To recapitulate: a number of important assumptions are underwriting the degradation of Australia's freshwater ecosystems. While all of these assumptions may well have once been correct, circumstances have changed over the last two centuries. Eleven of the most important of these assumptions are listed above in section 2.5. While some of these assumptions have been successfully challenged by the current water reform agenda, others remain more-or-less unscathed. Australian water managers have been slow to adjust to changing circumstances.

Of the critical management issues facing the water industry, I have argued that at least four need special attention:

• managing the cumulative effects of incremental water infrastructure development (in line with sustainability principles);
• the completion of comprehensive freshwater ecosystem inventories, and the development of systems of representative freshwater reserves;
• the integration of the management of surface and groundwaters, recognising that most Australian streams and rivers are fed primarily (most of the time) from surface groundwater; and
• a serious commitment to build quality assurance principles into water management processes, especially relating to compliance enforcement and auditing.

This is not to say that issues such as environmental flows or fish passage are unimportant - they are very important. To some extent they are components of the wider issue of managing cumulative effects. However, in my view, management processes around Australia are changing in response to some of these concerns. The four issues listed above are issues where management regimes have been particularly slow to respond to emerging needs.

Although management procedures to control cumulative effects are being considered by most States, no State has moved to implement all of the three critical components identified in Chapter 4.

On the question of the protection of biodiversity, the RFA program has at least established a system of terrestrial biodiversity reserves, and to this extent has gone a considerable way to meeting international and national commitments to protect terrestrial biodiversity. Fifteen percent of many important terrestrial ecosystems now lie in protected areas. However, the situation in the marine and freshwater environments is very different. Here, little progress has been made. Most Australian States (Queensland is the exception) have less than 1% of their marine waters allocated to protected areas, and an even smaller percentage to representative marine reserves. And in the freshwater area, the few representative ecosystems with adequate protection lie within large terrestrial reserves, such as those in Tasmania’s south-west World Heritage Area. The extent to which freshwater ecosystems encompassed in large terrestrial reserves meet CAR criteria has simply not been assessed in any Australian State.

Six of Australia’s eight jurisdictions have failed to establish reserves specifically to protect representative freshwater ecosystems, in spite of international and national commitments dating as far back as 1982, and in spite of commitments in the form of policy statements by all jurisdictions except South Australia and Tasmania (Tasmania has developed commitments in draft form). Victoria and the Australian Capital Territory are exceptions, as they are the only States to fund programs designed to put in place representative freshwater reserves. In Victoria’s case, the program (initiated by the 1987 State Conservation Strategy, and commenced in 1988) did not achieve its full objective of establishing a comprehensive system of representative reserves encompassing both still and flowing freshwater ecosystems (rivers and wetlands). However the framework established by the program does include fifteen Representative Rivers, and this framework could now be extended (given sufficient interest by the Victorian government) to establish a comprehensive and representative system.

While the limited Victorian reserves do not appear to be either comprehensive nor adequate (in the sense of providing coverage of all major representative freshwater ecosystems) the situation in most other States is that State governments have not adequately funded programs to implement even the limited commitments of the Ramsar convention (or the national biodiversity strategy) with respect to the establishment of comprehensive freshwater ecosystem inventories. All Australian States have developed wetland inventories, but even these, in all cases, are incomplete or inadequate when assessed by Ramsar criteria.

In the absence of comprehensive freshwater ecosystem inventories (and established systems of representative freshwater reserves) infrastructure assessment programs, looking (as they do) at discrete infrastructure proposals on a case by case basis (rather than taking a regional overview within a strategic planning context) will never be able to adequately protect freshwater biodiversity. Only State-wide strategies, based on comprehensive inventories, and including systems of representative freshwater reserves within each IBRA region, can be effective. Such strategies, where basin catchments cross State borders, need to be consistent across the whole basin.

Putting the issue of representative reserves aside for a moment, the other three issues of cumulative effects, the integration of the management of groundwater and surface water, and compliance enforcement and auditing - all need to be addressed within the processes and programs of integrated catchment management. This is currently not happening anywhere in Australia - with a few notable exceptions, mostly in New South Wales.

The only way to manage cumulative effects is to develop strategic limits on development, using sub-catchment, catchment and basin frameworks. Controlled developments should include dams, bores, levee banks, agricultural drains, and direct extraction. Clearance of deep-rooted vegetation, and expansion of irrigated agriculture, should also be included as the management processes become accepted.

These controls must be developed using a precautionary approach, and must be applied well ahead of the emergence of serious environmental problems - otherwise the human dimension of the ensuing problems becomes politically unmanageable. These three elements (catchment development limits, early application, and a precautionary approach) are critical – as argued in Chapter 4. This is the opposite of the current situation, where caps are only considered in 'highly stressed' catchments.

The obvious framework for setting caps already exists: integrated catchment management. Stakeholders must be involved. While States must provide clear, strong guidelines, stakeholder involvement and consultation is absolutely essential. States must also provide a statutory framework where catchment plans have a clear, important and effective role. Catchment plans developed on the basis of motherhood action statements and goodwill will not be effective.

The framework must also consider the issue of equity for those seriously disadvantaged - for instance by the wind-back of water allocations to provide environmental flows. In some cases, compensation must be paid.

Inventories and representative reserves

As set out above, Australia (and of course each of Australia’s constituent States) is committed to the concept of representative freshwater reserves at international and national levels. Moreover, Victoria, Western Australia, Queensland, the Northern Territory and New South Wales are committed by way of State strategies to develop systems of representative freshwater reserves.

However, these commitments are not being implemented in any focused way, in any State. Victoria has the most "advanced" framework, but even here the matter of freshwater reserves needs urgent revision. New South Wales began investigation of a freshwater reserves system, but the program has apparently been shelved for the moment. This may be revived by the soon-to-be released draft aquatic biodiversity strategy. Nevertheless, biodiversity surveys that will lead to a better freshwater inventory are progressing. Queensland is currently considering the development of a possible Rivers Policy which may carry the freshwater reserves concept. Western Australia and Tasmania are presently developing State biodiversity strategies, and Tasmania and South Australia (somewhat belatedly) are developing State wetlands strategies. These strategies could address the issue of representative reserves.

Policy commitments are one thing; putting a system of representative reserves in place, however, requires political commitment and funding.

Frameworks for the protection of freshwater ecosystems must, as Principle 8 of the national biodiversity strategy states, encompass both reserves and sympathetic management of remaining ecosystems. Additionally, reserves must extend beyond CAR reserves to encompass special-purpose freshwater reserves. Although a great deal has already been achieved in the establishment of existing wetland reserves, many important freshwater sites remain without adequate protection. Many of the values of freshwater systems can be protected to some degree by managing catchment and riparian condition, and protecting water flow and quality. However, some important site-specific natural values cannot be given adequate protection by these means, and consequently these areas need reserve status, coupled with specific management plans, to protect their important natural values.

Importantly, water flow and quality controls alone will not ensure fish passage, and are unlikely to protect the full suite of aquatic ecosystem values in the face of the incremental growth of small dams, bores and levee banks within catchments subject to vigorous human demand for water for agricultural, industrial and urban use.

Every Australian State should establish a system of comprehensive, adequate and representative freshwater reserves. In part, this will involve establishing "free flowing" rivers and streams. These reserves will not only protect the biodiversity of important representative ecosystems, but will provide benchmarks against which we can judge the effectiveness of our river management programs in other areas.

The first step to developing such a system of freshwater reserves has already been taken. All Australian States have wetland inventories, although (to a greater or lesser extent) these inventories remain unfinished. Many of these wetlands already have reserve status. The most urgent "next step" is to review and extend these inventories. Firstly, the review should be targeted at assessing the extent to which existing reserves provide a 'comprehensive, adequate and representative freshwater reserve system' - using the best data to hand. Such State reviews should use an agreed national classification method, and should aim to identify gaps, and provisionally select reserves to fill these gaps.

Secondly - again as a matter of urgency - existing inventories should be extended to encompass wetlands, rivers and streams, and aquifers. The development of comprehensive freshwater ecosystem inventories must, of course, include classifications to allow the selection of "representative types", and should also include condition indices and monitoring programs. As discussed above, there is a need to extend current work on ISC and ARC indices relating to stream condition to cover wetland and aquifer ecosystems.

Such inventories can then form the basis for the full development of systems of CAR freshwater reserves, and strategic conservation programs relating to infrastructure assessment regimes capable of managing the cumulative effects of incremental water infrastructure development.

As outlined above, the Commonwealth and several State governments have already made policy commitments to establish systems of representative freshwater reserves. These commitments should now be backed by funded programs to establish agreed procedures, and to identify and select representative reserves.

Integrated management of surface and groundwater

State governments have been, and to a large extent still are, managing groundwater and surface water as two separate resources. In some cases this is a reasonable assumption, especially when groundwater is confined deep underground, as is generally the case with the Great Artesian Basin. However, in most cases this assumption is simply not correct, as groundwater and surface flows are often inextricably and very directly linked. Australian rivers (unlike rivers of the Himalayas, which are fed by snowmelt, or rivers of southern New Zealand, which are fed by rainfall and snowmelt) most of the time, feed on surface groundwater. Yet the two aspects of the one resource have been, until very recently, managed by separate policies and programs, and sometimes even by different State agencies using different pieces of legislation. Fragmentation of the management of the resource in this way can only exacerbate the effects of the tyranny of small decisions, a powerful mechanism leading towards resource degradation and over-exploitation.

State water agencies must manage groundwater and surface water in a unified way. Water management plans, and water allocation plans, must be made with respect to the whole resource, and where significant linkage occurs, this must be reflected in total catchment/aquifer caps on water allocation. Groundwater environmental flows must be taken into account using a precautionary approach, to protect groundwater-dependent ecosystems. Aquifer condition assessment techniques must be developed, along the lines of those now in place for river assessment (for example, the Index of Stream Condition).

Compliance audit and enforcement programs

No system of resource management will be effective unless it is implemented, enforced as necessary, audited and reviewed. These are standard quality assurance techniques, built into internationally accepted standards such as the ISO 9000 (quality assurance) and ISO 14,000 series (EMS).

It is essential that programs be designed, and budgets provided, to audit and enforce controls. A situation where a farmer sees his neighbour build a substantial dam without the required approvals, and the State takes no effective action, undermines the entire water management regime. Under such conditions, cumulative effects become impossible to control.

Action by the Commonwealth

As discussed above, management of natural resources is - under the Australian Constitution - the province of the States, and the Commonwealth's means of influence is largely through the allocation of funding, as well as cooperative programs developed through consultative mechanisms like ANZECC and ARMCANZ. The Land and Water Resources Research and Development Corporation, the National Land and Water Resources Audit, the National Rivers Consortium, the Wetlands Program, and the Healthy Rivers Program are five freshwater programs within this latter category.

Bearing this in mind, and in light of the above discussion, the Commonwealth could take the lead with roughly sequential steps:

• The National Reserve System program could identify the issue for priority funding;
  
• The Land and Water Resources Research and Development Corporation, in conjunction with the National Land and Water Resources Audit, the National Wetlands Program and the Australian Biological Resources Study could fund a study (under a Commonwealth / State steering committee) to establish agreed classification methods for establishing major freshwater ecosystems types (and for measuring and reporting value). Such a project may also need to consider, as a first step, expanding or adapting the IBRA regionalisation framework (which is primarily designed around terrestrial ecosystems) to produce a regionalisation framework focused on freshwater ecosystems. Once those steps were underway;
  
• The National Reserve System program, and the National Wetlands program, could earmark NHT funding for States to develop comprehensive freshwater inventories as a basis for the establishment of identification, selection and establishment of representative freshwater reserves;
  
• The National Water Quality Management Strategy could, at the request of the National Rivers Consortium, expand its current "Implementation Guidelines" on ICM planning to incorporate the protection of biodiversity values into ICM plans. The Implementation Guidelines also need amendment to provide better focus on the management of cumulative effects;
  
• COAG water reform funding could be supplemented and/or earmarked to encourage the incorporation of biodiversity concerns into ICM plans, using the revised guidelines.
  
• The National River Health Program, in conjunction with the National Reserves System Program, and the National Wetland Program could allocate funding to assist States identify and select representative freshwater reserves, using the expanded inventories and revised methodology discussed above;
  
• The Land and Water Resources Research and Development Corporation and the National Wetlands Program could fund a second phase of the National Wetland Research and Development Program, focussing on improving on-ground wetland management for public and private wetland reserves;
  
• Management guidelines to address the difficult boundary issues involved in managing river reserves could also be examined by a project funded by LWRRDC, the Audit, or the National Rivers Consortium. National guidelines in these areas would assist in the development of management plans for representative reserves covering rivers in particular, although wetland and aquifer reserves also suffer from the impacts of land and water use within their catchments, well outside the actual boundaries of the reserves; and
  
• the National River Health Program could direct funding previously applied to determining river health to the question of aquifer health. This would be compatible with its objectives (see above).

Action by the States

At the State level, the most pressing concerns are:

• States should support and encourage the Commonwealth government to supply coordination and funding in the areas listed above;

• Existing wetland inventories should be expanded to include rivers, streams and aquifers, and resources should be allocated to support the joint Commonwealth/State classification methodology study suggested above;
  
• Groundwater and surface waters, when part of a connected resource, must be managed in a coordinated and coherent way, with water allocation and management plans taking account of both surface and groundwater stocks, flows, usage, threats and qualities;
  
• The States, through the National Rivers Consortium, should initiate the development of aquifer health assessment techniques. An Index of Stream Condition (ISC), developed in Victoria, is now being used in other States (discussed above), and is being extended by a current Audit project (discussed above). As part of the development of comprehensive ecosystem inventories, consideration should be given to developing similar indices for wetlands and aquifers;
  
• Discussions should be initiated in relation to expanding the current river environmental flow program to include groundwater environmental flows;
  
• Integrated catchment management programs should be modified to incorporate sustainability principles. With respect to the precautionary principle, this will involve managing cumulative effects by setting development caps on a catchment-by-catchment basis, within State and basin strategic frameworks - and well ahead of serious environmental problems;
  
• Integrated catchment management programs should be modified to incorporate sustainability principles. With respect to the principal of quality assurance, this will involve upgraded compliance audits and enforcement programs. For example, all States should audit and remove illegal dams; and
  
• Information on inventories, monitoring programs, and compliance and enforcement audits should be made readily accessible to all those involved with (a) water infrastructure environmental impact assessment, and (b) catchment planning.

Fish passage and environmental assessment:

While this paper has focused on four key issues, there are two key issues which the paper has not discussed in depth.

The first is fish passage. Many native fish undertake life-cycle journeys taking them from the estuaries at river mouths to the headwaters of these rivers. The construction of dams and weirs, generally speaking, makes these journeys difficult or impossible. Considerable advances have been made in the development of fishways in recent years; however this knowledge is generally not being widely applied (see Table 6.1 above). And fishways currently don't help fish travelling downstream over dam spillways. There is an urgent need to apply existing knowledge to facilitate fish passage, both through the construction of fish passes in dams and weirs of all sizes, and in simply removing unnecessary weirs. New dams, of all sizes, should be placed off-stream wherever possible, acknowledging the much higher construction costs involved. Compliance issues are also involved: on-stream farm dams are almost impossible to police with respect to allowed levels of water harvesting, while compliance enforcement is at least feasible in regard to off-stream dams.

The second issue relates to the environmental assessment processes applicable to large dams. Environmental impact assessments are traditionally confined to the direct effects of the dams themselves. There is an urgent need to extend assessments of large agricultural dams to encompass the long-term direct and indirect effects of both the dams and the irrigation proposals which the dams depend on for their financial viability.