Vegetation condition mapping and catchment management: The North Central Victorian experience. Ian Higgins, North Central Catchment Management Authority, PO Box 18, Huntly, Vic. 3551, Australia. Email:

Key words: landscape, monitoring, pattern, planning, priorities.

Introduction.  The Australian Government recognizes extent and condition of native vegetation as important surrogates for indigenous biodiversity (e.g. Hill 2001) and requires their use as two of the indicators of regional biodiversity status against which Commonwealth investments in regional biodiversity conservation can be evaluated.

These investments are largely directed to regional natural resource management agencies. Catchment Management Authorities (CMAs) have this function in Victoria. CMA funding resources cannot address all needs, so need to prioritize their conservation efforts across many patches of remnant vegetation. Using ‘top down’ planning has limitations when the planning body does not control or manage the land. Nevertheless, setting priorities from a theoretical perspective is considered worthwhile and still used in biodiversity conservation planning (e.g. Platt & Lowe 2002).

Using extent and condition as the basis of prioritization ignores variations in the type and significance of vegetation and habitat. (Victoria uses Ecological Vegetation Classes (EVCs) to classify native vegetation types.) To remedy this shortfall, the State of Victoria (2002) developed ‘Victoria's native vegetation management: a framework for action’ (VNVMF) to provide a shared understanding of the value or Conservation Significance of native vegetation. Under the VNVMF a remnant patch's Conservation Significance rating determines regulatory responses to clearance proposals. It is also a legitimate basis for prioritizing conservation actions. A major underpinning of this framework is the concept of Conservation Status, defined as the degree of threat of extinction to an EVC within a Bioregion.

Different methods for calculating Conservation Significance apply depending on what information is relevant to (or available for) the remnant patch in question. The VNVMF defines Conservation Significance as a function of the vegetation type's Conservation Status and its condition (OR habitat value for threatened species OR other values, e.g. land protection). The highest Conservation Significance rating calculated is the one applied.

Thus, vegetation condition mapping is extremely important to CMAs. It provides Conservation Significance based prioritization and a baseline against which changes in condition can be reported.

Prior to 2004, the North Central CMA only had Conservation Status mapping to inform its vegetation management priorities. Our aim here is to explore what difference vegetation condition mapping makes to mapping of conservation priorities; and potentially to investment decisions.

Methods.  Collaboration between the North Central, Mallee, Goulburn Broken and North-east CMAs led to funding for vegetation condition mapping across much (nine million hectares) of Northern Victoria. Within the North Central CMA's area, all of the Goldfields, Murray Fans, Murray Mallee and Victorian Riverina Bioregions and about one half of the Wimmera Bioregion were included. This covered around 87% of the mapped native vegetation of the region. The National Action Plan for Salinity and Water Quality (NAP) and the Natural Heritage Trust (NHT) in Victoria provided funding. The collaborating CMAs developed specifications for the condition mapping, employing the ‘Habitat Hectares’ condition assessment method described by Parkes et al. (2003). Habitat Hectare condition scores comprise both an internal site condition score and a spatial context score (encompassing size, shape and connectivity). The Arthur Rylah Institute for Environmental Research (ARI) successfully tendered for the project to develop the condition mapping. See Newell et al. (2006) for a more detailed description of the process and outcome. The ARI team supplied the vegetation condition mapping as a grid (raster) dataset. Each 30 m grid cell's value was the predicted Habitat Hectare score.

Development of a Conservation Significance based priorities map required the combination of the Habitat Hectare score with the Conservation Status of each EVC according to the principles of the VNVMF. The VNVMF specifies the range of Habitat Hectare scores for each Conservation Status category that a patch of remnant vegetation would need in order to achieve a particular Conservation Significance rating. For example, any vegetation patch of an EVC with an Endangered Conservation Status has a Conservation Significance of at least High, but if its Habitat Hectare score is more than 0.4, its Conservation Significance is Very High.

We intersected the vegetation condition (Habitat Hectare) information with pre-European EVC mapping, itself intersected with Victorian Bioregions (both Victorian government geospatial datasets) using ESRI ArcView 3.2 Spatial Analyst. Each grid cell was then assigned a Conservation Significance rating based on the combination of its Bioregional Conservation Status and its Habitat Hectare score.

We developed two maps of conservation priorities, one based on Conservation Status, the other on Conservation Significance. Each had a one (highest priority) to four (lowest priority) scale. Figure 1 shows how the priorities were matched to Conservation Status and Conservation Significance ratings.


Figure 1. Mapped native vegetation conservation priority across a portion of the North Central CMA using only Conservation Status (i.e. the threat of extinction to an EVC) (left) and Conservation Significance (i.e. a combination of information on conservation status and vegetation condition) (right).

Download figure to PowerPoint

We also classed the vegetation condition (Habitat Hectare) scores data into 10% equal interval condition classes as these match the thresholds used by the VNVMF framework's ‘rules’ for categorization of Conservation Significance. Using Microsoft Access 2000 s querying facilities we can prepare area summaries by bioregion, EVC, Conservation Status, Conservation Significance, vegetation condition class or any combination. This enables us to prepare a baseline for the indicator of condition preferred by the Australian Government (2004), viz. ‘. . . the proportion of each native vegetation type in each IBRA sub region that is estimated to be in specified condition classes . . .’

Results.  We produced two maps indicating conservation priorities: (1) A map of Conservation Status (based only on the threat of an EVC to extinction); and (2) A map of Conservation Significance (based on Conservation Status plus information on vegetation condition). The conservation significance ratings produced are a minimum. That is, with more information, such as knowledge of the presence of threatened species, the patch may have a higher significance. Therefore, within the accuracy of the mapping, the Conservation Significance should be at least as mapped.

Each map can be represented with four ranked conservation priorities. Large spatial and summary differences are apparent when the two priority maps are compared. A small portion of the North Central CMA region's mapped native vegetation is shown in Fig. 1. This illustrates differences in the spatial arrangement of conservation priorities that Conservation Status and Conservation Significance produce. The total areas of native vegetation assigned to each priority class also differ depending on the method used. For example, across our region, the total area of highest priority vegetation using the Conservation Significance method is only 17% of the total area of highest priority vegetation using the Conservation Status method. This is an artefact of the significance categorization scheme used in the VNVMF.

A set of baseline data was also prepared for the proposed vegetation condition indicator (see Australian Government 2004) for the majority of the region's vegetation types.

Discussion.  We are still exploring what ‘value adding’ vegetation condition mapping provides. It is already clear though, that for a relatively small investment an adequate vegetation condition baseline indicator can be prepared. The cost of the modelling and mapping exercise was around three cents per hectare across the studied parts of the four participating CMAs. In the North Central CMA, this was equivalent to about 18 cents per hectare of mapped vegetation.

Using vegetation condition mapping lends a Conservation Significance ‘view’ to our understanding of native vegetation in the landscape. However, the model that produces the condition mapping has a 51% predictive accuracy. Spatial inaccuracy of this order means that the vegetation condition modelling cannot be used at the small individual patch scale. Thus, it doesn't directly support decision making under the regulatory aspects of the VNVMF framework (although it has provided further evidence that our native vegetation estate is generally in very poor condition and reinforced the argument that further losses of extent or condition from human activity should be resisted, presumably through tighter regulation).

At landscape to regional scales though, the vegetation condition information usefully indicates how different vegetation types are faring. Conservation Significance mapping at these scales also provides a more meaningful view of where our priorities might lie.

For instance, in Fig. 1, several large blocks of native vegetation are only third priority from a Conservation Status perspective. The Conservation Significance view suggests these should actually be our second priority. Similar large areas of native vegetation now only remain on public land within the north central Victorian region. The CMA's current focus is on management of generally small patches of native vegetation on freehold land. Insights from our condition mapping may prompt re-examination of the levels of investment in public land native vegetation management for biodiversity conservation.

Using our method of incorporating information on vegetation condition into prioritization rigidly determines the total area of vegetation in each priority class. This may lead to undesirable proportions or unmanageable areas of high priority vegetation. It would be preferable to use a continuous numerical rating. This would represent Conservation Significance (and therefore priorities) as a spatial ‘surface’ (e.g. a contour map). Management agencies could then match the total area of high priority vegetation to the resources available for its management while still using transparent, defensible rule sets to determine where the highest priorities are located.

Acknowledgements.  Milos Pelikan of Spatial Vision and Peter McRostie of the North Central CMA for help with raster geoprocessing. Geoff Park of the North Central CMA, David Parkes of the Victorian Department of Sustainability and Environment and an anonymous reviewer for providing helpful comments on an earlier draft. Thanks also to Graeme Newell of the Department of Sustainability and Environment for helpful suggestions and encouragement.


  1. Top of page
  2. References