form a new organisation, the Loweswater Care Project,through which scientists could draw on the benefitsgained from being closely engaged with catchmentstakeholders from institutions and the community. Thismaximised the use of information, expertise and dataavailable to the modellers and drew on the knowledgeand experience of stakeholders to identify the importantfactors which needed to be included in the modelling.Residents’ memories of land use changes, changes inagricultural practice, environmental changes and theways in which relations within the community and thecomposition of the community of Loweswater havechanged in time contributed substantially to inform theresearch priorities.

The team developed a series of three simple linkedmodels, with outputs from one model leading into theother. The models linked land management processes andland cover to catchment hydrology and nutrient flow andfurther to algal populations in the lake. Modellers felt anunusual responsibility to make the model relevant toLoweswater Care Project participants, because of theclose connections forged, especially with those involved infarming the land.

Modelling incorporated information from farmers on farmmanagement and soil data, septic tanks data collected bya local resident, rainfall data collected by residents in thecatchment area, alongside information collected byscientists and the Environment Agency.

The modelling demonstrated to the Loweswater CareProject how land cover and land use in the catchmentimpact on algal populations and showed various scenariosof change of algal populations in response to changes inland cover and use.

Testing a community approach to catchment management Claire Waterton, Lancaster University

QUANTIFYINGWEIGHING UP AND MEASURING DIFFERENT DISCIPLINARY ANGLES

11

12 QUANTIFYING

SEVERAL PROJECTS TOOK ON THE CHALLENGE OF MEASURING AND QUANTIFYINGINTERDISCIPLINARITY THROUGH STRUCTURED METHODS.

In some projects experts from different disciplines and keystakeholders scored and ranked the relevance of researchfindings, topics and factors to study. In others they put avalue on the social and natural factors contributing to aparticular problem. By engaging experts from a wide

range of disciplines and stakeholders with differinginterests and concerns, researchers aimed to weigh up andnumerically define the different disciplinary angles of theproblems studied, whilst at the same time avoidingdisciplinary bias.

CASE STUDYEXPERT WEIGHING OF RISK FACTORS

The project studied microbial pollution risks ofwatercourses from livestock farming. Joint reasoning and learning about this environmental protection issuefrom different analytical and interpretive starting pointswas a fundamental aspect of the research process. Co-production of data, interpretations and outputsresulted in the development of a practical learning toolto mitigate microbial risks at both the farm and fieldlevels. It also informed a process of citizen science forthe public scrutiny of these risks.

Making sense of the underpinning drivers of pollution risksdepends on a wide assessment of the various physical,social and economic characteristics of farms that cancontribute to run-off and result in water pollution.

The team integrated the monitoring of watercourses forpotential pathogens with determining E. coli mobilisationfrom faecal material; and desk-based identification ofdiffuse microbial pollution mitigation measures withinterviewing farmers about their attitudes and practicestowards livestock management. Farm maps were usedduring interviews to represent the character of, andreasoning behind, management actions.

Expert elicitation was used to prioritise and assignweighting to key social and natural risk factors that existacross farm systems. Such risk factors are categorisedinto four components: infrastructure characteristics offarms; E. coli burden; the run-off potential of farm land;and farmers approaches and attitudes towards manure,land and animal management.

The results were translated into a risk assessment tool -nick-named the ‘kite’ tool to reflect the four componentsthat interact to influence risk - that can be used tographically represent cumulative risks posed by any farmenterprise, so that decisive interventions can be put inplace.

Risk also depends on a deeper and unresolved set ofuncertainties regarding what might constituteappropriate levels of intervention, and whereresponsibilities for action and investment lie. The projectinvestigated the scientific and political basis for actionagainst these microbial risks through a citizens’ jury,which reinforced the case for strong state support ofmicrobial risk management, but also added weight to thecase for cross-industry subsidies of mitigative action.

Sustainable and safe recycling of livestock wasteDavid Chadwick, Rothamsted Research, North Wyke

13QUANTIFYING

CASE STUDYBEST-WORST SCALING OF INTERVENTIONS FORINFECTIOUS DISEASES

The researchers applied an established market researchtool - best-worst scaling - to elicit and analyseperceptions of the effectiveness and practicality ofinterventions to manage E. coli O157 risk in farming andrural settings.

Candidate interventions were proposed by project teammembers based on findings from key stakeholderinterviews. This generated 99 interventions to manage E.coli O157 risk. Experts from a broad cross-section ofacademic disciplines - public health sector,environmental microbiology, epidemiology, veterinarysciences and land management – as well as manyfarmers in the study regions, then commented on therelative effectiveness of the proposed interventionsthrough best-worst scaling. This allowed each expert togive a differing perspective on the marking of the mosteffective process.

Best-worst scaling is a choice-based technique wherebyrespondents make repeated choices between sets ofoptions. In this project, experts in a first instance assessedtwelve options sets that each contained five interventions,indicating the most and least effective measures to reduceE. coli O157 in each set. This round reduced the number ofinterventions to the 30 interventions considered to bemost effective. The process was then repeated with

livestock farmers, who chose what they perceived to bethe most and least practically implementable interventionsin the field. Experts and farmers therefore combined theiropinions on an equal footing to generate a list ofinterventions that were considered to be both effectiveand practical to implement.

A selection of the top 30 interventions was modelledusing quantitative microbiological risk assessment todetermine their potential to reduce E. coli O157 exposureto humans.

This technique has the potential to be applied to assessinterventions associated with other infectious diseases.

Reducing E. coli risk in rural communitiesNorval Strachan, University of Aberdeen

14 QUANTIFYING

CASE STUDYSUSTAINABLE APPRAISAL FRAMEWORK

Future policies are likely to encourage more land useunder energy crops - principally willow, grown as shortrotation coppice, and the tall grass Miscanthus. Thesecrops will make an important contribution to the UK’scommitment to reducing CO2 emissions and are grownunder low input agriculture.

However, they are quite different from the arable cropsthat we are used to and it is not clear how planningdecisions based on climate, soil and water should bebalanced against impacts on the landscape, socialacceptance, biodiversity and rural economy. The problemis that these wider implications of land use change havenot been investigated, and there has been no attempt toidentify the full extent of the broad range of potentialimpacts, or more usefully, to highlight scenarios for landuse change which can minimise negative impacts andaccentuate positive impacts.

The dominant environmental governance goal over thelast two decades has been the achievement ofsustainable development. Such a goal is normallyconceptualised in terms of social, economic andenvironmental ‘pillars’, and recognises the need forhumanity to co-exist with nature. As such, aninvestigation of the implications of spatial change needsto combine evidence of impacts on the naturalenvironment within a socio-economic and politicalsetting that provides the context for sustainable landmanagement decisions.

To combine the social and natural sciences, sustainabilityappraisal was used to incorporate social, economic andenvironmental criteria and data in a single framework.This uses a workshop approach that enables participants

to question natural scientists, social scientists andeconomists to improve knowledge and understanding(the analytic component); thus facilitating moreinformed deliberation (the deliberative component) overthe key sustainability criteria to include.

Sustainability appraisal is adapted from spatial planning.It relies on examining the sustainability implications ofpolicy options in order to determine the best ones totake forward. Implications are determined by testing theoptions against social, economic and environmentalsustainability objectives measured through the use ofindicators.

In this project, the team agreed on scenarios to be tested(equivalent to policy options) with a broad range ofstakeholders including growers, government agencies,energy companies, non-governmental organisations,union members, etc. The same broad stakeholderengagement was used to develop the sustainabilityappraisal framework and to interpret the results.

Impacts of increasing land use under energy cropsAngela Karp, Rothamsted Research

15QUANTIFYING

CASE STUDYCHOICE EXPERIMENT TO VALUE CHANGES IN UPLANDLANDSCAPES

Recent work in the economics literature stemming frombehavioural psychology suggests that the act ofexperiencing a ‘good’ impacts on preference. That meansthat, for a given individual, their forecast or memories ofutility (relative satisfaction) are likely to differ from thosestated at the moment of experience. But in the past thistheory has only been tested using happiness-basedmeasures of utility, and not for environmental goods.

This project applied the choice experiment technique tovaluate changes in upland landscapes in the UK in orderto identify whether experience, at that moment or inmemory, impacts on the value associated with changesin ecosystem services under different managementregimes. Four treatments were employed using the samesample to measure decision utility (off-site), experiencedutility (on-site), and remembered utility at two differenttime intervals (off-site).

On-site treatment generates very different estimates ofpreferences than any of the off-site treatments. Whilstmeasurement of experienced utility is fraught with

difficulties, the approach taken allowed the identificationof experiential impacts on utility.

It was found that the act of experiencing anenvironmental good altered how individuals madedecisions about environmental resources. They changedtheir views on the cost to the environment, with moreemphasis placed on environmental goods. This resultmay have implications for the future use of experiencedutility as a basis for the valuation of public goods.

The sustainability of hill farmingNick Hanley and Dugald Tinch, University of Stirling

MODELLINGCREATING REPRESENTATIONS OF COMPLEX REALITIES

16

17MODELLING

PROJECTS ALSO APPLIED INTERDISCIPLINARITY AT THE DATA LEVEL, BY MODELLING SOCIAL,ECONOMIC, ECOLOGICAL, ENVIRONMENTAL AND HYDROLOGICAL FACTORS FROM INPUT DATAREPRESENTING DIFFERENT FACTORS OR PERSPECTIVES TO A PROBLEM.

In some instances this was done by modelling acombination of social and environmental data, either usinglinear or tabular modelling, or spatial modelling in ageographical information system. Other projects

incorporated local knowledge – qualitative or quantitative– as input data into models. Local or stakeholderknowledge was often used to fine-tune models and makethem better fit the reality.

CASE STUDYMODELLING THE IMPACTS OF THE WATERFRAMEWORK DIRECTIVE

This project brings together hydrology and economics toexamine the physical impacts of the EU WaterFramework Directive upon rivers and how the changes inland use needed to achieve a reduction in pollutants inwater are likely to impact upon already fragile farmingcommunities.

The project developed a methodology for integratedmodelling of the relationship between rural land use (andconsequent farm incomes) and water quality (includingdiffuse and point sources of nutrients, pesticides andfaecal matter and consequent ecological status). Thismethodology combines econometric statistical and linearprogramming analysis of a large cross-section and timeseries panel database of farm activity with hydrologicalmodels linking land use with consequent water quality.

The model is being used to provide policy guidance onstrategies for implementing the Water FrameworkDirective within the context of ongoing CommonAgricultural Policy reforms. Particular attention is givento the impact on land use, farm incomes and the ruraleconomy of alternate policy options.

The project is also assessing economic values for thesocial benefits that may be generated by implementationof the Water Framework Directive and testing thetransferability of these benefits assessments.

The project shows the impact of Water FrameworkDirective policy changes that aim to reduce diffusepollution on farm activities, farm income and waterquality. A hydrological model at catchment levelsimulates how pollutants from fertilisers or pesticidesleach into water and how this affects aquatic biology. Aneconometric shows land use changes, farm activities andincomes for farms in water catchment areas, as a resultof policy changes that aim to reduce diffuse pollution byreducing inorganic fertiliser application, livestock rates orconversion from arable to ungrazed grassland.

These models can be used to predict land use changes inresponse to shifts in environmental, policy, or marketforces; and to assess how such changes in agricultural landuse are likely to affect levels of diffuse pollution to rivers.

Modelling the impacts of the Water Framework DirectiveIan Bateman, University of East Anglia

18 MODELLING

CASE STUDYSIMULATING THE PRESENT FROM THE PAST

The threat to biodiversity and rural landscapes from treedisease epidemics is greater today than ever before.

Expanding international trade in plants, together withincreased passenger movements, has led to the entry ofvarious invasive pathogens into the UK in recent years.Many of these have the capacity to kill native trees in verylarge numbers. Tree disease management is complicatedby scientific uncertainty, the presence of large numbersof stakeholders, the difficulty of establishing clear lines ofinstitutional responsibility and a clash of public andprivate interests over who should pay.

However, tree disease epidemics are not new andpolicymakers have various historical precedents on whichthey can draw in seeking to avoid past mistakes andlengthen institutional memory. In the UK, the Dutch ElmDisease epidemic of the 1970 killed over 30 million trees.For scientists and public alike, it is probably one of themost dramatic domestic environmental events in theirlifetimes. This project has sought to integrate historicalanalysis into the heart of the current biosecurity debateby comparing the Dutch Elm Disease epidemic with theSudden Oak Death outbreak that is currently unfolding inthe UK.

For Dutch Elm Disease to be a rich source ofinterdisciplinary and policy-relevant knowledge tounderstand present day threats like Sudden Oak Death,there needs to be an integrated understanding of thebiological and socio-economic aspects of these differentdisease problems.

This was achieved through a linked historical andcontemporary analysis that began with a biophysical andsocio-economic reconstruction of the Dutch Elmoutbreak. Modelling work, directly informed by insightsfrom archival research and interviews with key actorsinvolved in the attempted management of the outbreakat the time, was undertaken to simulate the origins,

spread and eventual trajectory of the disease. Thisallowed the researchers to identify key events andphases of the outbreak, drawing on an interdisciplinaryunderstanding of the interaction between the biology,epidemiology and economics of the epidemic.

Further work explored the sensitivity of the outbreak todifferent courses of action. Different disciplines werebrought together in order to arrive at a fullunderstanding of the way in which interacting biologicaland institutional factors shaped the course of the diseaseand its outcomes.

This fresh analysis of Dutch Elm Disease sheds importantlight on the current Sudden Oak Death outbreak andexplains what policymakers are encountering in theirattempts to contain it. Despite biological differencesbetween the two disease systems, the researchdemonstrates the difficulties in both cases of earlydetection and the speed with which outbreaks becomeuncontainable once established in the widerenvironment. But it also points to an enduring lack ofpublic awareness of the underlying drivers of disease riskand need for a broader debate amongst stakeholders ofthe conflicts between freer trade in horticultural productsand effective biosecurity.

Lessons from Dutch Elm Disease in assessing the threatfrom Sudden Oak DeathClive Potter, Imperial College London

19MODELLING

CASE STUDYMODELLING AND MEASURING RURAL INEQUALITY

Achieving sustainable rural development depends on thedistribution of social, economic and environmental goodsand services that are needed to maintain and reinforcethe vitality of rural areas.

Inequality in such goods and services has importantimplications for individuals or groups of peopleexperiencing it, but also for society as a whole. In urbanareas, poor environments are associated frequently withdeprivation and social exclusion, but the relationshipbetween environment and deprivation in rural areas isless well understood.

Research that can inform evidence-based rural policy-making requires readily accessible data, from both socialand natural scientific disciplines, about the distributionand inequality of social, economic and environmentalconditions. The researchers quantified and measuredsuch inequalities throughout rural England by developinga high resolution spatial dataset containing: the naturaland constructed physical components of rural areas; thequalities and character of places and people in thecountryside; information about living and working there;and the political and economic context. After identifyingthose areas where inequalities were greatest theyinvestigated how rural residents experience the kinds ofinequalities identified and which inequalities theyperceive as inequitable.

One of the challenges was the apparent incompatibilityof spatial data collected by different academicdisciplines, due to the differing scale and nature of datacollection and the phenomena studied. This requires acritical understanding of data form and distribution.

Social data typically correspond to administrative orpolitical areas, which are often subject to temporalchange and not related to the landscape. Environmental

data such as land cover or biodiversity correspond toecological zones and are frequently organised as grids.For farmed areas, the farmers to whom socio-economicdata are attached are located spatially at points that maybe some distance from their land holdings, the areasassociated with environmental data. Also underlyingdistributions pose challenges, like irregular distributionsof land cover, land use or settlement, or continuousdistributions for air pollution. Variation also depends onhow data are collected and organised. The combinedexpertise of researchers with different disciplinarybackgrounds was essential in getting to grips with suchchallenges.

The team selected as its basic spatial unit the LowerSuper Output Areas. These are areas with consistentpopulation size (average 1,500 residents) but highlyvariable in size, designed for the collection andpublication of small area statistics for the 2001 Census ofPopulation in the UK. They mapped onto those areas arange of data related to economic activity, income andwealth, health and well-being, and ecology, land and theenvironment.

Social and environmental inequalities in rural areasMeg Huby, University of York

VISUALISATION VISUAL SCENARIOS OF FACTORS AND INFLUENCES CONTRIBUTING TO CHANGE

20

21MODELLING

PROJECTS HAVE USED MODELLING METHODS TO DEVELOP VISUAL SCENARIOS SHOWINGHOW DIFFERENT FACTORS OR INFLUENCES CAN CONTRIBUTE TO CHANGE.

This enables stakeholders and researchers from differentdisciplines to study the consequences or future options ofparticular scenarios, proposed changes or interventions.Such techniques particularly help to make scenarios more

accessible for stakeholders to understand and evaluateand gives them a degree of realism that could nototherwise be achieved.

CASE STUDYMODELLING WITH VISUALISED SCENARIOS ANDCOGNITIVE MAPS

To manage the countryside sustainably in future, oneneeds to understand how it is likely to change and why.Some of the biggest changes are likely to happen inupland areas, and may compromise the many importantbenefits they provide to society, such as clean water,carbon storage and the protection of internationallyimportant species of plants and animals.

This project combined knowledge from localstakeholders, policymakers and scientists to identify thecurrent needs and aspirations of those who live, workand play in three upland areas: Peak District NationalPark, Nidderdale Area of Outstanding Natural Beauty,and several catchments in Galloway.

By combining interviews, surveys and computermodelling, scenarios were developed that can be used asa starting point to discuss the opportunities and threatsthat climate change may bring and find viable optionsfor future upland land management and sustainable rurallivelihoods.

The team’s approach to interdisciplinary working wasdiverse. Co-researchers trained colleagues in theirdisciplinary methods. Joint researcher-stakeholder sitevisits enabled the exchange and integration ofknowledge. Stakeholders scrutinised information andpolicy briefs developed by research from publishedliterature. Conceptual models of system structure andfunction developed from researchers’ expertise, localknowledge and research literature formed the basis forthe development of computational models used toexplore scenarios.

An integrated computational model was developed fromdifferent disciplinary models: an economic agent-basedmodel of land manager’s behaviour; a hydrological soilerosion model of the effects of land managementbehaviour on soil properties and erosion; a habitatsuccession model of the effects of land management onplant species composition and succession; and a land usechoice model of land managers’ activities on individualparcels of land.

This integrated model was then linked to a carbon modelestimating carbon fluxes and was run across the threeproject areas. For each area a range of scenarios wereconsidered, including finding optimal carbonmanagement. The optimal carbon management scenariowas used to examine the viability of carbon offsettingusing peat restoration within the English Peak Districtand used to project scenarios forward to 2030 givenclimate change.

Model outputs were integrated with qualitative outputsto develop scenarios for upland futures. These scenarioswere visualised and developed into short films to elicitstakeholder feedback on adaption options for changes ofthe uplands.

The decision to use video as a medium to integrate localand scientific knowledge, was in itself based onstakeholder feedback. Documentary and animatedparticipatory video was created to communicate projectfindings and adaption options online and via socialmedia. At the same time, an interactive map-basedwebsite now allows people to share videos, photos andthoughts about what the uplands mean to them. A songand music video are currently being created tocommunicate the message “More than just a bog: carbontrap, sponge, history classroom”.

Sustainable Uplands: learning to manage future change Klaus Hubacek, University of Leeds and Mark Reed,University of Aberdeen

REACHING OUT ACROSS RESEARCH DISCIPLINES,ENGAGING STAKEHOLDERS, GAINING KNOWLEDGE,STUDYING COMPLEX REALITIES.

RELU DATA SUPPORT SERVICEUK Data ArchiveUniversity of EssexWivenhoe ParkColchester, CO4 3SQ

Email: DSS-RELU@essex.ac.ukTelephone: +44 (0)1206 872234

relu.data-archive.ac.uk