Cape York Peninsula Land Use Strategy
Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.
Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.
Cape York Peninsula Land Use Strategy
Abrahams, H., Mulvaney, M., Glasco, D., & Bugg, A.
Office of the Co-ordinator General of Queensland
Australian Heritage Commission, March 1995
Method details and results.
The method considered wilderness, biophysical naturalness and the area of each mapped vegetation unit. This was undertaken by using the ArcInfo 'Relate' function, between the NWI coverage (a lattice) and the Vegetation coverage (polygon based).
The indicators were:
For each vegetation polygon three attributes were calculated as follows:
To select the best examples of each vegetation class using these indicators the attributes for each polygon need to be compared with appropriate characteristics of each vegetation class. To achieve this for each of the 201 communities, the mean and standard deviation of polygon areas, polygon Wilderness Quality means, and polygon Biophysical Naturalness means were calculated.
The individual vegetation polygon attributes were then compared to the vegetation class-wide characteristics to select the larger classes with better condition and integrity, these then representing the better examples of that community on the Peninsula.
The first trial of this analysis selected those examples that were larger or of higher value than the mean for each of the vegetation community characteristics. This trial failed to select representative examples of each vegetation class, with only 50% approximately of classes having polygons that satisfied these criteria.
This occurred because the distribution of polygon areas was negatively skewed, ie. the majority of vegetation communities were mapped with a small number of very large polygons and a proportionately greater number of smaller polygons. The mean value for the polygon areas is driven by these few very large values. Thus only a very few polygons are above the mean size for a particular vegetation class. In about 50% of cases these few large polygons were below the mean for either wilderness quality or biophysical naturalness.
A further analysis was run by:
(When considered collectively, at a value of about 0.75 standard deviations below the mean areas there was a strong break-point on the area distribution curve. Selection of all polygons larger than one standard deviation below the mean failed to discriminate on this area basis as some 80% of polygons were selected while selection polygons only larger than the mean restricted the choices to such an extent that this area condition swamped the other considerations.)
This analysis resulted in the selection of the better representations of vegetation areas for the majority of the vegetation communities (see Table A4.1).
Using the above methodology no representative polygons were found for twenty-five of the 201 vegetation classes. Fifteen of these communities are rare or restricted and so not surprisingly had a smaller number of polygons to select from. Further analysis of these was not undertaken as rare vegetation communities have been considered in detail previously. (Section Four of the main report).
For the ten more common vegetation communities (Table A4.2) the analysis was modified by dropping consideration of the wilderness quality characteristics. Thus employing two criteria instead of three. The analysis by using only two characteristics was still selecting the larger areas of best integrity as being characteristic of their class (see Table A4.2). This analysis successfully determined areas for seven of these communities.
Table A4.2 Representative Vegetation Classes
Requiring additional analysis
Vegetation D1 Polygon Total D1 Class Frequency Area Hectares 95 1 258,800.87 121 2 3,249.38 64 1 3,491.22 158 2 5,420.55 62 1 2,312.76 150 1 10,352.40 139 4 245.42
The remaining three vegetation classes (two of which are uncommon) were considered independently to identify the best polygons on the basis of the three variables. Vegetation classes 5 & 167 were uncommon and Vegetation Class 49 being common overall but not as the polygon dominant vegetation class. The rule set used for these classes is outlined in Table A4.3, following.
Table A4.3 Rule Sets for Vegetation Classes, 5, 49 & 167.
Vegetation Area Consideration Biophysical Polygons Class Naturalness Selected 5 > 0.75 STD below Mean Equal to the mean 1 rounded to 2 decimal points (4.22) 49 > 0.75 STD below Mean Equal to 3 (below 1 the mean, 3.02, for this vegetation class) and, independent of area Highest value 1 polygon (4.6) above mean (3.02) 167 N/A Equal to or 2 greater than 3. The mean for this class being 2.01.