The previous sections have dealt with management targeted at biological aspects of biodiversity at the small, farm-scale. As we have previously emphasised, different species operate at different spatial scales and genetic variation within species can occur across their natural range. This is particularly important in rangeland areas where ecological and landscape processes are highly variable (spatially and temporally) and, consequently, the patterning of biodiversity is dynamic (Stafford-Smith and Ash 2005). It is therefore difficult to find a set of principles to guide the management of biodiversity at the landscape scale or that are applicable across each of the CMA areas. Because we have primarily discussed management that targets small-scale, paddock or whole-farm activities, it could be argued that such approaches are essentially only patching up the landscape. Local vegetation enhancement or revegetation activities, for example, could be seen simply as defensive strategies which do not consider the broader landscape.

We acknowledged early in this review that while National Parks represent an important component of species conservation, they fail to adequately address conservation of fragmented landscapes such as those found in agricultural areas or the conservation of genetic diversity across non-fragmented landscapes such as the rangelands. Where the conservation of biodiversity is integrated within agricultural landscapes (as described in previous sections) through changes in management (‘well-managed paddock model‘), it may slow the rate of biodiversity loss, but there is little guarantee that ‘best management practices’ will be maintained with changes in land ownership.

Thiele and Prober (1999) propose a Conservation Management Network (CMN) to provide a mechanism that takes advantages of both the National Park model and the ‘Well Managed paddock’ approach. A CMN, is a network of remnants, their managers and other interested parties co-ordinated through a single administrative point (e.g. the Grassy-Box Woodland CMN is housed within the Department of Environment and Conservation). The advantages of such an approach are numerous:

• Biodiversity is conserved at an ecosystem level and should capture ecosystem, species and population biodiversity
• Biodiversity is conserved at both the small and larger, landscape scale
• Multiple instruments for preservation can be employed
• Management can be co-ordinated across remnants, and management of one remnant can inform the management of another
• A single point for brokerage of information

The CMN model provides a mechanism to share information and responsibility for conservation of biodiversity at the large scale, both of which are vital ingredients for successful community-based projects (see Marais et al. 2007 for examples of project failure due to a lack of shared knowledge among stakeholders). Despite the shared knowledge created by the CMN model, major barriers still exist to large-scale restoration activities. Apart from funding, these include the ‘shifting baseline syndrome’, the scale and complexity of restoration and the long-term/open-ended nature of restoration. Manning et al. (2006b) suggest ambitious long-term ‘stretch goals’ and ‘backcasting’ or visualisation of the restoration endpoints as two possible techniques for approaching large-scale restoration activities. Currently, the potential use of software packages that allows the loss or restoration of vegetation to be visualised is being examined, and potentially offer a powerful method for the prevention of continued biodiversity loss and for restoration planning (A. Manning pers. comm.)

Recently, Lindenmayer et al. 2008 documented principles for the integration of biodiversity at a landscape level but acknowledged these are strongly context-dependent. That is, they cannot necessarily be applied in all landscapes and serve more as a check list (Table below).

Table: Guiding principles for landscape conservation (Adapted from Lindenmayer et al. 2008)

The importance of classifying the landscape to develop large-scale management guidelines has been highlighted in a number of recent publications (McIntyre and Hobbs 1999; Hobbs and McIntyre 2005). These authors suggest the use of a large-scale agro-climatic classification aligned to IBRA bioregions to identify landscapes with common characteristics, sharing similar threats and approaches to management (Hobbs and McIntyre 2005). At a smaller regional scale, patterns of landscape modification can be identified beyond a simple habitat vs. non habitat classification to ensure consideration of the degree of disturbance is incorporated into management objectives (McIntyre and Hobbs (1999)). For example, landscapes may be classified as intact, variegated, fragments or relictual based on the degree of habitat destruction, connectivity modification and pattern of modification of the remaining habitat.

How regional natural resource management bodies ensure that management at these larger scales is linked to management on farms remains problematic. The CMN model is one possibility but it probably requires administration at a higher level than the CMA (eg by a government Department) if all of the relevant landscapes are to be included. Briggs (2001) proposes local councils are the best bodies to provide the link between the two scales. Here, it is argued that councils have a statutory ability to undertake landscape and biodiversity management. While this has occurred in some situations (e.g. Ecotracks (2007)), this author also acknowledges that most council’s are under-resourced to meet such an obligation. It is however possible that alliance between CMAs, agencies and councils may support such a mechanism.

Environment Management Systems (EMS) many provide another model but their adoption within the agricultural sector has been slow, probably due to their complexity and the level of detail required for planning and monitoring. In some sectors they have been replaced with Best Management Practices (BMP). However, there a number of aspects of such approaches that may facilitate large-scale integration of biodiversity and production. For example, an EMS or BMP may incorporate an audit component that serves as a reference point against which changes in management may be monitored. EMS and BMP approaches may also embed catchment biodiversity plans allowing on-farm management activities to be couched within the broader regional planning.

It is doubtful if the costs and benefits of biodiversity are shared equitably among agricultural sectors let alone between urban and rural communities, given the difficulties in definition of biodiversity and the differences in its value to different sectors outlined at the beginning of this review. Within the agricultural sector there are difficulties in grasping the utility value of biodiversity. This is largely because the values of natural capital and ecosystem services supported by biodiversity are not costed and remain unappreciated, and the interdependence of agriculture and biodiversity unrecognised.