Norton and Miller (2000) argue that different parts of the landscape yield different production and conservation values and that land use can be rearranged to reflect these differences. To illustrate how this can occur, the simplest rearrangement may involve agricultural land managed specifically for commodity production (typically intensive enterprises using high inputs), and biodiversity restricted to areas set aside from agriculture (land sparing). The most complex rearrangement can incorporate heterogeneity (‘wild-life friendly farming’). In this latter example, patches of vegetation are retained, a diversity of crops may be planted, paddock sizes may be smaller, scattered trees or vegetation at paddock margins retained and, typically, conservation and production areas become less well defined (Fischer et al. 2008). A major contrast between these two extremes is that ‘wild-life friendly’ farming characteristically involves greater variability (heterogeneity) at smaller spatial scales than land sparing. However, the notion of heterogeneity may not simply be confined to a description of habitat heterogeneity (Oliver et al. (2007)) but extended to include ecological heterogeneity, where the landscape supports many ecological processes (Benton et al. 2003). Thus, where land capability varies and land use is to be matched to this variation, it also implies variation in land use or management. Where landscapes are homogeneous (biodiversity is low), common land use may exist. Thus, a central idea of matching land use to land capability is recognising that heterogeneity is the key to maintaining biodiversity.

There is a lot of discussion about the importance of maintaining or building ‘connectivity’ in the landscape and its relationship to the preservation of biodiversity. However, different species have different requirements in terms of connectivity (e.g. differing spatial arrangements and composition of remnant vegetation). For example, the common bird species Cacatua galerita (Sulphur Crested Cockatoo) requires open grasslands while Platycercus elegans (Crimson Rosella) requires woodland habitat (Dorrough et al. (2008a)) . Thus, maintaining the connectivity of, say, grasslands and woodlands may be of equal importance for the preservation of habitat to support a suite of species.

Although the specific habitat requirements of plants and animals may differ, if the environment is sufficiently heterogeneous at all spatial scales, different taxa will find their own habitats (Part and Soderstrom (1999)). Using birds as an example again, a recent British study found that complex landscapes with diverse vegetation structure and composition resulted in higher species richness with populations that were more stable or persistent (Devictor and Jiguet (2007)). In Australia, over past decades a good deal of time has been spent tree planting, largely as strip plantings along fence lines. While the biodiversity benefits of these activities may be debated, (see section on remnant management), this has essentially resulted in a similar landscape ‘design’, or management practice across whole catchments. Despite the benefits of tree planting in highly modified landscapes, the common land management practice of pastures with paddock perimeter tree planting is still working to homogenise landscapes (and minimise further heterogeneity).

However, ecological heterogeneity is also lost as agricultural intensification occurs. For example, inappropriate grazing management in the WCMA has resulted in widespread native woody shrub invasion and an associated decline in farm profitability and loss of floristic biodiversity. Here, the competitive interaction between woody shrubs and perennial grasses is shifted to favour unpalatable woody weeds as grazing pressures are increased. In the NCMA and BR-GCMA, intensification of cropping activities in some areas has led to large contiguous areas dominated by tillage under a common management, increasing the vulnerability of cropping enterprises to variable climatic conditions. The global evidence that links farm biodiversity decline to agricultural intensification and landscape homogenisation is clearly recognised (Benton et al. (2003) ; Millennium Ecosystem Assessment (2005)). Thus, the re-establishment of heterogeneity in agricultural areas may provide an improvement in biodiversity. Theoretically, all agricultural practices can be tailored to increase rather than eliminate heterogeneity. Currently, CMA’s are not focusing explicitly on the creation and management of heterogeneity on agricultural lands.

Where land use is to be rearranged, the capability of different land management units must be recognised. This can be facilitated by identifying differences in land capability either through training within courses that specifically match land use to land capability or as part of whole property planning. It is important to note that for the most productive agricultural areas (highest productive capability) there is likely to be a conflict between achieving biodiversity gains and losing agricultural production. The capacity to fence off areas to meet the management requirements for different enterprises e.g. for saltbush lamb production [Case Study 1] or for conservation [Case Study 2] is fundamental to the matching of land use and land capability on specific land units. There are other examples where fencing has been used with the specific aim of excluding additional grazing pressure from kangaroos or the exclusion of livestock for water point management [Case Studies 3 and 4]. While there has been some past criticism over the allocation of funds for fencing for biodiversity enhancement, in most case studies reported here part funding for fencing costs had been crucial in achieving both production and conservation/regeneration outcomes. ’Wildlife-friendly fencing’ essentially involves the removal/replacement of barbed wire or making it more visible (Wildlife Friendly Fencing (2008)). While there has essentially been no research evaluating the effectiveness of ‘wild-life friendly fencing’, many landholders we spoke to were using portable electric fencing without barbed wire.