CHAPTER 12
Concluding reflections and planning the next research cycle

[Chapter 12 in: Allen, W.J. allenw@landcare.cri.nz (2001) Working together for environmental management: the role of information sharing and collaborative learning. PhD (Development Studies), Massey University.]

As outlined in Chapter 1, this thesis represents an action research inquiry into how an adaptive management or 'learning by doing' approach, consistent with the concept of sustainable development, can be initiated and implemented in complex, regional or large scale contexts. In particular the inquiry focusses on the social and institutional issues that arise in ensuring adequate stakeholder participation in generating and managing information to support collaborative decision making and subsequent change on-the-ground. Action research provides an appropriate methodology for an intervention-based approach which is concerned with how the different groups and individuals involved design and implement action in relation to one another.

This starting point highlights that action research is focussed on possibility rather than prediction. Because it is not value neutral it is therefore important to state any underlying values in advance. In this sense, most action research can be seen to be guided by two fundamental principles: i) that there is a need to democratise the knowledge process -- so people normally shut out from research and information become involved in the research itself, learning how to obtain information and how to use it; and ii) that it acknowledges a social change emphasis -- whereby the goals of research are to engage in action that reverses inequalities, empowers the have-nots, and ultimately transform society so decision-making becomes more transparent and democratic. Within these broad principles my work as outlined in this thesis is undertaken within an environmental research institute, so the focus is on finding ways to improve people's relationship to the environment, and help environmental decision-making to be built on the improved use of technical information.

As a starting point for this research an initial framework of ideas and concepts was outlined as consistent with the aims of adaptive management, and capable of guiding different stakeholder groups to work collaboratively to identify and implement more sustainable resource management practices. What can be regarded as the first version of the Integrated System for Knowledge Management (ISKM) approach (Figures 4.3 and A1.2) sets out a number of steps suggested as necessary components within any approach designed to achieve this. In particular this approach acknowledged that: i) relevant and practical strategies for action could only be developed through a co-operative and integrated process which combined knowledge from both manager experience and conventional science; ii) there was a need to document these for the benefit of the wider user-community, along with supporting information, through a user-friendly and accessible computer-based information system; and iii) the continuing input of new science and management-based experimentation was needed to maintain the relevance of such an information system over time.

This framework was also shown (Chapter 4) to be supported with underlying concepts of the need for participation, emphasising the importance of local knowledge, experiential learning and systems thinking. This framework was then applied, and refined, through a case study approach to guide the development of ecologically-based research and development efforts. The research involved one main case study (tussock grasslands) and three smaller, but related, ones (black stilt, Tb vector control, NRM-changelinks website development). The tussock grasslands case study began in June 1994 and remains ongoing at the completion of this thesis inquiry in June 2000.

Consistent with an action research approach, the inquiry design was shown to be emergent, progressively developing as it was influenced by the events that took place during the case studies and by the subsequent analyses that are made. Each subsequent chapter represents one plan-act-reflect cycle within this larger inquiry. Thus, within the tussock grasslands case study Chapters 5, 6, 8 and 11 deal with a number of different issues that arose from the implementation of ISKM in a decision-making environment characterised by multiple social perspectives. These were, respectively, how to manage: i) forums which support constructive community dialogue; ii) evaluation processes which meet the need of the different parties involved; iii) multi-stakeholder information networks; and iv) integration of both 'soft' and 'hard' inquiry processes within research initiatives.

Chapter 7 looked at conflict through one of the smaller-bounded case studies using a dispute resolution exercise around the management of a rare wading bird (black stilt). Chapter 9 involved the application of ISKM to improve the identification and uptake of Tb vector control, and also how groups can be supported as part of this process. Finally, Chapter 10 used the experience gained in the development of the NRM-changelinks website to investigate the potential use of the Internet to leverage improved information dissemination and networking. In all of these cases the issues under investigation were related back to how they fit into the larger tussock grasslands case study.

As its name implies, action research inquiries can be viewed as having two main outcomes: action and research. The focus is action to improve a situation and the research is the conscious effort, as part of the process, to formulate public knowledge that casts light on the functioning of the client system or the action research process itself, or both. These outcomes are generated through the iterative use of a 'plan--act--reflect' cycle of collaborative inquiry, which in this thesis is illustrated by each of chapters 4-11. The subsequent material in this concluding chapter provides a wider cycle of reflection which first reviews the 'act' outcomes of these different cycles and then provides some thoughts that emerge from this study to help guide the planning of future inquiry cycles.

What is significant about the tussock grasslands case study is not that it has resulted in a regional adaptive management approach to tussock grasslands management -- it has not yet -- but rather that those involved (researchers, farmers, conservation managers and local government staff representatives) have learnt more about the processes and issues involved in working together and sharing information, and continue to seek ways to implement adaptive management.

During this process the ISKM framework has been progressively refined (see Figures A1.2, 1.2, 5.7, 8.1, 9.5 and 12.1, respectively), and it has been used to support the efforts of increasingly larger ecologically-based research programmes. Its initial development and use was undertaken to help a team of researchers (in the wider sense of the term) address the issue of an invasive weed (Hieracium spp.) in the tussock grasslands of the South Island high country. Subsequently the ISKM approach was used to underpin the work done of the Tussock Grasslands Programme, which was until June 2000 the major research programme in this area of New Zealand. Through this programme the application of ISKM was extended to address more general issues of tussock grassland dynamics, nutrient flows, and ways to better integrate conservation and pastoralism in this scenic region. This work remains ongoing as the 'montane objective' within an expanded research programme, 'Changing landscapes and restoration of biodiversity', which represents the main focus of New Zealand research into the protection and enhancement of biodiversity in productive landscapes.

Moreover, the design of several major science programmes has been influenced by this work. For example, the importance of ensuring that the 'participatory' component of a science programme is integrally linked with other aspects of the research, and that the outcomes of stakeholder involvement are fed into the research design to influence subsequent activities and strategies as discussed in Chapter 11 is now evident in the Landcare Research-managed biodiversity in productive landscapes (see above) and 'Integrated land and water resource management in complex catchments' programmes. Equally the action research nature of this research to date, particularly as it relates to drawing out public knowledge for use in other environmental management situations, is evidenced by a number of published papers that have addressed those social and institutional factors having an impact on the implementation of ISKM (see for example http://nrm.massey.ac.nz/changelinks/rel_pap.html).

Within the tussock grasslands of the South Island high country several stakeholders from outside the science sector who have demonstrated their commitment to contribute to the ongoing development of the ISKM approach. Key among these are the farming groups who are using the condition assessment model and have agreed to look at how the monitoring results from their properties can be shared as part of a wider 'learning' process. These groups are, in turn, supported in their efforts by the Otago Regional Council. Similarly, the Department of Conservation continues to support the ongoing development of an Internet-based Tussock Grasslands Management Information System (http://tussocks.net.nz/ ), which is seen as complementing these efforts.

In the Waitaki/Mackenzie basins activities to build trust between the Department of Conservation and the local farming community have been initiated following a conflict management exercise (see Chapter 7). This is part of an ongoing process to help the two groups improve their communication and the subsequent management of wildlife in the district. While both parties would undoubtedly agree that this will be a long process, a subsequent evaluation I undertook with agency staff highlighted improvements that had been implemented (unpublished minutes 9 December 1999). These included the publication of a regular newsletter, the holding of open days jointly organised by the Department of Conservation and local farmers, and an increasing focus on how conservation, farming and tourism activities could be integrated.

In the work funded by the Animal Health Board to improve the identification and uptake of effective ferret control efforts in North Canterbury (see Chapter 9), the question of how agencies could better support community-based groups to provide a vehicle for improved information sharing and collaborative learning to influence behaviour change was investigated. The subsequent findings are now used to support the functioning of Tb vector control groups throughout New Zealand (e.g. Oliver et al. 2000).

Finally, Chapter 10 documented the lessons learnt through the development of the NRM-changelinks website (http://nrm.massey.ac.nz/changelinks/) as an exercise to look at the potential for Internet-based information sharing and networking in the area of developing collaborative approaches for environmental management. In terms of outcomes, this has now become one of the larger participatory resources on the Internet in terms of site traffic (see site statistics http://www.sitemeter.com/statsapp1/default.asp?action=stats&site=webtracks).

Collectively, these examples of outcomes in practice help to verify the findings of the action research inquiry. By definition it is a science of implementation, and as practitioners (science programmes, agencies, community groups, etc.) take up the results they are confirming their confidence in them -- or at least, their intent to pursue them to see what happens! This does not, of course, mean that the inquiry is finished, as the approach also aims to leave practitioners with the capacity to question and improve those practices. The same applies to the action research process itself: it is important to show that the process is enabling more targeted questions to be developed as the inquiry progresses.

In this regard we can see that, from the rather open-ended approach to the action research inquiry process that began with the Hieracium Management Programme, subsequent activities have become more clearly specified. For example, in the research programmes cited above our future action research inquiries are focussed on three linked areas, which appear to have the potential to leverage improved information flows and collaboration in natural resource management:

• improving participation and the use of local knowledge in the research process,
• improving the dissemination and use of this knowledge in the wider community through improved networking and collaboration (including the use of the Internet),
• and capacity building (supporting the above approaches) -- through participatory monitoring and evaluation.

From my own perspective these emerging directions for exploration have developed through the experiences documented in this thesis, and similarly the way in which they will be investigated in practice builds lessons and insights gained from this study. Accordingly, this final chapter can be seen as a wider process of reflection covering the research undertaken over the past six years as a guide to planning the implementation of future action research-based initiatives centred around these activities.

As discussed in Chapter 2, contemporary development-literature promotes a more embracing development paradigm that places people at the centre and seeks to empower stakeholders to influence and share control over development initiatives and the decisions and resources which affect them. Even as macroeconomic policies and trends continue to shape resource development opportunities, the move away from centralised planning by governments is requiring R&D initiatives to work towards empowering communities to deal with their own needs. In this regard, the challenge for researchers is to work with communities and undertake inquiries that begin with the search for solutions to social (community) problems -- placing an emphasis on problem context and identification.

Because one of the main issues related to establishing such a collaborative approach within the wider social and institutional contexts of catchments and regions is one of implementation, an action research approach (see Chapter 3) provides an appropriate methodology. This is directly applicable to the study of how individuals and groups design and implement action in relation to one another.

Moreover, there is an increasing realisation that new sources of 'expert knowledge' and databases are needed to identify persistent resource management practices more clearly (see Chapters 4 and 5). In many cases, the knowledge that is required about the past and present state of our natural resources, and about the relationships between social and environmental systems, is held within local communities and implementation and policy agencies. Accordingly, it follows that the task of organising information to understand better the links between natural resource management, social realities and ecological dynamics should be a collaborative venture between research scientists and the different stakeholders involved.

In this regard, a first version of the Integrated Systems for Knowledge Management (ISKM) framework was outlined at the beginning of this inquiry (Chapter 4) as an approach for supporting such a collaborative approach for managing information. A second version, which included positioning information technology as a supporting rather than a central component was set out in Chapter 5. Subsequently, Chapters 9 and 10 highlighted the growing role of the Internet in allowing people to create, annotate, link together and share information from a variety of sources and media. It appears to have considerable potential in multi-stakeholder situations to extend information-sharing, learning and networking. A third version emphasised the need to build relationships for change and identify clear roles for all the parties concerned at the beginning of such collaborative initiatives (see especially Chapters 9 and 11).

Finally, a fourth version is outlined below (Figure 12.1) which acknowledges that implementation activities and their subsequent monitoring and evaluation should be seen as separate activities. As Chapters 6 and 9 illustrate, effective collaborative environmental initiatives are the ones that pay attention to both the task and the process, and so meet the needs that the different participants have in both areas. The task can be defined as what those involved have to do (e.g. reduce pest numbers). The process is concerned with how people and groups work together and maintain relationships. Experience shows that people often neglect process issues (commonly to concentrate on the task). However, both task and process will suffer if they are split from each other.

Similarly monitoring and evaluation need to be seen as distinct but related activities. Monitoring provides the raw data to answer questions. But in and of itself, it is a useless and expensive exercise. Evaluation puts data to use and thus gives them value. Evaluation is where the learning occurs, questions are answered, recommendations made, and improvements suggested.

In practice the use of the steps within ISKM remind us that an information system cannot be regarded only in terms of its transfer component (often a paper, or a computer-based model/DSS). Rather, as Chapters 2 and 4 highlight, such a system is better viewed as a 'social system' within which people interact to create new knowledge, and broaden their perspective of the world. This is a significant perception distinction for science whereby dissemination (information technology/extension) and users form elements of a larger knowledge system (Figure 4.2). This concept is synonymous with the idea put forward in Chapter 8 that learning is socially constructed, occurring through interactions between individuals, between individuals and groups, and between groups of groups. Learners learn to function in a community by developing a shared language and acquiring the community's subjective viewpoint.

Learning, in this sense has two components: its process and the outcomes of that process. Change can be observed as an outcome of learning. This, in turn, must be viewed as an accumulative process which builds on existing practices and norms through interactive learning. While information is central to this process, Chapters 9 and 11 remind us that there are some supporting social processes that are required for this to happen. These include forums to develop a shared understanding around issues, management of a moderate degree of conflict and the provision of a supportive environment. Central to the notion of this supportive environment is the concept of social capital -- the framework that supports the process of learning through interaction -- and which requires the formation of networking paths that are both horizontal (across agencies and sectors) and vertical (agencies to communities to individuals). The quality of the social processes and relationships that social capital supplies -- within which learning interactions take place -- is especially influential on the quality of the learning outcomes in collaborative approaches.

Taken one step further, this suggests that this social capital plays an important role in influencing change, and sustaining a social and institutional environment that is ready to adapt. Equally, it explains why change is much harder to achieve in some situations than in others. In many cases stakeholders will lack the culture for participation in multi-stakeholder processes. Thus, building capacity (or social capital) for participation should in many cases be seen as a first step. That capacity should not just be assumed to be there.

Nor is this capacity-building just a function for science. It is something that needs to be built into all development activities -- public health, education, environmental management, etc. Thus each sector will contribute to the development of social capital, which will, in turn, provide a richer social environment for subsequent efforts to operate in.

In this context efforts for change will need to be centred around supporting groups of people working together. As Pretty (1998) emphasises, true participatory projects are those which empower people by building skills, interests and capacities that continue even after the project ends. This implies the institutionalisation of such initiatives and the corresponding capacity for activities to spread beyond the immediate project in both space and time.

Increasingly the role of groups as a catalyst for change is becoming well accepted in terms of environmental management. These groups may be formally constituted (e.g. a landcare group, or agency team), or they may comprise members of a working group that has come together to undertake a one-off task. Such groups will have been formed for a range of reasons: to build trust between different parties, to develop best-practice guidelines, to establish community monitoring schemes, to develop a shared vision across a district or catchment, or to learn to use the Internet.

However, if we are serious about the need to foster a more collective approach to environmental management that is capable of the transformational change being sought, we have to do more than just work together to undertake specific projects. Roughly put, the collective vision that emerges is one which establishes an ongoing process for sound environmental management within and among the many groups involved in some way (see Chapter 9). For this to happen groups need to develop the capacity to move beyond the completion of task-bounded activities to more actively catalyse change within their immediate membership first, and to spread that culture to others in their respective groups over the longer term.

More than any other activity and by its very nature, building the capacity for groups to mature in this way depends for its effectiveness on participant ownership and commitment. Its success will rely on the use of participatory and formative evaluation exercises (see Chapters 3 and 6) that strengthen the ability of groups and group-members for ongoing self-assessment and correction. It is by engaging in such exercises that groups will be able to progress through the continuum outlined in Chapter 9, moving from dependency to interdependency. The monitoring and evaluation component of environmental research and development programmes, then, needs to be equally about building capacity, diagnosing constraints and opportunities and trying to make programmes grow and expand, as it is about measuring and describing on-the-ground progress against pre-set targets.

Evaluation has a value beyond the immediate role of supporting capacity development within the immediate group. Because of their nature, collaborative initiatives are only made possible with support from a number of different parties, all of whom need to be kept informed of progress and outcomes. Funders need evidence that their investments are paying off. In particular, there is a need to develop intermediate indicators of success (e.g. within the time-frame of funding cycles) for process-oriented initiatives such as capacity building. Equally, other stakeholders who are giving of their time to help the particular effort (e.g. land managers providing information, agency staff facilitating projects) are also important audiences for information about the progress of the initiative. They too need evidence that their input is having an effect -- at the least, to maintain their motivation for continued involvement.

Good evaluation is also needed to generate useful feedback to guide implementation. Managers need feedback to assess progress, assist with planning, and guide ongoing refinements to operations. Moreover, such collaborative initiatives are essentially experiments providing opportunities for practitioners and action researchers to test their knowledge and experience. In this way much can be learnt about fundamental and cross-cutting questions concerning the best way to model programmes, or to examine more closely the role that 'social capital' and 'capacity-building' can play in helping achieve more environmentally sound management. This information, in turn, can be fed back to shape future policy and research agendas.

As we go about developing these lessons we also need to remain aware (Chapter 10) that learning from single case studies is problematic. There is, therefore, a need for action researchers to undertake cross-case analysis, which can provide more valuable and robust lessons by sharing reflections across programmes and projects (Figure 12.2). In this way our understanding of the variety of forms that interventions can take will be increased, shedding light on implementation issues, and increasing user confidence in the external validity of findings.

Figure 12.2 An across-case study approach to action research.

Moreover, there are other reasons why the action research component needs to be managed 'in conjunction with' research and development programs, rather than as a totally dependent component. By definition the goal of the action research team is likely in practice to be slightly different from the research or development programme. The latter is more likely to be aimed at developing outcomes in a particular topic area, while the action research component is equally concerned with looking for broader process lessons that can help with implementation issues across topic areas. Within this broader process, therefore, it is also important that science be seen as one of the 'stakeholders', and not as outside the process of change.

Moreover, if too much emphasis is placed on developing agreements and fostering improved working relationships to meet project deadlines it is likely to merely reinforce the status quo of the existing system (see Chapter 6). Action research, in this context, can merely lend itself to single-loop learning, focussing on changing individual and collective action strategies, while leaving underlying values and norms unchanged. In the end, this may be counter-productive in facilitating double-loop learning, which involves a more critical inquiry into and changing of underlying goals, values and performance measures, as well as strategies and assumptions.

As collaborative learning approaches are scaled up they will bring different challenges for action researchers. Most action research efforts that are reported involve the action researcher as closely connected with the changes being studied. However, as Ledford and Mohrman (1993) point out, in large-scale action there is a need to develop a strategy for learning about loosely coupled activities that occur in multiple locations. Increasingly in these situations, the client system will become predominantly policy makers, rather than managers and groups/teams.

This is particularly true in the New Zealand situation where action research studies such as those described in this thesis are undertaken more from the point of view of research, than to fill an extension function. Science programmes neither have the resources, nor the mandate, to undertake environmental extension/education in this area. It is therefore necessary to work alongside agency groups (often local government) who have the mandate and resources to use research findings to 'make a difference' on the ground. To help in these larger scale policy situations, we need better measures of process change as outlined above and some of these will have to be developed using quantitative methods.

As shown in this thesis, good information management and the development of constructive learning environments are key to bringing about change in environmental management. If these changes are to be achieved, individuals and communities must be supportive and directly involved in research and decision making. In these cases action researchers can play a major role in providing the tools and approaches to ensure that policy initiatives can be 'seen primarily as experiments, and dealt with as complex and uncertain ventures in which the participation of those who are expected to benefit is essential' (Rondinelli 1983).

If we assume that in the short term there will be no major shifts in financial resources to the environmental or development sectors, nor will current policies be massively altered to change the status quo, then we need other strategies for empowering people and changing current practices. The use of action research approaches to find out how to improve information flows, and strengthen and harness many existing aspects of social relationships in environment and development, may work to foster constructive change.

POSTSCRIPT (September 2001)
Some final reflections

Since this thesis was submitted more experience has been gained in the implementation of ISKM in the tussock grasslands case study, and there has been more time for reflection. In particular the Internet-based Tussock Grasslands MIS has now been made publicly available providing more lessons about both the Internet and the wider ISKM process. Similarly, the focus of our future action research inquiries have expanded to address the need for social capital highlighted in the closing chapters of this study.

With the benefit of hindsight, my thesis construction and timing underplayed the achievements and significance in developing an Internet-based MIS as an integral output of the ISKM process. At the initiation of the main tussock grasslands case study in 1995 the researchers began by outlining ISKM (Chapter 4 and Appendix I) as a participatory framework for developing a comprehensive management information system (MIS) to underpin adaptive management. The Internet was chosen as a platform because of its potential for providing access, easy updates, and supporting learning and communication across different groups. However, as evidenced by the content of subsequent chapters, the significance of an Internet-based MIS as an output was perhaps overshadowed by a more process-oriented focus on different aspects of stakeholder involvement in managing information. Moreover, the Tussock Grasslands MIS was only made available for public access on the Internet in June 2000, and that the site was not actively promoted until after this thesis was submitted in October 2000.

One major reason for the delay in implementation was the emphasis placed on trying to develop a comprehensive MIS before releasing it. This was not consistent with the original notion of using a prototyping approach for development (p. 60). Accordingly, the subsequent MIS not only took a long time to develop, its size and interlinking pages also made it difficult to referee. A lesson from this is to take advantage of the Internet's ability to accommodate progressive site development, and make future material available by posting even single pages as they are completed.

A subsequent evaluation of its use by Department of Conservation staff was undertaken in April 2001 by a colleague, Chris Jacobson. This showed that the MIS is being used by staff to support their decision making. Staff stated that it is a valuable resource and that similar sites in other areas of interests should be initiated. A number of specific requests for improvement were made (see Tussock Grassland MIS evaluations http://www.tussocks.net.nz/evaluation1.html ), and these are being addressed during the 2001/02 year.

The lessons learnt in the development of the Tussock Grasslands MIS highlight the need for science agencies to take advantage of the Internet to support stakeholders in accessing and debating information pertaining to complex environmental issues. One major problem for environmental decision makers is that information held by different stakeholders (local, tradition and science) is rarely available on a collective basis (e.g. Chapter 4 p.59). In this regard the Internet provides us with a new and convenient system for managing complex information which allows people to create, annotate, link and share information from a number of disparate sources and media. Similarly, the linking abilities of the Internet enable scientists, and other information providers, to display any new piece of information in relation to how it addresses knowledge gaps in a wider context. This is important as solutions to emerging environmental issues are rarely provided through the development of discrete pieces of information and technologies. Rather, the act of developing new ways forward is more likely to be characterised by the need for debate and ongoing information distillation and synthesis among different stakeholder groups concerned with the linkages between different pieces of information, management systems and scales.

The need for this debate is often not appreciated by scientists who often see the use of the Internet as yet one more way of 'getting the right information out there' (p. 160). However, as the farmer group leaders involved with our Tb vector control case study indicated, 'their vision for the Internet MIS was as a focal point around which to build more opportunities for farmer/scientist discussion and learning' (p. 160). This is consistent with the steps outlined in the ISKM framework for engendering a collaborative approach to generating and managing information, through which different groups and individuals interact to learn together and broaden their perspectives of the world.

With the recent evaluation of the Tussock Grassland MIS (http://www.tussocks.net.nz/), we can see that this case study has involved the use of all the steps and feedback loops outlined in the latest version of ISKM (Figure 12.1). It has resulted in a demonstrable information system that is being used in practice, and is being improved with user feedback and new information. From a research point of view the case study has contributed to the development of a participatory approach to information management that emphasises a number of key steps applicable to developing the understanding, knowledge and action needed to address environmental issues constructively. Looking to the future, ISKM can provide a common framework that enables action researchers working in different case studies to develop process lessons relating to the various steps involved. This is important if we are to learn lessons across case studies (pp. 219-220). It also provides a guide to help science leaders looking to improve the responsiveness of their research programmes to end user needs, and the subsequent participatory management of that information through to its provision on the Internet.

Even when science technologies (e.g. best practices, DSSs, models) have been developed with a high degree of awareness of stakeholder needs using processes such as ISKM, getting this information used to support management decision making at a wider level is still a major problem. Research teams can at best only work with a few representatives of stakeholder groups, providing limited opportunities for engendering social learning beyond this immediate level of engagement. This is particularly true in relation to many environmental management issues characterised by large geographic scales, many players, multiple perspectives on the situation and where science and other information is subject to diverse and contested interpretations.

In this regard, Internet-based material only provides the potential for different stakeholder groups to more readily access information. In the Tb vector control case study for example some group leaders and facilitators recognised that Internet-based material could provide a community resource which could be easily updated and shared with others - who may not have Internet access - through their involvement with groups (p. 160). The advantages of the technology are not in creating new 'virtual' communities, but in strengthening already existing social networks (p. 186). This is illustrated in Figure 12.1 which points out that processes such as ISKM need to operate in a social environment that supports learning.

This diagram highlights that while information is key to learning and subsequent informed and collective action, such learning will only happen at a societal level if it is supported by social capital (trust among the different players involved, mechanisms to develop shared understanding, and strong horizontal and vertical networks between agencies and stakeholder groups). In turn, this implies a need to ensure that the different interest groups involved have adequate capacity to participate in such multi-stakeholder processes. Therefore agencies seeking to support improved stakeholder participation in R&D, both at operational and policy levels, need to support both process outcomes (creating the conditions for participation) and task outcomes (getting information flows in place on-the-ground).

Exploring how agencies can achieve this increased level of societal capacity for participation is then another important area for profitable action research study. Because science agencies and programmes do not generally have the resources or mandate to work at this scale, this will require action researchers to work with environmental agency staff as they seek to support regional and national management and change initiatives. In these situations action researchers need to negotiate a role for themselves as evaluation specialists assisting those involved in such multi-stakeholder processes to assess progress and guide ongoing programme improvement. Such evaluations will also serve to build capacity to support improved participatory processes, as well as developing lessons that can be used to shape future initiatives.