SATS-Symposium-Peter Cullen

Matching the development aspirations of a burgeoning population with the biophysical constraints of our planet is the greatest challenge faced by the human race. Can this planet support a doubling of our population to 12 billion people; can Australia support a doubling to 40 million people? Will we know in time to take the necessary action? We are embarking upon a journey towards sustainability, without having a clear idea as to the end point of our journey. Like the explorers of old we need to navigate carefully to keep track of our progress and to let us adjust our course as we evaluate our progress.

Clearly there are social, economic and environmental aspects of this challenge, and while this paper addresses only the latter, it does so in the understanding of their interconnectedness and the need to develop institutional arrangements that integrate these elements.

I start from the premise that we are a long way from sustainability at present (Australian State of the Environment Committee, 2001). Consider the environmental footprint of Sydney or Melbourne as they grow seemingly uncontrollably towards 6-8 million people. The supply of water and sewage services will challenge current approaches and the management of air quality may well be beyond present transport and energy technologies. I acknowledge however that we have started this journey, and we know how to reduce many of the externalities of agriculture and energy production, even if our progress in implementation is slow.

The US National Research Council's Board on Sustainable Development (1999) identified water and air pollution as the top priority issues in developed countries, with ozone depletion and climate change also ranked high. For many of the less industrialised countries, droughts or floods, disease epidemics and the availability of local living resources are critical.

Australia State of the Environment 2001 confirms these issues as relevant in the Australian context, and identifies the following issues for unfavorable comment on our progress.

Atmosphere	Ozone depletion Greenhouse gas increased by 17 per cent between 1990 and 1999 Dust and particulates
Inland and coastal waters	Over-extraction of water. In 2000, about a quarter of Australia's surface water was classed as highly used or overused. Invasive species Salinity Loss of habitat Nutrient loads and algal blooms
Land and biodiversity	Net loss of vegetation cover. In 1999, about 470,000 hectares of native vegetation were cleared, an annual rate some 40 per cent higher than in 1991. Land degradation – erosion, salinity and acidity. In 2000, about 5.7 million hectares of land were affected by, or at high risk of developing, dryland salinity Invasive species

These issues are a mix of symptoms, processes and causes, and they are not independent. The clearing of native vegetation leads to a loss of biodiversity, accelerated soil erosion and hence dust, and to dryland salinity. These chains of impacts make prediction difficult. The simplification of our natural ecosystems to meet human needs leads to loss of habitat and makes them vulnerable to invasive species. These are the issues that have been largely beyond the capacity of our fragmented institutional arrangements to address.

Knowledge is fundamental to our move towards sustainability. Environmental science needs to move beyond explaining how things have gone wrong and how bad they are, to providing better signposts for future action. There are three aspects to consider:

In considering our knowledge needs, it is clear they fall into two categories. We need knowledge of our natural resources, and what is their present condition. Secondly, we would like a capacity to predict how the resource base might change in response to various pressures and management actions.

We have just completed a major national effort in the National Land and Water Resources Audit and the State of the Environment report. These are major exercises to draw together what we know about the condition of the natural resources of our country. This work has led to breakthroughs in modelling environments at the landscape scale as well as in developing protocols for managing and sharing data. It has produced a wealth of data, which awaits further analysis to convert it to knowledge.

In areas like that of river health we are developing exciting new measures that incorporate both biological outcomes (commonly of a structural nature in terms of invertebrate and fish populations) along with drivers of change such as flow, water quality and habitat availability. The audit has broken new ground here.

The audit has been difficult because of the absence of appropriate data sets. Monitoring the condition of resources is fundamental to adaptive management and yet has not attracted the best brains or the resources to do it well. The data is largely collected at the State level and is collected because of a recognised need to manage some already obvious problem. Often the data collected by State agencies is at different scales and degrees of resolution, making it difficult to assemble as national data sets.

The Federal Government has a legislated requirement to produce a State of the Environment report every 5 years, and has recently agreed to a continuation of the National Land and Water Resources Audit. The audit will build on our capacity to integrate data sets collected by the States and to present them in a user-friendly format to those seeking access. It is important that these two activities be better integrated so that the data collected by the audit is available in time to inform the interpretative report of the SoE process.

Finding 1
It is commendable that the Federal Government has agreed to an ongoing audit function that will collate data and present it in a compatible form. It is important that governments address some of the knowledge gaps identified in the State of the Environment report and the National Land and Water Resources Audit and ensure we start collecting some of this data. It is important that the audit and the SoE activities are seen as complementary.

Consider the calls to develop our northern rivers; very little ecological work has been done to understand these systems, and yet developers want to rush in and exploit them. They are likely to make mistakes as great as the ones made over the last century in the Murray-Darling Basin. It is surprising that an iconic river like Coopers Creek that traverses two States should only have two stream gauging stations.

Finding 2
It is apparent that there will be development pressures on presently undeveloped areas, and it is essential that we have an adequate understanding of these resources and how they might respond before investment is encouraged. State and federal governments need to identify potential development areas and invest in knowledge for at least a decade before development proceeds.

There is now a range of remote sensing technologies letting us address some of these issues in a cost-effective manner and advances can be expected. The geographic information systems (GIS) tools now available let us store and manipulate the large amounts of data that are becoming available. The web-based technologies developed by the audit mean large amounts of information are readily available over the internet at very low cost. The audit has broken new ground here.

But there are huge knowledge gaps. The collapse of support for the discipline of taxonomy means our capacity to even name many of the organisms we are attempting to manage is beyond us. We have a limited capacity to maintain national collections in many groups of organisms. Universities have withdrawn from this area because of the lack of job opportunities for graduates; we now hardly even have advocates for the taxonomic disciplines. The scientific community itself has not given this a high priority, and there are many who believe we get a better investment by studying processes.

To summarise, we know how to collect much of the data about resource condition, and have been developing some exciting new tools. The limiting factor is funding to undertake the data collection and analysis and an institutional structure that sees data as an important area for investment.

These are expensive and perhaps risky investments, and need to be undertaken by State and federal governments. It may be possible to recoup the costs of these investments through a resource rental arrangement in areas appropriate for development. In the absence of such government investment, the burden of proof for demonstrating sustainability should rest with those providing development.

Knowledge about the resources and their condition is necessary, but is not sufficient for decision-making. We need to develop a predictive capacity so that we can understand the long-term outcomes from various management alternatives in time to guide decision-makers before irreversible change is induced.

We would like to know how the condition of our various resources has changed over time. This is difficult because we generally do not have long runs of the appropriate data to let us see many of the trends of interest. Much of the data is very variable, being driven by climatic factors and it is difficult to detect trends from the background noise.

The concept that the environment provides a range of 'ecosystem services' has been popularised by Constanza et al. (1997), who provided preliminary dollar values for the range of services provided by different ecosystems.

There are some obvious ecosystem services provided by healthy river-floodplain systems:

The capacity of ecosystems to provide these ecosystem services is dependent on the biodiversity of the system. We have learned that when we simplify an aquatic system by transforming a flowing river into a series of weir pools we lose biodiversity, and we open up the system for domination by exotic invasive species like carp or other undesirable organisms such as blue-green algae. Conserving a wide suite of organisms in an ecosystem ensures there will always be some than can do well under any particular conditions of flow, nutrient status, temperature and light conditions. As we create aquatic 'monocultures' by our water management activities which stabilise flows and create a limited number of habitats we lose species so there may not be any organisms to take advantage of particular conditions that later arise (Gunderson et al., 1995). This puts the system at risk of domination by undesirable species, and this may then be impossible to reverse.

Part of this problem is our ignorance of both the species we might lose, and how they contribute to the functioning of the entire ecosystem. Who would have thought that a simple fungus might be critical, until penicillin was discovered? We just do not know what services many of the organisms at risk might be able to provide us with. Our ignorance is more profound at the level of the ecological community. Paul Ehrlich provided the analogy with the rivets in an aircraft wing. We can lose a certain number of rivets and the aircraft will keep flying; lose more and it may crash. Once crashed it is not possible to put it back together again.

Developing a predictive understanding in large complex systems is likely to require a paradigm shift, with the realisation that biodiversity and landscape functions are emergent properties of a complex adaptive system (Harris, 2002). Harris talks of a pandemonium of interactions producing great variability at a range of scales. Old notions of stability and 'equilibrium' are no longer appropriate as we appreciate the interactions and the complexity of the environment. These systems frequently exhibit hysterisis effects, with processes that may not be reversible, or which reach points of no return.

The predictive capacity we need requires ecological understanding. Ecology is the science of interrelationships in nature. It is an integrating science, not a descriptive science.

While we have a growing understanding of what are the key drivers of change in both our terrestrial and aquatic ecosystems, at this stage we have limited predictive capacity. This is due to several factors, including multiple causation, the interactions in the system, the stochastic nature of the climatic drivers and other ecological processes and the lag times between action and outcome. With dryland salinity, for example, the action of removing deep-rooted native vegetation may not show as salinised land for 30-50 years.

As an example, consider the challenge of predicting river health. We now have tools for assessing river health that incorporate biological outcomes (initially invertebrate populations, but now moving to include fish and other biota). These measures of river health commonly also include the drivers of change, which include water flow, water quality and aquatic habitat. We also understand the impact of a variety of human actions on river health, although we don’t as yet have simple functions relating causes to outcomes except in a most general way. If 10 per cent of irrigation water is taken from productive agriculture and returned to the environment what will be the outcomes in terms of river health? How should we deliver it to get the best ecological outcomes? We do know that if the water is delivered as a medium-sized flood it will lead to both fish breeding and bird breeding events, as well as have benefits to wetlands. But we do not need medium-sized floods each year, so the time base needs to be carefully considered, and will vary for different communities. Is there any point restoring some forms of environmental flows if the water is cold after storage in a deep and stratified reservoir? Is there any point restoring flows if the riparian area has been damaged? How can we advise on the most cost-effective interventions to restore river health in these multi-stressed systems?

There are exciting but challenging opportunities here for science to deliver tools for managers. This requires us to develop and validate models of various types that can simplify complex interactions and provide advice to policy makers. We may even link these biophysical models with social and economic models to build scenarios of the future to guide decision making.

Finding 3
Ecology must move beyond description of phenomena and identification of connections to developing a predictive capacity in multi-stressed systems.

I doubt that science as it is presently organised and funded can deliver the knowledge to let us move towards ecological sustainability. Much of our scientific effort is fragmented, with groups below optimal size, especially in our universities. Much of the funding that is available is short term and inappropriate for the questions being addressed. There is also a serious disconnection between the management agencies that conduct whole system interventions and the science teams that could assess conditions before and after, and define linkages.

It is inappropriate to study processes that might occur over time-scales of 10-20 years in 3-year projects. We have very few long-term ecological sites that can be used to benchmark changes. We have few long-term experimental sites where we can apply management interventions in a controlled manner to find out their impacts. These are difficult long-term experiments; difficult to design, difficult to negotiate with managers, difficult to fund and difficult to interpret.

Franklin (1989) has provided a useful typology of the sorts of ecological problems for which we need long-term research:

Finding 4
Developing a predictive capacity in the highly variable Australian environment will require patient investment in long-term ecological reference sites and experimental sites.

Identifying the appropriate boundaries for consideration of any environmental problem is also worthy of effort. In considering the contribution of agricultural science for instance, we see a science that has been a triumph at the paddock scale and a disaster at the landscape scale. The emphasis on short-term production at the expense of longer-term land degradation, and the almost total exclusion of consideration of externalities means that the loss of soil and nutrients from the farm has been inadequately considered. The downstream impacts of salt and agricultural chemicals have been largely ignored by agriculture, despite the fact that impacts are often on other farmers.

Finding 5
A systems approach to environmental problems will put more attention of defining appropriate system boundaries, both spatially and temporally. Scientists should argue the case for the scales proposed in a grant application on the basis of the phenomena being studied rather than on resource requirements.

Science owes much of its success to its ability to split problems into individual processes that can be studied in detail and in some isolation. We have developed scientific disciplines to enable cluster of scientists to work together. It is these disciplinary clusters that define the dominant paradigm at any time, and provide the quality control processes through peer review. Much of our present knowledge has been generated within this traditional disciplinary structure.

The disciplines are maintained by the structure of the training we provide to new entrants. University departments and scientific societies have developed to maintain this disciplinary structure. In science there is a propensity to splinter knowledge into sub-disciplines that seek to isolate and get their own departmental labels and specialist societies. The tribal nature of the peer-reviewing process also works to maintain ownership of certain types of knowledge within particular groups.

Many of the larger problems now facing society are not as amenable to solution through disciplinary research, and require the intellectual contributions of several disciplines if progress is to be made (Gibbons et al., 1994). We are moving beyond the simple cooperative model where disciplinary experts each take on some aspect of a problem and then work in disciplinary isolation with the occasional interconnection. The emerging trans-disciplinary relies on frequent interaction and stimulation across the disciplinary boundary. This is often difficult for scientists, and requires more attention to training in leadership and group processes. Scientists need to become comfortable in sharing their insights and preliminary thoughts on particular issues.

The new mode of knowledge generation poses challenges to universities in teaching people to work across, and to manage, the intellectual interfaces that we have previously largely treated as solid walls. The disciplinary structure of the universities produces isolated specialists who know almost everything about small bits of the whole system, but have no capacity to put this jigsaw together. The isolation of the various biological sciences from each other and from the physical sciences means there is little hope of developing a predictive understanding at the landscape scale.

We have been slow to develop the predictive understanding we need because we have not organised our ecological science effort to provide the large multi-disciplinary, whole system experimental studies at field scales. These experiments may take one or more decades to complete, and require the cooperation of management agencies carrying out large-scale landscape manipulations. Such studies appear beyond our present organisational and funding arrangements.

Finding 6
Major environmental problems may be best addressed through multi-disciplinary teams that have a high level of interaction as they work together to address the problem. This will require some innovative organisational arrangements for science institutions and funding bodies.

The Federal Government is establishing a new regional catchment basis for natural resource management, which is likely to guide both Federal and State Government investment in the coming years. Communities are being encouraged to develop regional plans based on the best available knowledge, that incorporates State and national priorities. If the plans can meet various requirements for accreditation, then governments may choose to provide block funding, not as grants but as investments to achieve defined outcomes. Payments may be partially dependent on measurable outcomes, providing new challenges for monitoring programs.

The technical challenge of assembling and delivering the appropriate technical information in ways that it can be useful to regional communities is an obvious challenge for science.

Publicly funded rural extension services that once provided technical knowledge have been largely replaced with private suppliers with a clear focus on profitability. Public good extension has moved to providing facilitation for group processes rather than substantive knowledge.

The technology transfer challenge is widely recognised and the general assumption is that scientists themselves should now take on this extension function. This assumes that scientists have the skills and motivation to do this work. It assumes that the research project itself is the appropriate unit of knowledge to transfer, and that it can be done within the life of the funded research. All of these assumptions are questionable.

In exploring how professionals seek new knowledge, we have found that the plain English summary of a research project is not the preferred mechanism. Recipients want the new knowledge to be embedded within the existing knowledge base and presented within the context of the decisions they make (Cullen et al., 2001). This sort of knowledge delivery is a specialised task and has led to the development of knowledge brokers in some organisations to prepare and deliver these materials. Such people are motivated to focus existing knowledge to solve a problem rather than selling their favourite research tool or their next set of experiments.

Finding 7
Delivering appropriate knowledge to those who can use it to guide their decision making is not being done effectively at present, and new approaches and appropriate resourcing are required.

We have a long journey ahead of us if we are to move towards sustainability. Science is fundamental to providing the knowledge we need to chart our way and yet is not providing what is needed at present.

We need to organise our science effort so that we can deliver science at the landscape scale rather than the plot scale. We often need longer-term science, at the level of decades rather than the traditional 3-year funding cycle. We need to be able to assemble multi-disciplinary teams to integrate the various disciplines that can contribute to environmental problems. We need smarter ways of packaging the knowledge we do have and delivering it to those who can use it.

We are dealing with complex interactions in ecosystems, often with long lag times. We can describe many of these systems but our predictive capacity is still weak. This is the challenge for environmental science in the journey towards sustainability.

Australian State of the Environment Committee (2001) Australia State of the Environment 2001. Independent report to the Commonwealth Minister for the Environment and Heritage. Canberra: CSIRO Publishing (on behalf of the Department of the Environment and Heritage).

Board on Sustainable Development, Policy Division, National Research Council (1999) Our common journey: a transition toward sustainability. Washington, DC: National Academy Press.

Constanza, R et al. (1997) The value of the world's ecosystem services and natural capital. Nature 387: 253-260.

Cullen, Peter, Peter Cottingham, Jane Doolan, Brendan Edgar, Christine Ellis, Melanie Fisher, Dianne Flett, David Johnson, Lynne Sealie, Sue Stocklmayer, Frank Vanclay and John Whittington (2001) Knowledge seeking strategies of natural resource professionals workshop synthesis. Canberra: CRC for Freshwater Ecology National Rivers Consortium.

Franklin, J F (1989) Importance and justification of long term studies in ecology. In Likens, Gene (ed.) Long term studies in ecology: approaches and alternatives. New York: Springer Verlag.

Gibbons, M C Limoges, H Nowotny, S Schwartzman, P Scott and M Trow (1994) The new production of knowledge. London: Sage Publications.

Gunderson, L H, C S Hollings and S S Light (1995) Barriers and bridges to the renewal of ecosystems and institutions. New York: Columbia University Press.

Harris, Graham (2002) Ensuring sustainability – paradigm shifts and big hairy goals. Enviro 2000 Conference. Melbourne.